If you ever see your name next to a 9 am workshop slot, beg for a different one.

That was my first lesson from Appsforum this month.

Getting people to a workshop at 9am is hard on day 2 of an event is hard. I forget that not everyone is an annoyingly over-energetic morning person like me.

But the people who did come were brilliant. Properly engaged, asking so many questions that we got through only one of the six exercises I’d planned.

And at the end, someone asked me where they could learn more from me, which made me all warm and fuzzy 😬.

That happened three times this month: at App Growth Summit, at Appsforum, and after my positioning workshop at the WiE Summit.

It’s my favourite signal, and it beats a full room every time. A full room tells you the title worked. “Where can I learn more?” tells you the session did.

(Yes, I’ve just described acquisition versus activation. I can’t switch it off.)

Here’s the rest of September.

What’s been happening with clients

I rounded off the growth audit this month.

Every audit is a big undertaking, and there’s always a stretch in the middle that’s pure information overload, where I’m piecing a hundred different things together into one story.

I sometimes feel guilty that I don’t do growth audits more quickly. With AI, that should be possible, right? But then Iqbal Ali reminded me of something at our Women in Experimentation Summit:

That the efficiency of AI comes at the cost of understanding.

The manual process of going through the data and piecing it all together is what helps me create a good audit. Automating that process would strip out the very thing that creates the value I’m delivering.

And the two founders made it a joy. Engaged, curious, and full of great questions. They also told me what had surprised them — in a good way:

I’m excited to see what they do next, and I’ve made them promise to keep me posted. I’m now rounding off another audit focused on improving trial-to-paid conversion rate.

Another audit!

I didn’t have long to miss working on the last audit before diving into a new one. The previous one was focused on identifying where they should focus to grow faster and profitably. This one was more specific: how to improve trial-to-paid conversion rate.

For specific challenges like this, I still like to look more broadly at the quantitative data, but I spend more time on qualitative research (user testing and reviewing user recordings) as well as a more in-depth heuristic analysis.

I’m excited to run through it with the founder today in a workshop, where we will discuss the key findings and prioritize the ideas further together.

Where I’ve been

App Growth Summit London. My first time attending. I met a wide range of people and caught up with Owen Farmer, whom I interviewed for a feature on how he scaled WordPal to nearly $1 million in ARR, starting at 13 and fully bootstrapped. Keep an eye out for that one.

Appsforum London. A panel on monetization and a workshop on activation (yes, the 9am one). I’m now revamping that workshop for App Growth Annual in New York next month, with everything the early risers taught me.

Online, hosting the Women in Experimentation (WiE) Summit. More than 800 people signed up, but more importantly (because signups are, of course, a vanity metric), the engagement in the sessions was amazing.

I introduced several of the speakers and ran my workshop on working out your positioning.

What I’ve been writing

The quiet article-publishing streak is over. We’re back in full force!

Lorelei, my lovely editor at RevenueCat, is back from parental leave, and now that the WiE Summit is done, I’m writing again for Convert.

Are you ready? It’s a lot.

Curse of the first renewal (RevenueCat). Whether someone renews is mostly decided during the first activation window: roughly five days for weekly plans, two weeks for monthly plans and the first month for annual plans. Miss that window and you’re relying on hope.

Why Your Pricing Tests Keep Failing (Convert). Pricing decisions stack like a pyramid, and most teams test the top layers while the foundations are quietly broken.

Why would someone pay for your app when AI does it for free? (RevenueCat,). Chegg lost half its stock value to ChatGPT while Duolingo kept growing. The difference is what a blank text box can’t do: structure, memory, habit, precision, connection, and blending physical with digital. There’s a two-minute Blank Box Test in there to help you work out which ones are yours.

“Not enough usage” is the churn reason nobody is fixing (RevenueCat). It’s one of the biggest reasons people cancel, and most teams answer it with a discount. But usage churn is a habit that broke somewhere: the trigger never fired, the action never stuck, the reward went flat, or the value never landed. A discount fixes none of those.

Money-back guarantee vs free trial (RevenueCat). A guarantee won’t replace your free trial, but on paid web funnels, it can beat one. Meta learns to find people who actually pay, and far fewer people claim the refund than you’d fear. How long to offer it, how to run it, and when to skip it entirely.

This month’s flip-flops

The W Stories podcast. I joined Viktoria Kharlamova and Tetiana Isaienko to talk about being the first employee at a growth agency, back before anyone really knew what growth hacking was, impostor syndrome, and why I recover on a mountain rather than a sofa.

The Price Power Playbook: Monetization Edition (Botsi with Ramdam). 42 monetization insights from the Price Power Podcast, each built around a real test. Mine is about a client who told me they weren’t a yearly product.

Their yearly subscriptions went from 16% to 39%. I’m in very good company, alongside Alice Muir, Dan Layfield and Marcus Burke, among others.

This month’s flip-flops

Every month I share not just my wins but my challenges, so here are my monthly flip-flops. This month, technology and I did not get along.

1. I recorded my workshop. Granola recorded nothing.

I definitely pressed record for the Appsforum activation workshop. Granola captured precisely none of it, leaving me with an empty recording.

Frustrating, because that recording was how I planned to spot what I wanted to change for New York. Thankfully, I scribbled notes straight afterward, so the revamp ideas survived.

2. I gave a five-minute intro to nobody.

The WiE Summit was online, and I spent five minutes warmly introducing the event before realizing I was still backstage… the switch to live hadn’t happened 🙈.

So I got to go ahead and repeat myself.

Maybe the lesson this month is simply to get better at clicking buttons…

3. I didn’t get a role I went for

This is a hard one to share, but I think it happens to so many of us that we should share it openly. I applied for a product growth role at a wellness app scaleup I was very excited about.

The role was actually more junior, but I was hoping there was a chance to reshape it to a more senior role. They were super kind about it, but ended up going for someone who better matched the level of the role.

What I’ve been enjoying

Starting the month in the Netherlands. Friends in Rotterdam, seeing my mum and grandma; it's always nice to go back.

A little break in France and Spain. I took a small week off and finally got to meet my new nephew. Given they live near the Spanish border, we combined it with a trip to see friends there.

Signing up for Hyrox. How can it be that, after talking about wanting to do one for nearly a year, it still felt like an impulse decision when we realized there were tickets left for London?

I was in an in-person meeting and came out to a Hyrox charge on the bank account. My husband had managed to get us a spot in the mixed doubles in London in December.

So we rushed to start training, beginning with a Monday 6:50am killer. My heart rate (I kid you not) has never been that high. My husband, meanwhile, has already asked me approximately ten times: “Why did we sign up for this?”

It’s reminded me how important it is to have a fitness goal. I can already notice that I’m taking more care with what I eat (I have a major sweet tooth) and becoming more conscious of my training.

Obsessed with paint-by-numbers. My creative outlet for the last few years has mainly been crocheting gifts for friends who are having babies, which, in your early 30s, is a lot. Lately, though, I’ve been trying my hand at paint-by-numbers, and I’m definitely obsessed.

It’s such a great ADHD-friendly thing to do while on a call with someone or listening to a podcast.

If you missed my newsletters this month

Looking ahead to October

Conference season continues, but with one trip instead of several.

App Growth Annual, New York: running the revamped activation workshop.

No other trips planned, which I’m quietly delighted about. I like my home time too — and I have big plans for the garden.

Then it’s Business of Apps in Berlin in November, where I’m still planning a dinner with people from the app growth space.

If you’ll be in New York or Berlin, let me know. I’d love to meet you.

I joke about the 9 am slot. I’d probably take it again, but in retrospect, I’d hustle a lot more to make sure I got people there.

At the end of the day, though, whether 2, 5, or 10 people show up, I’ll always give a 9 am slot my all.

I just hope I don’t have any evening people in the room, because they’re going to find my morning chirpiness very annoying.

Till next time,
Daphne

Full transparency: I have zero musical talent, no keyboard, and about six dead months of teenage guitar to my name.

The most musical I get is leading karaoke enthusiastically and badly.

I’m not an undiscovered talent by any means.

But I decided to test Simply Piano’s web onboarding anyway, screen by screen.

Here’s the thesis it left me with: you can’t dump the whole app’s how-to on them at once; you need to let it unfold step by step.

Most onboarding flows are full of screens that ask. Few are full of screens that use and guide you to understand the value step by step. Don’t worry, it’ll become clearer as we go.

So let’s see the key things they got right, and if you want the full onboarding flow, I’ve broken it down here. (Warning: it’s a big one.)

5 things this funnel does well

1. It reassures with proof, not platitudes

I told it I have no keyboard and have never played piano. Instead of a generic “don’t worry,” it told me how many other users are learning at my exact level.

A number beats a platitude every time.

2. It explains the product without a carousel

Instead of one swipeable screen listing six features, it uses a short sequence of single-idea screens: step-by-step lessons, songs you’ll know, five-minute sessions, visible progress.

One idea per screen, while the earlier reassurance is still fresh.

3. The profile screen proves it was listening

The in-between profile screen plays back the answers I’d given.

Most apps either skip this screen or fake it with a spinner and vague copy.

4. It gives a fast, specific win before the paywall

Right before asking for money, it lets me choose my first song from a set of options tailored to what I’d already told it. It pairs that with a concrete promise: I’ll be able to play it in one day. That’s far more compelling than a generic “start your journey.”

Nice touch: the piano in the background changes to the song I picked.

3 things I’d fix (if data backs it up)

Nothing’s perfect, so let’s discuss what I’d change on Simply Piano’s app.

1. The paywall stacks too many decisions into one screen

A free trial on the annual plan but not the monthly, fine. Savings shown clearly, fine.

But then a family plan appears with zero prior context; nobody had asked if I had one, and a trial-extension option gets layered on top.

Three decisions on one screen undo the “designed by teachers, here to guide you” reassurance stacked right above it.

2. Rounded numbers can decrease trust: “5,000”, “10+”

I personally find this looks fake. What actually convinced me was the row of real artist photos and names underneath. As well as stating the new ones added, something like “273 songs added in the last month” would prove to me that they have a wide range of ever-updating content.

3. Never answered how I’ll use it

I told the app I didn't have a keyboard, and while it did a good job reassuring me (see above), it never explained how I'd actually use the app without one. I know from the Sub Club podcast that there are online ways to practise, but this wasn't mentioned during onboarding, which left me unsure how I'd get started.

💡 What I’d recommend this week

If you love a good onboarding breakdown, so do I. You can find five more web-to-app breakdowns here.

Every screen in an onboarding quiz either collects information, uses it, or provides value.

Collecting is easy and cheap to build. Using it is what actually earns trust, and Simply Piano only pulls it off well most of the time. Not only that, they help you understand the app better as you go to get a feel for what it looks like and how it works.

If you’re auditing your own funnel this week, go screen by screen and ask which bucket each one falls into.

If more than half are “collecting,” you’ve built a survey, not an onboarding.

Speak soon,
Daphne

When you think of running price tests, what comes to mind? Adjusting prices up or down? Maybe testing discount levels?

That’s where most of us start, and it’s exactly the wrong place.

I’ve audited and experimented across a lot of subscription apps, and the pattern repeats: teams jump straight to price points and discount percentages while the foundations underneath are broken.

The test comes back flat, and everyone concludes “pricing isn’t the problem.”

It usually is. They’re just testing the wrong layer.

So in this week’s Convert article, I share the mental model I use: pricing decisions stack like a pyramid. Four layers, each one dependent on the layer below it.

Read the full article

The four layers

1. Value metric. What are you actually charging for? It sounds like a packaging detail, but it’s the single biggest pricing decision you’ll make. Does it grow with the value customers receive, or punish them for using the product?

2. Packaging. How your plans are structured. Every extra option at the paywall is cognitive load; every option has to earn its place.

3. Price point. The actual number, backed by willingness-to-pay research (Van Westendorp, MaxDiff) rather than guesses.

4. Optimization. Anchoring, urgency, discount framing. Legitimate levers, but they’re refinements, not foundations.

Most pricing problems live in layers one and two. Most pricing tests happen in layers three and four.

That gap is why so many teams run test after test and conclude their users are just price-sensitive. They usually aren’t. The pyramid is just upside down.

In the article, I break down what to test at each layer, why usage-based pricing can quietly drive churn even when the dashboards look great, and how to keep your price tests clean enough to learn from.

Read the full article

Recommendation

This week’s newsletter is sponsored by Convert, an incredible testing platform. If you have the traffic to run A/B tests, it’s a genuinely great tool.

And if you’re pricing an e-commerce store rather than an app, their piece on 5 pricing lessons learned from 500 A/B tests pairs well with this one: real data on which price tests move revenue, and which just add noise.

Learn more

Most teams don’t have a price-sensitivity problem, they have an order problem. Work the pyramid from the bottom up.

Till next time,

Daphne

The other week I went through the full onboarding for three competing manifestation apps back-to-back, prepping for a client session.

By the third one, I genuinely couldn’t have told you which was which, and not just because I was extremely overcaffeinated (yeah… detoxing coffee is still not truly happening; decaf is not the same).

Each featured:

• The same promise to change my life.

• The same generic demographic questions with zero explanation of why they mattered.

One conversational flow asked about my sexuality before it asked what I actually wanted from the app. None of your business, by the way; I’m happily married, and my love life is not the area I’m here to fix.

Then, when I answered "I don't know" to one of its big, existential questions, it chirped back: "That's a good start! Expand on that in one or two sentences."

I just told you I don’t know…

Yes, this is a sample size of one. But it's also why this matters more now than ever.

Your users are doing exactly what I did

Over 14,000+ new apps launch every month.

Downloading and trying one has never been lower effort. So users treat your app the way I treated those three: install, poke around, compare, leave, and sometimes come back.

The State of Subscription Apps 2026 data shows it clearly:

• Monthly reactivation jumped nearly 50% in a year, from 13.7% to 20.1%

• AI apps earn a 41% higher first-year LTV per payer ($30.16 vs $21.37), but their monthly plans retain 36% worse over 12 months

• Users try, get disappointed elsewhere, and circle back: reactivation driven by comparative value, not life events

Churn used to be a funeral, now it’s an audition. They’re not gone; they’re checking whether anyone in your category actually deserves them.

We have higher standards for what we let take up storage on our phones than we do for what we let take up space in our dating lives.

Three mindset shifts to improve your app

1. The audition happens in the first session.

55.4% of 3-day trial cancellations happen on Day 0. Not Day 3, but Day 0.

If someone can’t tell within one session what you’ll do for them that the other apps won’t, they’ve already toggled off auto-renew.

2. “As good as the competition” onboarding is now a losing position.

If your onboarding feels like the category default (like the three I tested), users have no reason to pick you over the app they tried yesterday.

I review apps with the following question in mind:

What would I remember about this app tomorrow?

3. Your competitors’ one-star reviews are a free roadmap

Everyone copies what the category leader does well. Almost nobody reads what their users say they do badly.

That’s where the “come back to us instead” opportunity lives, in the thing the whole category is failing at that you could actually fix. I still meet too few apps investing in review mining.

The uncomfortable part? This cuts both ways.

Your churned users are a warm list, not a dead one, but only if something is different when they come back.

A win-back email pointing at the same onboarding they already left is just asking them to churn twice.

Users have never been this unfaithful. That’s not the threat it sounds like; it’s the biggest second chance the market has ever handed you.

Make the initial audition count, it’s your chance to stand out.

Worth a review (💡 What I’d recommend this week)

I’m not going to name and shame the apps; instead, I'll point you to some that I think do this well.

These are web-to-app funnels (to save you downloading a load of apps), but the principles apply. Here are two examples:

1. Reading.com

Great at bringing the actual app experience into the flow; I talk about it here.

2. Jumpspeak

Originally, I wanted to include this in my “5 web-to-app funnel examples” article, but I was trying not to make it all about quizzes, so they sadly got left out.

The first part feels fairly generic, but pay attention to how the copy mirrors your profile back to you.

It creates the feeling that the app understands your situation, your needs, and the level of support you’re looking for, making the experience feel much more personalised.

Till next time,
Daphne

When teams ask me about freemium, they usually jump straight to the same question: what should we offer for free vs paid?

It’s the wrong first question.

Before you map individual features to tiers, you need to decide something more fundamental: how does your free tier relate to your paid tier?

Get this wrong, and you end up with a free experience that feels either too stingy (users never get enough value to care) or too generous (they never need to upgrade).

Get it right, and the feature mapping becomes obvious.

There are three broad architectures. Each suits a different type of product.

1. The taster model: same product, with limits

Here, users get the full product experience up to a certain point.

The free version doesn’t feel stripped down — it feels complete enough to use and get real value. The upgrade trigger comes from usage growth, not from missing features.

Loom is a great example. I used it free for about a year, maxing out at 5-minute recordings and 25 saved videos. I didn’t use it often, so I’d occasionally delete older videos to make room.

Over time, I started seeing the value — and I reached a point where five minutes felt frustrating, and sending four separate videos to a client didn’t exactly scream professional. I upgraded and have been a loyal customer for years.

Zoom works the same way (40-minute limit on free calls), as does Notion (same editor, but with block limits).

This works when:

• Your product’s value increases with usage volume → more storage, longer recordings, more messages

• Your core value is what drives word of mouth, so free users need to experience the real thing

• There’s a natural inflection point where casual use becomes serious use

The risk: set limits too high, and users never hit them.

Set the limits too low and the free tier feels like a trial with extra steps.

2. The split model: different features for different jobs

Here, free and paid users get distinctly different capabilities. Premium isn’t “more of the same” — it’s a different set of tools for a different use case.

CapCut does this well. Basic video editing is free, but as I edited and wanted fancier subtitles and better tools, those were clearly premium features. I could edit without them, but for business use where I wanted polish, paid made sense.

Surfline, the surfing app, works similarly.

Casual surfers only need a basic forecast — the free app handles that fine. Serious surf fanatics want extended forecasts, wave timelines, and detailed swell data. As a casual surfer myself, I just want to know whether I can surf (and hope to be able to stand up).

But as those casual users fall in love with surfing, they may become paid users later.

This works when:

• You serve distinct user segments with different needs — casual vs power user, personal vs professional

• Paid features have a meaningfully higher value that justifies a separate tier

• Free features are strong enough to stand alone, not just a teaser for what’s locked

The risk: if the split feels artificial — if users constantly bump into locked features during their normal workflow — it creates frustration.

Locked features should feel like a different destination, not a roadblock on the same path.

3. The hybrid model: taster limits + split features

Many of the best freemium apps combine both approaches.

Some features are available to everyone with usage limits, while others are exclusively premium.

This gives you two types of upgrade triggers: “I need more of what I already use” and “I want access to something new.”

Slack does this. You get the full workspace experience, but message history is limited to 90 days and you’re capped on integrations (taster). Yet Slack Connect — cross-organisation channels — is fully premium, because it’s a business feature that signals a team should be paying (split).

ChatGPT works similarly. Free users get capable models with usage limits (taster), while advanced reasoning tools and higher-volume image generation are premium-only (split).

Duolingo too: free users get all lessons with ads and limited “hearts” (taster), while unlimited practice and offline access are premium-only (split).

This works when:

• You have both a broad casual audience and a smaller power-user segment

• Some features scale with usage while others are binary — you either need them or you don’t

• You want multiple upgrade triggers at different points in the journey

How to choose

If you’re unsure, these questions help:

1. Does your product’s value increase with usage volume? More messages, more storage, more recordings? Lean toward the taster model. Set limits where casual use flips into serious use.

2. Does your product serve clearly different user types? Hobby vs professional, individual vs team? Lean toward the split model. Gate features by segment, not usage.

3. Is your user base broad, with a wide spectrum of needs? Lean toward the hybrid model. Use taster limits for the core experience and split features for power tools.

And remember: your architecture can change over time. It’s much easier to tighten a generous free tier than to loosen a restrictive one.

Recommendation

This newsletter is adapted from a longer piece I wrote for RevenueCat on freemium strategy.

If you’re actively working on your freemium model, it covers the full process: deciding if freemium is right for you, defining your goals, mapping features to tiers, and creating a “Bill of Rights” for your free users (a concept I love from Life360).

Read the full guide

The feature mapping matters, but the architecture shapes everything.

Daphne

I turned 33 this month. And instead of the usual thanks-for-the-wishes post, I asked for something.

Full transparency: I miss being part of a team.

Consulting gives me so much variety, but I miss being in the room, learning from the people around me, and seeing one product all the way through.

So I’ve quietly been making room in my portfolio for one bigger engagement. Going deeper with one subscription app, ideally wellness or impact, a team of 15 to 100, a product growth role.

More days a week, part-time or full-time, remote or hybrid in London. I’ve done this before. It’s the work I love most.

I double-checked that post with three people; as writing “I miss being part of a team” out loud felt like admitting something wasn’t working. (It is working. I just want a different shape for part of it.)

I never expected what would happen next, as people really showed up:

• Suggestions sent by people I’ve never met.

• Intros I’d never have found on my own.

• Two conversations I’m still in.

• A few messages that were just “same, actually.”

None of that happens if I post “33 learnings of this year”

So that’s the theme of this month, and honestly, the thing I’d most like you to take from this issue: be specific out loud.

Whether you want a role, a speaking slot, a co-founder, or feedback on your paywall, people are generally willing to help.

They’re just not psychic.

The more precise you are about what you’re after, the easier you make it for someone to hand you the exact thing.

Nobody can help you with a secret. The best things that have ever happened to me, I’ve dared to ask for.

Storytime over, now here’s the rest of August.

What’s been happening with clients

I’ve started an audit with a new client, and I’m loving it. They’ve built a custom in-house tool that's been great to use and nerd out over.

It’s also where I’ve had to hold myself back. There’s so much in there. My instinct is to keep pulling threads and follow every interesting one.

This month I’ve been actively practicing not doing that. Depth on the questions that matter, rather than depth on everything.

I love doing growth audits, and given that I’ve been focused on keeping space, they are the main type of projects I’m accepting right now.

What I’ve been presenting

Andrew Capland and I ran Rise of the Growth Engineer on 19 August, a workshop on how growth roles are being rewired right now.

The career ladder is looping rather than climbing, and we mapped four archetypes: the founding growth hire, the player coach, the strategist, and the growth engineer.

The questions were the best part. Three worth repeating:

“Do you need a coding background to be a growth engineer?”

No. Andrew’s example was someone shipping to production without one. Curiosity and the confidence to try things matter far more than a CS degree.

“How do businesses actually perceive growth profiles?”

Many still quietly mean marketing when they say growth. (Regular readers will know this is the hill I’ll die on)

The move: define your version of growth before the conversation gets going, then test for about 90% overlap with what they actually need. If it’s 50%, that’s not a negotiation; that’s a different job.

“Isn’t there a risk companies hunt for a unicorn and under-hire?”

Yes, and this is the honest one. It’s the growth hacker myth of 2015 in new clothes (I’m still ashamed I used to call myself a growth hacker all those years ago).

Growth engineers tend to be strong on data and automation and lighter on customer research and strategy. You get quick wins and a slow strategic mess.

If you missed it, don’t worry; there is a recording.

What I’ve been writing

Three articles were written this month. None of them are out yet, but I am hoping to share a few new ones next month:

Why would someone pay for your app when AI does it for free? (RevenueCat). Chegg and Duolingo faced the same threat but had opposite outcomes. This will share 6 ways to differentiate from a blank text box.

A guide to physical products and apps (RevenueCat). Five levels of physical, from brand merch to membership-native hardware. This was a dream to write with my e-commerce and app background.

Why Your Pricing Tests Keep Failing (Convert). Pricing decisions stack like a pyramid, and most teams test the top layers while the bottom ones are quietly broken.

I’ll share each one here when it goes live.

What I’ve been up to with Women in Experimentation

Deep in prep for the WiE Summit (9–10 September). I’m organising it with the WiE team and Convert, introducing some of the speakers, and running a short session of my own on how to work out your positioning, whether you’re in-house or consulting.

It’s a topic I end up advising a lot of fellow consultants on. Positioning gets treated as a branding exercise when it’s really a decision about what you say no to.

Sign up for free to attend.

Sign up

This month’s flip-flops

Every month I share not just my wins but my challenges, so here are my monthly flip-flops:

1. I tried to shorten an onboarding that didn’t need shortening. We were revamping onboarding and seeing drop-off, so I did the obvious thing and started cutting questions that weren’t adding value.

The founder stopped me. The problem wasn’t how many questions there were. It was one question, and the mental load of answering it.

Onboarding length isn’t a question count. It’s the intensity of what you’re asking, and that goes up fast when the topic is sensitive, or emotional, or full of terminology people don’t have yet.

Four easy questions and one hard one is not five questions. It’s one hard question with four warm-ups.

Simple wins, and this was a good reminder of it.

2. I got disproportionately embarrassed about one slide.

During this month's webinar, a slide in the presentation was missing some information. Small. Nobody died. And I felt genuinely embarrassed, in a way that was completely out of proportion to what actually happened.

I think that’s new, and I think AI is why. When the tools are this good, a mistake stops reading as “you’re human” and starts reading as “you didn’t even check.” The irony is that I ran that deck through Claude 3 times and checked it myself…

The bar for what counts as sloppy has quietly moved, and I’ve absorbed that without ever agreeing to it.

I’ve been rereading The Perfectionist’s Guide to Losing Control by Katherine Morgan Schafler this month, which is either good timing or the reason I noticed at all.

Her argument isn’t “stop being a perfectionist,” which never works and mostly just adds guilt on top. It’s that there are different kinds of perfectionism, and the job is to work with yours rather than against it.

What I’ve been enjoying

The birthday itself. My husband planned a surprise weekend away, which was absolutely amazing. New Forest and the Isle of Wight, lots of dog walks, and no laptop.

Doing more things with other people. The webinar with Andrew. Connecting with Ioana Iordache to discuss collaborating and planning a dinner in Berlin before Business of Apps in November, as well as continuing to work with Alice Muir, which has been a real pleasure.

Freelancing gets lonely, especially for someone as extroverted as I am.

Online isn’t the same as in person, but it’s a lot better than nothing, and I’ve stopped treating these as a nice-to-have.

I’ve also gotten to see Rosie Hoggmascall in person more. My goal is to drag her out of the office in time, but it’s as much for me as it is for her. It’s nice to have friends in the space who get it.

We even did yoga in the park the other day, followed by a lovely picnic she’d arranged.

If you missed my newsletters this month

Looking ahead to September

You’re receiving this while I’m working from the Netherlands, catching up with friends and family.

Then the first week and a half of September is conferences, back-to-back:

• App Growth Summit London (3 Sep): attending, purely as a human in the audience

• Appsforum London (7–8 Sep): running a workshop and hosting a panel

• WiE Summit (9–10 Sep): organising, introducing speakers, and my positioning session

It will be a busy few months conference-wise.

Then a workshop at App Growth Annual in New York in October, followed by another at Business of Apps in Berlin.

I’m also planning a dinner during the Berlin trip with people in the app growth space, because I badly underestimated how big that scene is there, and I’d like to meet some of them IRL.

If you’re at any of these, tell me. I’d love to meet you. (See: entire theme of this newsletter.)

Also in September: the audit gets handed in, discussing a few new potential bigger projects / roles, and I finally get to meet my new nephew in France!

In July I wrote about deliberately doing less. This is what I used that new space for. Not more things. Better ones, and more of them with other people in the room.

And if there’s something you want this coming autumn, say it out loud to someone, specifically.

Worst case, nothing happens.

That’s also what happens if you stay quiet.

Till next time,
Daphne

A few weeks ago, I mentioned that one of my clients had been featured on a podcast.

Time to tell the full story.

The client is Natal, a pre- and postnatal fitness app that I support in its growth alongside Alice Muir.

Their founder, Nancy Anderson, appeared on Sub Club — my favorite podcast in the subscription app space (I’ve had the honor of being on it myself, both in a regular episode and more recently a Sub Club Live), which made the whole thing even more exciting.

And there I was, out walking my dog, listening to someone I work with casually share a number most founders would probably frame and hang above their desk:

They removed their free trial. Monthly subscriptions grew 2,000%.

Yes, you read that right. Not 20%. Two thousand.

Hearing it made me weirdly proud, like watching a friend absolutely nail a wedding speech. I can’t always share client numbers, so when a client like Natal is open about both their challenges and successes, I can’t help but celebrate it.

Everything below comes from that episode. These were the biggest lessons and moments that stood out to me from Nancy’s story.

1. The free trial was a delay, not a converter

Natal removed the free trial from their monthly plan entirely; that is the irony of it all. Monthly subscriptions increased by 2,000%, quarterly subscriptions by 46%, and annual subscriptions by 21%. It worked so well that they made the change permanent.

My favorite part? The idea wasn’t even theirs.

They heard it on the same podcast, in an episode featuring the Zumba team. Nancy’s reaction was basically: “Huh, that makes sense. Let’s try it.” Even she was surprised by how well it worked.

So why did it work?

Because by the time a Natal user reaches the paywall, the selling is already done. 93% of people who reach their web checkout go on to download the app. Their trial conversion rate is around 68%, compared with an industry average of roughly 38%.

When someone arrives that convinced, a free trial doesn’t necessarily build confidence — it just delays the commitment they were already prepared to make.

Now for the uncomfortable flip side: if your users are arriving cold from a Meta ad and your onboarding experience has to do all the persuading, removing your trial will probably hurt. A lot.

The lesson isn’t “remove your free trial.” The lesson is that the conviction built before the paywall matters more than the paywall itself. What’s interesting is we are now taking this a step further and testing around paid trials instead for the web-to-app funnel.

2. The payment method that unlocked the “I can’t afford it” audience

This is one almost nobody talks about, but it’s a powerful lever in the US:

Natal started accepting HSA payments (Health Savings Accounts), which allow eligible users to pay for healthcare-related expenses with pre-tax money. In practice, that can make the app feel 30–40% cheaper for those users.

The demand was hiding in plain sight. For years, women had been paying full price and then manually submitting receipts to their insurance providers.

Nancy heard about it constantly through her DMs. So she wanted to help make it easier for those women to take advantage of an HSA.

Here’s the part I love: they launched it quietly. No big announcement. No campaign. Nothing. Within two weeks, tens of thousands in revenue came from people simply discovering the option at checkout.

Think about what we usually mean when we talk about a “growth lever”: a new channel, a new feature, a new acquisition strategy. This was none of those.

It was a payment method that converted “I want this so badly, but I can’t afford it” messages into subscribers. There’s even a built-in seasonal push: HSA dollars often expire at the end of the year, creating a natural Q4 urgency around using those funds before they disappear.

(Yes, this is US-specific. But the underlying question applies everywhere: who already wants to pay you, and what friction is stopping them?)

3. The lever that no dashboard can measure

Nancy has tried to outsource content production many times. It keeps not working. It keeps not working. In a niche as sensitive as pre- and postnatal health, words matter. A lot.

And nobody can capture her voice, expertise, and understanding of her audience quite like she can. Having been in many meetings with her, I’m still in awe of her knowledge and passion for what she does.

So she kept doing it herself. In the podcast episode, she shared that a recent audit showed their organic content outperforms competitors by 90%.

Her most honest line from the episode was that growth would be easier if she “could get tech bros to buy into” trust, because trust is almost impossible to quantify in a meeting. You can’t A/B test it. You can’t put it neatly into a spreadsheet.

But you can see the impact everywhere downstream: the 68% trial conversion rate, the paywall that works without a trial, the checkout visitors who actually go on to download.

Founder-led content can be slower and difficult to scale. It’s also almost impossible to attribute, but it’s powerful. We saw this at Heights too, when I was Head of Growth; much of the initial traction came from founder-led initiatives.

Worth a listen (Recommendation)

The full episode is genuinely worth your commute or dog walk: How Removing the Free Trial Grew Monthly Subs 2,000% — Nancy Anderson on Sub Club.

Nancy is brilliant at explaining why she still writes her own content and reads her own DMs. (And no, she didn’t ask me to share it; I offered because I’m so proud of what they’ve built.)

If you want the origin story of the no-trial idea, the Zumba episode of Sub Club is where the Natal team first heard it. A nice reminder that good ideas are lying around everywhere — the skill is knowing which ones fit your app.

Please don’t copy Natal

I know, I know. After 800 words or so of “look how well this worked,” here I am telling you not to do it.

The irony is not lost on me: Natal’s biggest win of the year came from copying a podcast tip.

And it worked because it built on years of trust-building. Their audience was arriving at the paywall already convinced. The exact same test on an app fuelled by cold paid traffic could go very wrong.

I’ve spent years helping founders improve their trial-to-paid conversion, but I’ve realized we need to be more critical and ask whether the trial should exist in the first place.

That’s the real takeaway: your monetization setup is full of defaults you’ve never questioned. Whether that is:

• Trial length

• Trial existence

• Plan mix

• Payment methods

• Pricing

So don’t copy the answer; rather, copy the behavior: look at your data, understand your customers, and keep testing.

Till next time,
Daphne

A quick extra email from me this week, because I'm excited to be hosting a webinar this Wednesday together with Andrew Capland.

Recently, I pulled apart 20 live Head of Growth job specs and compared them to the same analysis I ran in 2022.

It made me realize just how much the Head of Growth role has changed and that we need to be talking about it more. Andrew had similar insights from coaching 100+ growth leaders.

Together, we realized it is now 4 roles. So 4 people with the same job title:

1. 1 that is the founding growth hire: broad scope, no playbook, nobody above them who’s done the job.

2. 1 leads a small team of six and still jumps into the hands-on work.

3. 1 advises three companies and focuses on strategy.

4. 1 is the entire team.

I’ve been at least two of these over the last few years (maybe 3).

The title survived the last years, but almost nothing underneath it did:

• The execution work (PPC, SEO, offline) has mostly left the spec.

• Strategy, experimentation, and unit economics stayed.

• AI-native is, of course, on the rise.

• And the salary range for this one title now runs from £70K to £250K, because it isn’t one job anymore. It’s four.

Knowing which version of the job you’re in or need to hire — changes everything about what you should learn next.

So this Wednesday, Andrew Capland and I are breaking it all down live:

• The 4 roles the Head of Growth job is splitting into, each told through a real person’s week

• What the spec data says you can stop worrying about, and the few skills that now carry the whole job

• The career ladder bending into a loop, and the senior IC roles that didn’t exist a few years ago

• What to do next based on which version matches you

• Plus live Q&A

Wednesday 19 August · 5pm UK / 12pm ET · Zoom · 60 min · Free

Save your seat

Can’t make it live? Register anyway; everyone who signs up gets the recording and a one-page cheatsheet/summary.

And if you’re a founder rather than a growth leader, come anyway.

You’re probably about to write one of these job specs, and hiring the wrong one of the four is an expensive mistake.

We promise no “AI WILL TAKE YOUR JOB” fear-mongering, just valuable and actionable insights.

Hope to see you there,

Daphne

Inviting you to my free webinar with Andrew Capland on August 19th at 5pm (GMT+1), all about the rise of growth engineers. More on that to come, but don't forget to sign up!

Sign up for my free webinar

Hi there,

I asked Kameron Tanseli to walk me through an ordinary Tuesday. He runs growth engineering at Flora, and built it at Fyxer.ai while it went from $1m to $32m ARR.

Here’s what he sent back, more or less as written:

• Wake up. Double espresso from the Sage Bambino, over ice (snap, same coffee machine!). Gym, breakfast.

• Open PostHog. Check the metrics, check the running experiments. (His note next to this line: “peeking lol.”)

• Come up with two or three ideas for the backlog and add them to Linear. He’ll often point an LLM at PostHog and BigQuery to do the data science part for him.

• Hand the build to Devin, or write it in OpenCode with AI assistance, depending on complexity.

• Devin QAs every pull request end to end, with video and screenshots.

• Lunch, and a walk in nature.

• Merge to production. Launch the experiment. Post it in the #experiments Slack channel.

• Meetings from 3pm to coordinate bigger launches with other squads.

• 6pm: brain dead, go cook dinner.

Read it again and notice what isn’t there:

There’s almost no building in it.

He specs, an agent builds, an agent tests, he merges. The part I always assumed was the job, the part founders think they’re paying an engineer for, has been squeezed into an afternoon and mostly handed to something that doesn’t sleep.

So what actually fills the day

Ben Fanning, a growth engineer at Fyxer, doesn’t describe a schedule at all. He describes a cycle:

• Data first

• Then ideation

• Then design

• Then the build

• Then analysis

• Then improving the tooling that does all of the above.

“There’s a lot of days where you do all of these things in one day, which is part of the joy of growth engineering.”

Six stages, one person, one day. Which only works because five of the six are decisions and one is typing.

Three things eat the hours.

1. Deciding what’s worth testing. Kameron adds two or three ideas to the backlog a day. Almost none become experiments. He runs one, maybe two a fortnight. The gap between those two numbers is the actual job: analysing, understanding and prioritising.

2. Checking it’s real before it’s built. Kameron and I approach this similarly: running a few quick calculations before choosing experiments goes a long way toward ensuring you have enough traffic to test and that the opportunity is worth it (MDE = Minimum Detectable Effect).

   

3. Checking it’s real after it’s built. At Flora, Devin signs up as a customer for every pull request and sends screenshots of each variant. Kameron’s example: it tried the Stripe checkout, found a coupon leaking in from a different experiment, fixed it, and sent him the new screenshots. His reaction was “I don’t need QA engineers anymore.”

Nobody has ever enjoyed that job. It’s the first one I’d hand over too.

Why one person can hold all of it

Ben describes the role as the centre of a Venn diagram: part software engineering, part experimentation and data, part product and user psychology. Drop a circle and you land in a different job.

Nobody arrives holding all three. There are three routes in, he says: engineering, data, or business. And the most common one now is business, “especially now with AI making the coming from the technical side potentially less important.”

The gate is moving, not closing.

Three things to steal, without hiring anyone

Most of you reading this are five to fifteen people; you might hire one growth engineer, but you won’t have a full team.

So here is what you can learn:

1. Give one person one metric and don’t let them change it. At Flora, one person’s entire job, week after week, is the percentage of new signups who generate an image. Weekly dashboard, weekly experiments, keep going until the number moves. Not a theme. Not an area. One number.

2. Draw a box around it. Kameron’s rule for avoiding collisions is brutally simple: “I will say to an engineer, you are from sign up to paywall. That is your area. You can draw a box around it. No one else is going to really touch it.”

3. Size the test before you build it. You don’t need Fyxer’s platform. You need an honest answer to how long a result would take at your traffic. If it’s four months, you’re testing in the wrong place. Go further down the funnel, where the numbers are smaller, and the money is bigger.

None of those needs a new hire. All three need someone to decide.

What I’d recommend this week

Join us on 19 August. I’m doing a free webinar with Andrew Capland on how these growth roles are pulling apart into different jobs with the same title.

Join the webinar

The building has quietly become the cheap part. What’s left is knowing which two of your twenty ideas are worth twelve days, and being honest with yourself when the answer comes back.

Speak soon,
Daphne

A while ago I was standing in the kitchen, checking several weather apps on my phone, trying to work out whether I should walk the dog now.

I turned around to find my father-in-law looking at me like I’d lost it: “Why are you doing that? Just look out the window.”

He wasn’t wrong. We have a book's worth of these types of down-to-earth quotes by him (“What type of music do you like to listen to while driving?” “Silence.”)

But it stuck with me. Because for me, the obvious move is always:

• Open the app (or three)

• Check a website

• Ask someone.

For him, the obvious move is: look outside. Go outside.

Same job (”what’s the weather doing?”). Completely different idea of the obvious solution.

That four-word takedown is the whole game in early growth, and it’s the thing I watch founders miss again and again.

You think you’re competing with the other apps in your category. You’re not. You’re competing with whatever your user already does to get the job done. And usually, that thing is free, instant, and good enough.

Your competitor isn’t another app

Time and time again, I hear apps compare themselves to other apps. But I challenge you to look a little further. Once you see it, it’s everywhere:

A meditation app isn’t really up against other meditation apps. It’s up against going for a walk, a yoga class, or flopping on the sofa with the TV to switch off.

A recipe app is competing with old school cookbooks. Duolingo is competing against in-person lessons. And a weather app is competing against, well, the sky.

An app I worked on for the over-65s, helping them stay active, had one giant competitor: YouTube. Their users’ kids kept sending them free workout videos.

How we actually beat YouTube

Let’s talk about that active app and our huge challenge of going up against a mogul like YouTube. We couldn’t win on “short, easy videos.” YouTube already had those, for free, forever.

So we stopped competing with the category and started competing with the real alternative. We focused on the thing YouTube couldn’t offer: a coach who knew their specific struggles:

• Could they keep their balance?

• Carry the shopping home?

• Get down on the floor to play with the grandkids, and back up again?

Our best-performing ad wasn’t about workouts at all. It was “keep up with your grandkids.” (Honestly, that still gets me.)

Because that’s the job. YouTube doesn’t promise that. A pile of free videos doesn’t promise that.

Why this isn’t just a nice reframe

I’m not trying to be pedantic, and it’s not just about mindset. This is about setting yourself up for success.

If you benchmark yourself against other apps, you build a slightly better version of the category, only to wonder why nobody switches.

The window is free and already in their hand. “Slightly better than the other app” loses to “free and instant” every single time.

To win, you don’t have to beat the category. You have to be, obviously, specifically better than the one thing they actually do instead.

What to do this week

1. List what they do instead. Write down everything a user does to get this job done without you. Include “nothing,” “the manual way,” and “ask a person.” For weather: the window, stepping outside, Google, a friend, the pre-installed weather app.

2. Find their default. Of that list, which one do they reach for first? That, not the App Store top charts, is your real competitor.

3. Beat that one thing, out loud. Make your value obviously better than that specific alternative, and say so. Not “a better weather app.” More like “stop guessing whether you’ll get rained on halfway through your run.”

And steal this question for your next five user interviews: “Before you had anything like this, what did you actually do?” Their answer is your competitor. (It’s almost never another app.)

What I'd recommend this week

If this clicked, the classic version is the McDonald's milkshake study (the story that kicked off Jobs To Be Done).

The team realised their milkshake wasn't competing with other milkshakes at all. It was competing with bananas, bagels, and a boring morning commute.

Clayton Christensen's Competing Against Luck is the deeper read if you want the full thinking behind it.

I fall for this too. I assumed everyone uses a weather app.

It took my father-in-law and a window to remind me that most people just... don’t.

Sometimes the thing your product is really up against is a behavior so obvious your users don’t even call it a solution.

So before you study a competitor’s paywall and onboarding for the hundredth time, go simpler. Find the free, boring, good-enough thing your users already do.

At the moment, this is often AI; AI is your competitor. So you need to understand how they use it instead and what you can do to offer real value on top of that.

See you next week,
Daphne

June was for tidying. July, it turns out, was for building.

Not building more — we’ve already established my feelings on more.

Building better foundations:

A business that actually supports me and my ADHD instead of one I have to constantly wrestle with. And, for the first time in a while, a real plan for what I want the next few years to look like.

The nudge came from this newsletter by Andrew Capland, reflecting on five years of working for himself. It made me realise I’d been so busy running my business that I’d stopped asking where I actually wanted it to go. We’ve been having some great chats about it and are even hosting a free webinar together next month.

So that's been July's quieter project, ticking along in the background between experiments, the cohort, and a fox with some very strong opinions about my garden (more on that later).

Here’s everything that happened.

What’s been happening with clients

July has been a month of experiments.

We've been running some really interesting tests around onboarding and paid trials, with a mix of expected results and a few surprises. Some experiments gave us the lift we were hoping for, while others raised even better questions.

That's one of my favourite things about experimentation: even when the outcome isn't a clear win, you still learn something valuable. Sometimes the most useful result is finding out what doesn't move the needle.

What I've been presenting

Sub Club Live with Alice Muir and David Barnard on reactivation! We went live on July 2nd and covered everything from whether you should even prioritise reactivation in the first place to the tactics we've seen actually work.

Presenting in 30°C+ heat was... an experience (I have never looked shinier on camera), but I did my best to bring the energy.

If you missed it, you can watch the session back here.

What I’ve been writing

Nothing new went live this month; a planned pause on the editorial side meant a few pieces were pushed back.

That doesn't mean the writing stopped, though.

Two new pieces are on their way — one on money-back guarantees and one on growth stalls — plus I’ve been preparing some extra Growth Waves content with some fun breakdowns and advice I can’t wait to share.

What I’ve been up to with Women in Experimentation

The personal branding cohort kicked off on July 10th!

You might remember I nearly talked us out of running it. I’m so glad I didn’t, because this group is such a motivated bunch. Every session leaves me more energised than I started (which, for someone who carefully guards her energy, says a lot).

One of our free spot participants, Miracle, shared such a sweet post.

This is exactly why we wanted to offer free spots to people who are rebuilding.

The job market is a tricky place at the moment, so it feels good to be able to help women out there who are struggling and we have plans to do more free webinars and sessions with WiE to support them further.

This month’s flip-flops

As always, the messy reality:

I triple-checked course access. Google Drive had other plans. After retiring my self-serve courses (as promised in last month’s issue), I moved everything over to Google Drive and checked the permissions no fewer than three times.

Some wonderfully mysterious permission settings still managed to lock people out. The irony of breaking access while trying to simplify everything was not lost on me.

It turns out I didn’t have the original course videos. So retiring the courses meant redownloading 180 (!) videos from Podia to organise. Joys.

The silver lining: it pushed me to build a second brain in Claude that now automatically organises all my relevant files. Future me should have a much easier time.

My survey estimates let me down. I’m usually quite good at estimating survey response rates. This month, for one client, they’re coming in super low.

So I’m revamping and reviewing to see what we can do.

The gardening saga continues. I successfully grew mint and rocket this year (redemption after this year’s graveyard of seedlings!)... and then the foxes dug up and ate all my rocket. Why, foxes, why? The mint lives on. For now.

My true saving grace is my fig tree; I’ve proudly picked already 26 figs (yes… I’m keeping count).

What I’ve been enjoying

Silverstone! I headed to Silverstone with a former colleague from Heights (now a great friend), our partners, and a lot of excitement.

I got to see all my favourite drivers, which almost made sleeping in a tent worth it. Almost. Camping and I are still not on speaking terms.

Our annual road trip to France with friends. The tradition continues: the same core group of friends (plus a few new additions), a tiny village, a champagne festival, and a weekend full of good food, a few morning runs around Reims, and catching up.

London remembered it's summer. It was hot. Complaining about the weather is basically a British rite of passage, so I'm assimilating nicely. The long evenings, though? I'll happily keep those.

Attended a breathwork and sound session with Rosie Hoggmascall. This was actually my birthday gift to Rosie, but I think I needed it more than she did. I'm fairly certain I drifted off towards the end 😅... which I'm choosing to believe means it worked.

A day-long gardening course. Yes, before you point it out: I did a gardening course and still lost my rocket to foxes. The course was after that and is my comeback story 🦊.

The focus was on designing a low-maintenance, food-growing garden, and we even sketched out our own plans. I'm feeling cautiously optimistic.

Completed the Ben AI accelerator course. This is where the second brain came from. Investing in actually learning this stuff properly, instead of duct-taping tools together, has already been paying off.

I became an auntie! After a difficult pregnancy, my sister welcomed a beautiful baby boy, Lucas, on 21 July. He’s still a little one, but slowly growing :).

If you missed my newsletters this month

Looking ahead to August

• A new growth audit — always exciting to dig into a fresh app with fresh eyes.

• A major feature launch for a client, very excited for them.

• A webinar! About the three directions growth roles are going towards; you can sign up for free here.

• More building and improving, of course.

AND… my birthday. Turning 33 on the 13th August and decided to stay young by going climbing in trees 🌲🌳.

If June taught me that endings make space, July taught me what the space is for.

Not filling it back up, but building something that fits.

Till next time,
Daphne

I’ve been dancing around this for months.

If you’ve been reading Growth Waves for a while, you may have noticed a shift:

• The teardowns? All apps.

• The client examples? All apps.

• The problems I get excited about — paywalls, onboarding, day-7 retention, activation loops? All apps.

I just never officially said it.

So: here it is. I’m apps only now.

How I got here

My career path has never been a straight line. (Whose is?)

It went:

1. Studying business and then Finance

2. Realizing I definitely didn’t want to go into the world of banking

3. Joining a growth agency to work across every type of brand imaginable (from B2B to apps to e-commerce)

4. Head of Growth at an e-commerce subscription startup

5. Then back out as a consultant and advisor, working across both worlds of apps and e-commerce

For a while, that felt like a strength. I could speak both languages. And because I was only working with B2C brands, the audience challenges often felt similar, whether it was an app or an e-commerce business.

But over the last year, I started noticing something… When a lead came in, was it an app? I’d get excited. I’d download it. Test it. Pull it apart. Think about the onboarding, the paywall, the activation moment, and the retention loops. I’d want to understand every part of it.

I didn’t feel the same way about e-commerce anymore.

I tried to ignore that for a while. E-commerce has been good to me. I’ve spent almost 10 years building experience in that world.

Letting that go felt a bit like… a breakup.

Not because anything was wrong. Not because I didn’t appreciate it. Just because sometimes you grow in a different direction, and you have to be honest about what you’re excited to build next.

What e-commerce gave me that I’m not leaving behind

Here’s what I’ve realized, though: I’m not actually leaving e-commerce behind. I’m taking the best parts of it with me.

Years of DTC work taught me things that are genuinely invaluable for apps:

1. Thinking in full customer journeys, not isolated screens.

   In e-commerce, the pre-purchase experience matters just as much as what happens after someone buys.

   You’re constantly thinking about the entire journey — from first touchpoint to conversion to repeat purchase.

   That framing completely changed how I approach app onboarding.

2. Web funnels and landing page psychology.

   A lot of app teams discover web funnels later. I’ve already spent years obsessing over them.

   Understanding messaging, friction, intent, and what makes someone take the next step is just as relevant for an app as it is for a checkout flow.

3. Retention when you can’t measure everything perfectly.

   E-commerce attribution is notoriously messy. You don’t always know exactly why someone bought, what influenced them, or which touchpoint mattered most.

   You learn to make smart decisions with imperfect data. That skill transfers.

4. Marketing can’t solve product problems.

   I learned this one the hard way. (And it became, roughly, my entire philosophy on growth, as you may have noticed.)

   The biggest growth opportunities usually aren’t hiding in another campaign. They’re in the product experience itself.

5. Both are in the wellness space.

   I focused on wellness and eco-friendly e-commerce brands; my plan and love is still for wellness apps.

That crossover background is genuinely what makes me different in the app world. I think about real customers, not just funnels.

I’ve seen what works across both models. And I think that combination is pretty powerful.

Why apps won me over

There were a few things that kept pulling me toward apps, and I’d be underselling it if I didn’t name them.

1. I can actually impact the product!!!

   In e-commerce, growth consultants often get pushed into the marketing lane — ads, email, acquisition — because working on the actual product is a completely different skill set.

   With apps, I’m often in the room (or the Notion doc) when product decisions are being made. That’s where I do my best work.

2. The data is better.

   Not perfect, but better. Behavioural data, cohort analysis, event tracking — you can actually see what users are doing inside the product and use that to make decisions. I get so much joy of running in-depth analyses using tools like PostHog and Amplitude.

   In e-commerce, so much of the customer journey happens in the dark.

3. Growth compounds.

   In e-commerce, you’re often chasing spikes. A great campaign. A seasonal moment. A viral hit. Growth loops exist, but they’re harder to build.

   With apps, if you fix retention at the right stage, the impact compounds.

   Improving the product experience can change the entire model. That compounding dynamic energises me in a way I didn’t expect.

4. No stock issues.

   Our best growth month when I was Head of Growth was followed by a stock outage and having to turn everything off.

   I cannot tell you how much mental energy I spent on this across previous clients dealing with running out of stock... Never again.

What this means in practice

The e-commerce brand I still work with is one where I can focus on exactly the parts I’m best at: subscription growth, churn prevention, user research and the full customer journey.

Those are the pieces of e-commerce that map directly to my app work. I’m only continuing with brands I’ve already worked with — relationships where I know the business, understand the customer, and can have the biggest impact.

Going forward, when I am ready to take on a new client (I’m currently only available for one-off strategy sessions or growth audits), I’ll be focusing that space on consumer apps, specifically wellness apps.

Because if you’re a consumer app founder, this is exactly where I want to be spending my time.

Activation. Retention. Monetisation. Growth experiments. Growth hiring. This is the work that gets me out of bed.

And if you’ve been around here for a while…

Thank you. Genuinely.

Growth Waves started as a newsletter for both worlds; I called it D2C and tried to speak to e-commerce and app brands. But I definitely started off skewing to e-commerce.

I’m not deleting the e-commerce archives. There’s still a lot of value there. But going forward, Growth Waves is an app-focused newsletter.

If that’s not what you originally signed up for, that’s completely fair.

And if it is? I think we’re going to have a lot of fun.

Till next time,

Daphne

I was just 1km into a 40km hike when the storm hit.

My boots were already soaked, squelching with every step. I was halfway up the first peak, with another 39km and the better part of 10 hours still ahead of me. (Yes, I’d checked the weather. Saturday “looked fine.” Saturday lied.)

So I had three options:

• Pivot: Turn back and try again on Sunday.

• Adjust: Slow down, find shelter, change the plan.

• Or Stay the Course and keep climbing.

I kept climbing and finished the hike.

Recently, I was thinking about that moment on the trail because I’d just sat down with a founder facing a different version of the same decision.

Most of us reach points where we have to choose a direction. Instead, we convince ourselves we can climb three mountains at once.

During our strategy call, he explained he was building in one of the most crowded categories imaginable. (I’ll keep him anonymous, but think weather apps. Almost everyone has one on their phone.)

Which was exactly why he felt stuck.

He had at least five different audiences he could build for. And when almost everyone uses your type of product, the possibilities feel endless.

So he was trying to serve all of them at once.

That’s the trap.

Five audiences means five slightly different products, each pulling in a different direction. Your roadmap slowly fills with competing priorities. You build a little for everyone and something indispensable for no one.

So I gave him the same thing I reach for whenever I’m overwhelmed by too many options: a way to make the decision smaller.

Here’s the thing, almost no one says out loud: the fear of focus is really the fear of permanence.

We don’t avoid choosing an audience because we can’t decide. We avoid it because choosing feels final. “If I commit to runners, I’m abandoning everyone else forever.” So we stay broad and water the product down for all of them.

But committing to an audience isn’t a tattoo. It’s a direction of travel. And there are only ever three things you can do with a direction.

The 3-way check: stay, refine, or pivot

You pick an audience, then run this every few weeks. Look at the evidence, then choose one:

1. Stay the course. The signals are good. Your target users are activating, the feedback’s right, there’s payment intent. Don’t second-guess it. Put your energy into execution.

2. Refine. The direction’s right, the details are off. Maybe it’s not athletes training for an event, just people who want to move more. Maybe it’s one use case, not five. Same mountain, better route.

3. Pivot. The evidence killed your core bet. Your target users tried it and didn’t see the value. Change the problem, the user, or the solution. On purpose, not in a panic.

That’s it. Three options. The goal isn’t to be right on day one. It’s about committing hard enough to actually learn, then steering.

It’s how the apps you admire got there:

• Calm could have been “wellness for everyone.” They committed, saw Sleep Stories take off, and doubled down.

• PhotoRoom could have been “Photoshop, but easier.” They picked one job (background removal) for one audience (non-designers selling online who need speed, not perfection), then expanded.

• Ladder (I use it most days) had real product-market fit with fans of specific influencer coaches. They hit roughly $1M ARR and a ceiling. To grow, they pivoted to a broader audience who didn’t want to pay extra for 1-on-1 access.

None of them got it right up front. They committed, then adjusted.

What to actually do this week

If you’ve got multiple audiences you’re juggling and you’re scared to pick, do this:

1. Score them.

For each audience, rate three things out of 5:

1. How painful the problem is

2. How badly it’s being solved today (aka are other apps not meeting their needs)

3. How willing they are to pay.

Use that to prioritise the highest potential audience.

2. Write it as a hypothesis.

“We’re building for [audience] who struggle with [problem]. We’ll know we’re right if [specific signal].”

A hypothesis is easy to change and can be proven/disproven.

3. Decide your evidence in advance.

Before you ship anything, write down what would make you stay, refine, or pivot. Otherwise you’ll just read the data to confirm what you hoped. We all do it, let’s not pretend otherwise.

I’m a pivot’er

Quick confession, so you don’t think I’m preaching from some serene strategic mountaintop: I’ve repositioned my own consultancy more times than I’d like to admit.

I’ve dropped services, added new ones, niched down, then niched differently.

You know what happened? Nothing.

No one cared. (Sorry, but it’s oddly freeing.)

The hard part isn’t choosing the perfect direction. It’s choosing one, knowing you can always change it later.

What I’d recommend this week

If this hit a nerve, I get properly practical on this in my free course with RevenueCat: Early product decisions: How to build an app people will pay for.

It walks through how to choose your first audience, define what “working” actually looks like, and run the stay/refine/pivot check without spiralling into endless second-guessing.

It’s the exact framework I kept reaching for during that strategy session.

Check out the course

I’d also highly recommend watching the Pivot scene from Friends again. Pivoting is far less scary when you have the phrase “Pivot, Pivot, Pivaat!” in your head.

One last thing I told that founder, because it’s the real risk when you’re trying to do too much: don’t fall for the solution before you’ve nailed the problem.

He’d sketched out a genuinely clever feature, the kind that makes you go “Oh, that’s actually different,” and I loved it. But a clever feature for an audience you haven’t committed to is just a nicer way of going off-topic.

Problem first. Then solution.

So: pick the audience with the strongest pull. Commit until the evidence tells you otherwise. Then decide whether to stay, refine, or pivot.

You’re not signing your life away. You’re simply choosing which path to take.

And when the storm rolls in, you can always change direction. (Ideally in drier boots than mine.)

See you next week,

Daphne

I’m a growth advisor. And I’ve decided to stop growing.

Not the advice you’d expect from someone whose entire job is helping companies grow. But here we are.

Today, instead of telling you how to expand your brand, I’m going to tell you why I’m deliberately putting a ceiling on my own. Bear with me.

When “more” started to hurt

I fell into freelancing accidentally and stayed for years. Every single year, I grew. I never wanted to become an agency, so that capped things nicely, but I still kept doing more. More courses, more talks, more projects.

Until more started hurting me.

• My newer courses cannibalised the sales of my old ones.

• My website grew bigger but felt constantly out of date.

• I had a list of projects and ideas I never had time to touch.

• I made mistakes in the rush to ship — wrong links, a feedback form none of you could actually access (oops).

Better beats more

Then I read Company of One. It describes how a business can become a beast: you grow and grow until you lose control of it. The answer isn’t more. It’s better.

More is tempting, especially with AI. But more doesn’t always mean better.

I want to be better. My revenue has grown, and I don’t need it to grow beyond this. I don’t want to hire a team I have to manage, or do everything at half quality.

I know not everyone gets to make this choice right now. Plenty of people are having “less” forced on them, not choosing it — and I don’t take for granted that I get to choose.

The point of a Company of One isn’t growth at all costs (ironic, coming from a growth advisor). It’s growing to the level that feels right for you, and shaping your life around it, instead of the other way round.

Why now?

Because I love what I do. Truly. But I also love long walks with my dog, working out, visiting new places, and time with my family and friends.

The last two years made that non-negotiable. I lost my stepfather. My husband had cancer (he’s been all clear now for 1.5 years). My older sister has been in hospital for 10 weeks with pregnancy complications.

If I’d grown faster, I couldn’t have been there for them. Even now at times I’m feeling the strain of it all.

What I’m stripping back

So I’m doing something that sounds strange for a growth person. I’m stripping things back.

Over the coming months, I’m going to be:

1. Removing my self-serve courses.

Don’t panic. Hear me out.

So much has changed since AI, and since then I’ve launched my CXL and RevenueCat courses to high demand. So it just doesn’t justify keeping my self-serve courses.

I’ll keep my CXL and RevenueCat courses, and I’m working on shifting the rest to Google Drive Access. If you bought one: don’t worry, you’ll keep your access, and I’ll share some additional gifts as a thank you.

2. Simplifying my website and how I work with brands.

Less noise, more up-to-date content on my actual approach and who I am. Removing the courses will help with this, and then it’s time for a full refresh.

3. Testing simpler LinkedIn content.

My carousels have been precious to me; they’re my mini stories and guides, and the way I love to consume content as someone with ADHD. But LinkedIn’s new formatting has made them almost impossible to read. So I’m testing simpler text-and-image posts: still scannable, but a better fit for what actually works there now. As well as carousels with larger fonts.

4. Removing inactive subscribers.

My newsletter grows over time, but the open rate has slipped over time. It’s hard to let go of list size, but a smaller, engaged list beats a big, quiet one. I’ve just tidied up inactive subscribers. This will also make it easier for me to see which content you like and which you dislike.

5. Reducing my client load.

I love my current clients, and I’m not going anywhere. I keep my client list small on purpose, so I can do right by the ones I have. At some point one of them will outgrow me — they’ll need a full-time hire or a different kind of expert for their next phase. When that day came last month, I chose not to take on a new client.

The missing piece

There’s one puzzle piece I haven’t solved. I want to help smaller startups too — the earlier-stage founders who don’t fit ongoing client work, and who I rarely have room for anyway.

To be clear: this isn’t me adding more on top. It’s redirecting. I’ve tested one-off strategy sessions, courses, workshops and growth audits. Strategy sessions and audits worked best, but they don’t scale.

So I don’t have the answer yet, and that’s fine. It will come later.

If you’re feeling the pressure to do more right now — more output, more channels, more everything, because AI makes “more” so easy — take this as permission to do less, better, too.

First, I want to focus on you. This newsletter. Showing up better here, adding more value, more consistency. I’ve already built a Claude dashboard to keep a better eye on performance.

Worth a read (this week’s recommendations)

The book that kicked all of this off is Company of One by Paul Jarvis. If any of this hit a nerve, start there — it’s the clearest case I’ve found for staying small on purpose.

Two more, if you want to sit with the idea:

• Four Thousand Weeks by Oliver Burkeman — on making peace with the fact that you can’t do it all. The one that helped me most this year.

• Essentialism by Greg McKeown — the disciplined pursuit of less, but better.

So here’s my question for you: if you let yourself, what would you strip back?

Hit reply and tell me; I read every answer.

Thank you for being part of my Company of One.

Company of One doesn’t mean doing it alone with no help — I have a lot of help.

It means a company that resists traditional growth: not chasing more, not hiring to add capacity, but getting better.

My younger sister has a tattoo on her arm: This is me trying.

Well, this is me trying to be better,

Daphne

Endings scare me. More than I’d like to admit. I slow down my reading to avoid finishing a beloved book; I avoid watching a TV show’s finale, and I hate goodbyes with long-distance friends.

So when one of my longest-standing clients rounded off our work this month, earlier than I’d initially planned for, my first reaction wasn’t pride.

It was panic. (Not my finest growth-consultant moment, I know.)

It took me a few days to see it clearly. The work hadn’t gone wrong; it was time for them to hire someone full-time to take over. I’d just thought I’d have a few more weeks to round off.

It was tempting to immediately look for a replacement, but, as if by serendipity, I was fortunate enough to be asked about a few potential projects.

But I stopped myself from taking them on.

I’d just been carrying too much, and I needed time to tidy up. To reorganize. To work out what I actually want more of, and what I’m finally ready to put down.

So that became the real theme of June: tidying up.

I know it’s nearly July and technically far too late (and hot) for spring cleaning. But clearing the clutter in my business, my life, and my head has felt so good.

So much has changed lately, in the world and in my own life, and I’m only human. I needed the room to process. This month, I finally took it.

Here’s everything that happened in June.

What I’ve been writing

Two articles went live this month:

The First 7 Days Are The Most Underrated CRO Opportunity (Convert)
Takeaway: The first week is where retention is quietly won or lost. Get those seven days right and everything downstream gets easier.

Monthly Plans Might Be Your Best Option (RevenueCat)
Takeaway: Annual isn’t always the smart default. Sometimes, monthly is the better call, and I make the case for when.

What’s been happening with clients

The client I mentioned at the top? Here’s the part that makes the ending easier to swallow.

Over the year or so we worked together, their sales grew 167%. Subscription revenue? Up 210%.

They’ve now grown to the point where they’re ready to hand the reins to someone else for a while.

And when I finally stop panicking long enough to actually look at it, that’s exactly what you want to happen. A client outgrowing you in the best way is a strange, proud, slightly sad thing all at once. The baby brand is flying out of the nest.

The lesson I keep relearning: an ending isn’t a failure. Sometimes it’s just space.

And this particular ending makes space that I feel like I’ve needed.

Such as time to build more automations to take the manual, repetitive tasks off my plate. Less admin, more thinking.

Clearing the clutter so there’s room for the work that actually needs me.

What I’ve been up to with Women in Experimentation

I’ve been deep in prep for a personal branding cohort, and I genuinely can’t wait for it. Every sign-up that comes through makes me giddy with excitement.

One reason it matters so much to me is that we’re giving spots to people who’ve lost their jobs or are going through a tough time.

Personal branding shouldn’t be a luxury for when everything’s already going well. It matters most when you’re rebuilding. That one sits very close to my heart.

If that’s you, or you know someone who could benefit, here’s the form for the free spots.

And if you’d like to join the cohort in general, you can apply here.

I’ve also been working with the judges to select the speakers for the WiE Summit. Reading through the submissions and getting to shape who gets the stage has been a huge challenge (120+ applicants) but also a quiet highlights of my month.

A little recognition

I won Best Educational Content for Community Content with Guest Test at the Best in Test Awards.

I’ve poured so much time into writing and content over the years, often wondering if any of it actually lands, so seeing that recognised meant more than I expected. I also got a badge, which is pretty cool.

This month’s flip flops

The panic when the client ended. I’ve already told you this one. My first instinct when something rounds off is to spiral rather than celebrate. I’m working on catching it faster. Progress, not perfection.

I tell founders to do less, then build myself a to-do list a mile long. I spent part of this month mapping out all the AI tools and automations I want to build, and ended up with a backlog so big I’d never actually start any of it.

I built some cool new things with AI this month.

Curious? Let me know, and I’ll include a “What I built” section in the next recap.

The irony of the “you need fewer, better things” person doing exactly the opposite was not lost on me. So I went back and made every idea smaller. The smaller the task, the more likely I’ll actually do it. (Funny how the advice I give is always easier to give than to take.) More on this next week.

I nearly stopped us from doing the Personal Branding Cohort. I let my own doubts fill my head: what if we don’t have PMF? What if no one signs up? What if my challenges on LinkedIn recently make me the wrong person for this?

I’m so glad Lucia and Marcella encouraged us to go ahead, because seeing the stories of the people who signed up for both paid and free spots has been such a motivator.

Just because I can do it doesn’t mean I should. I got asked a lot this month for meetings outside my usual time, process changes, and extra tasks. In the last few months,

I’ve been a bit too quick to say yes, but it’s come at a cost. This still happened a few times in June, but at least this month I was aware of it and managed to say no much more often than before.

What I’ve been enjoying

A month of less travel, which has been so nice. The one trip was at the very start of June, to see my sister who’s been in hospital in Bordeaux. Everything since has been lovely and local, and I’ve been soaking up the London weather while it lasts.

The highlight: I got to see Lily Allen with my younger sister. She’s a journalist and somehow managed to get us tickets, which was just incredible.

I’ve been a bit obsessed with her latest album, mostly because of how raw and emotional it is. The whole show was one beautiful theatre performance from start to finish.

If you missed my newsletters this month

Looking ahead to July

First up, a live SubClub episode tonight together with Alice Muir on Reactivation Strategies.

Then there are a few more articles on their way, focusing on fixing your first renewal and reducing churn due to ‘low usage’. First renewals and low usage are two of the most under-watched moments in the subscription lifecycle, and I’ve got a lot to say about both.

And of course, the live personal branding cohort is in July too.

Otherwise?

More tidying.
More AI automations (I’ve been nerding out building a second brain).

If June taught me anything, it’s that endings aren’t the enemy. Sometimes they’re just space, to tidy up, to reorganise, to choose what you want more of.

I’m only human (a fact I’m relearning constantly), and I needed the room to process. I’m taking it.

Till next time,
Daphne

Yes, 80% of trials start on Day 0. Yes, the onboarding paywall matters. And yes, you should absolutely optimise it.

But here’s something I’ve noticed working with apps: teams pour all their energy into the onboarding paywall, while barely thinking about what happens after.

For freemium apps especially, that’s a missed opportunity.

The onboarding paywall captures users at their moment of highest curiosity.

But feature-based and contextual paywalls catch users at their moment of highest need, and that’s often a far more powerful point to convert.

Vahe Baghdasaryan talks about these three paywall types in his Paywall Optimization Masterclass, and it struck me how under-discussed the latter two are.

So here’s to the underappreciated paywalls, and how to approach them properly.

Why the moment of need beats the moment of download

When someone downloads your app, they’re curious and maybe motivated, but they haven’t experienced it yet. They don’t know what they’re missing.

Now compare that to the moment when a user actually tries to do something and discovers it’s premium.

At that point, they’ve already decided they want it. The intent is real, not hypothetical. Interestingly, these later subscribers often have a higher LTV than those converting on Day 0.

There are two types of non-onboarding paywalls worth understanding.

Feature vs Contextual paywalls

Vahe splits non-onboarding paywalls into two types:

1. Feature-based paywalls

These trigger when a user tries to access a specific premium feature. The key is that they hit the paywall because they want that feature, not because the app decided to interrupt them with a sales pitch.

Here’s how X does it when you try to access offline videos:

It makes it clear this is premium and pushes you to upgrade, taking advantage of the exact moment you wanted that feature.

CapCut, the video editing software, does this cleanly as well. Basic editing is free, but fancier subtitles or specific effects are clearly marked as premium. You hit the paywall because you tried to use something, not because the app decided it was time to sell.

The principle: make the paywall about what the user wants, not about what you’re trying to sell.

2. Contextual paywalls

These paywalls appear based on behaviour, timing, or milestones, not just feature access.

Fitbod, a workout app, shows a paywall after three free workouts. By then, you’ve started building a habit and seen the value. The paywall hits at a moment of demonstrated commitment, not speculative interest.

Duolingo’s hearts system creates natural paywall moments when users run out and want to keep practising. The timing isn’t arbitrary; it’s when motivation is highest.

The principle: surface the upgrade when intent peaks, not when it’s convenient for you.

Contextual paywalls are about being aware of the moment, the user’s behaviour, and where they are in the journey, and presenting the offer when it’s most meaningful.

Combine the two for impact

For me, the two types of paywalls blur a bit, and it’s the combination that’s really powerful.

I remember signing up for AllTrails, the hiking app, when I lost connection to my offline map. I got a brief moment of signal in the middle of nowhere in Crete (minutes before I ended up being chased by wild dogs, but that’s a story for another time).

I tried to download the map, saw it was premium, and thought, “F** it.” I upgraded immediately — no way I was getting lost alone in Crete.

Here’s how their paywall works when you try to download a map:

Versus when you try to see the community heatmap (where others have walked):

Different visuals and plans were pushed; it’s adjusted to the feature and the moment.

Combining context and a feature is a powerful converter.

How to think about paywall placement

A few questions worth asking when thinking about paywalls:

• Where do users hit a wall? Look at where free users try to do something and can’t. Those are natural paywall moments. If you’re not surfacing an upgrade offer there, you’re missing high-intent conversions.

• How many times does someone see your paywall? Some users need multiple exposures before converting. One interaction isn’t always enough. Check the percentage of free users who see a paywall per session.

• Does your paywall match the moment? A generic “Upgrade to Premium” doesn’t work as well as “Unlock this route” or “Get unlimited hearts.” The more specific it is to what users are trying to do, the better the conversion.

• Is it clear what is premium? This sounds obvious, but it comes up all the time: upgrade options buried in settings or hidden behind tiny icons. If users don’t know premium exists, they can’t buy it. Don’t wait for them to click through; make it obvious.

This isn’t about more paywalls, it’s about better

For smaller apps, yes, visibility often needs improvement. But for most, the bigger opportunity is matching paywalls to the right moments. Tools like RevenueCat Paywalls and Superwall.

When you align paywalls with features and context, users don’t feel sold to; they feel like you’re meeting their specific need.

Recommendation

In every edition of Growth Waves, I share a resource related to the week’s topic.

Jacob Rushfinn wrote a fantastic deep dive on dynamic and contextual paywalls, with concrete examples from Strava, Calm, and Duolingo. He also highlights how AllTrails handles this well and even includes a whimsical board of contextual paywalls.

If you’re looking for inspiration on making paywalls feel more relevant to user behaviour, it’s definitely worth a read.

Read more

I would also definitely recommend Vahe’s course.

The onboarding paywall gets attention because it's measurable, and it's where most trials start.

But for freemium apps, the paywalls that appear at moments of real need often convert the users who weren't ready on Day 0 — but are ready now.

Daphne

A friend of mine was looking for a new senior role in data and Business Intelligence. He’s the kind of brilliant, experienced operator you’d trust to run your entire analytics function without blinking.

So, naturally, I did what any good friend does: went full detective mode.

Scrolling job boards, stalking company careers pages, asking around to see what’s actually out there.

And that’s when I kept coming across a role that made me think: what the…

Growth Engineer

I stared at it for a while. Closed the tab. Opened it again. Read the job spec twice.

A few days later, Ioana from AI Goodies posted something about design engineers, and it clicked. This is not a one-off; this is a shift.

So I went down the rabbit hole to figure out what’s actually going on here, and more importantly, whether this is:

a) a real evolution in how companies build and grow
or
b) another shiny “fix everything” title, like growth hacker (speaking as a former one… I was young, okay?)

So what even is a Growth Engineer?

In theory, a growth engineer is exactly what it sounds like: an engineer focused on growth. Sure, that tracks.

But theory and job specs are very different things, as anyone who’s ever hired a Head of Growth will know. I looked at five real Growth Engineer job specs across companies in the UK, France, and the US, and mapped out the responsibilities the same way I did for Head of Growth roles back in 2022. Here’s what I found:

The core appears consistently across almost every role:

• A/B testing and experimentation

• Full-stack feature development

• Analytics instrumentation

• Cross-functional work with product and design

That’s the job. That’s the base.

But then things start to diverge.

Companies 1 and 4 are building product-led growth machines: engineers who ship customer-facing features, run experiments, and own the conversion funnel end-to-end.

Company 2 sits closer to a data-heavy growth analyst with strong engineering chops; they want someone owning CAC, activation rates, and retention cohorts alongside the build.

And then there’s Company 3. Their Growth Engineer, GTM AI role is almost a different job entirely. It’s about building AI agents and automated workflows that power the sales and marketing motion, closer to RevOps engineering than product growth.

No A/B testing in sight. The only thing it shares with the others is the word “growth.”

Which brings me to what worries me.

My concern: the Head of Growth problem, but for engineers

Cast your mind back. Remember when every startup needed a Head of Growth? Or a Growth Hacker? And they were expected to be a data analyst, performance marketer, CRO specialist, product manager, and brand strategist all at once?

I wrote about why that goes wrong in 2022 — still relevant.

The mistake wasn’t hiring for growth. It was conflating “growth is important” with “one person can own all of it.”

I’m starting to see the same pattern here. Some of these Growth Engineer job specs are asking for someone who can:

• Build and ship full-stack product features

• Design and run rigorous A/B tests

• Write SQL and own analytics dashboards

• Build AI automation workflows

• Partner with marketing on GTM strategy

• Mentor junior engineers

Companies 2 and 5, especially.

That is four different jobs. And much like the Head of Growth situation, when you ask one person to do all of it, you usually end up with a jack of all trades and a master of none.

When the role genuinely makes sense

That said, I don’t think this is a bad trend. I think it’s real, and I think it’s the result of two things converging at once.

Ioana made a point in her article about design engineers that applies directly here:

The role isn’t a product of AI, but AI has dramatically accelerated both its visibility and its accessibility — by collapsing the distance between what different disciplines can each do.

A growth person who used to depend entirely on engineering resources to ship an experiment can now build more themselves.

An engineer who used to think purely about code quality can now engage more meaningfully with growth questions.

The gap is narrowing on both sides.

Being a generalist used to be a bad thing; it’s now becoming a strength. AI helps with deep knowledge/perfecting.

Now, understanding how it fits together and challenging the bigger picture is the strength, not specialist knowledge, and asking the right questions to make sure you approach it right

I’ve heard of brands already running fully autonomous growth experiments, an AI agent that handles research, analysis, and test setup with no dev support needed at all. That’s not science fiction anymore.

Someone needs to build and oversee those systems. That someone is increasingly being referred to as a Growth Engineer.

So yes, the role makes sense. But only if it’s scoped correctly.

The three flavours, and why you need to pick one

If you’re hiring a Growth Engineer, or thinking about it, the most important question to answer first is: which type do you actually need?

Even the best generalist will be slightly stronger in one area or another.

Flavour 1: The product growth engineer (Company 1 & 4)

Primarily a frontend or full-stack engineer who lives and breathes experimentation. Strong on shipping, strong on the product funnel.

Their currency is experiment velocity and conversion lift. Hire this person if your growth is primarily product-led and you need someone to build and test features fast.

Flavour 2: The data-led growth engineer (Company 2)

Analytically heavier, owns the metrics, comfortable in SQL and dashboards as much as in code.

Hire this person if you’re at a stage where you need someone to connect the dots on data to build smarter experiments, not just run them.

Flavour 3: The GTM/AI engineer (Company 3)

Much closer to a technical RevOps or GTM systems role (Should they even be called a growth engineer?).

They’re building the infrastructure that powers sales and marketing, not the product experience. Hire this person if your growth bottleneck is in go-to-market execution, not the product funnel.

Conflating these three will get you into trouble. Just like a Head of Growth who’s brilliant at paid acquisition is not the same as one who’s brilliant at product-led loops — the title sounds the same, but the skills are completely different.

The other thing to nail down before you hire: where does the role sit? Growth Engineers who report into engineering tend to be held to engineering standards.

Those who report into growth or marketing tend to skew more strategically but can struggle to get prioritisation.

Neither is wrong, but it’s worth being intentional about it.

My take

Growth Engineer is a real role.

But like a lot of new roles that emerge in the startup ecosystem, it’s currently a container that different companies are filling with very different things.

The underlying insight is correct: as AI speeds up building, the best growth people will increasingly be those who can build, not just strategise. I’m building prototypes and landing pages these days without dev and automating more and more.

And the best engineers focused on growth will increasingly need to think like growth people, not just ship features.

But if you hire a Growth Engineer hoping they’ll be your full-stack developer, your data analyst, your experimentation lead, your AI automation builder, and your GTM strategist?

You’re going to end up with the same frustrated, stretched-too-thin hire that the growth hacker era gave us.

First, define the flavour. Then hire.

Recommendation

In every edition of Growth Waves, I share a resource related to the week’s topic.

Two recommendations this week, since they sit nicely together:

Ioana’s deep dive on design engineers — a different role, but the same underlying shift. If you found this interesting, her piece adds useful context on where engineering and creative disciplines are converging more broadly.

PostHog’s take on what a growth engineer is — interestingly, their take also includes quite a bit of technical SEO. They have some great points about the characteristics a good growth engineer should have.

If you’re considering a Growth Engineer hire, write down the five things you’d most want this person spending their time on.

Then look at the chart above and see which flavour that maps to. If it maps to all three, you’ve found your scoping problem before it becomes a hiring mistake.

Till next week,
Daphne

This isn’t a normal email from me. It’s an extra one this week.

No frameworks, no teardowns today. Just a personal moment I wanted to share, and something I’ve been building quietly that’s finally ready.

First: I won an award last week

The Guess the Test Community Educational Content Award 🎉.

I awkwardly stuttered my way through a short acceptance speech (there were so many incredible names there, and I genuinely didn’t think I’d win). And it took a moment to really take it in.

Maybe once the physical award arrives it’ll feel more real 😅.

Here’s why it matters to me:

I have spent years creating content about growth and experimentation. Most of it free.

Not because I had to, but because I genuinely love teaching.

Even at university, I’d take on extra classes as a side job just because helping people learn something gave me a feeling I couldn’t replicate anywhere else.

Winning that felt like recognition of something I care about deeply. And it gave me the push to finally announce this properly.

Second: I’m running a personal branding cohort. And I want you to know about it.

Alongside Lucia van den Brink and Marcella Sullivan through Women in Experimentation, I’ve been building something for women in growth, product, marketing, data, and experimentation.

One thing I’ve always taught — informally, through WiE, through the UPSPEAK challenge with Convert — is how to build a personal brand and post on LinkedIn.

Because it’s brought me enormous value, and I know how hard it is to dare to speak up, especially in fields where women are already fighting to be heard.

So we’re scaling it up. Properly.

The Personal Branding Cohort: July 10–24

What you get:

• A positioning statement that makes you instantly recognisable

• A content strategy built on your actual expertise

• A fully optimised LinkedIn profile

• Your first post written, workshopped, and published before it ends

• 2 1.5 hour live sessions + a full video course + a workbook to keep

Price: £258

Enroll here

One more thing, and this is the part I’m most proud of…

Not everyone can invest right now. But that doesn’t mean they need this less.

For every ticket sold, we’re giving a free seat to someone who’s out of work or going through a tough time financially.

If that’s you, please apply. No questions asked.

Apply for a free seat

And if you know someone who’d benefit, paid or free, please share this with them.

It’s really appreciated. See you again on Thursday, where we’ll dive into a completely different topic: what is a growth engineer, and do you need one?

Daphne

One of the most impactful experiments we ran at Heights had nothing to do with optimizing the conversion rate.

It was a single line of copy on the product page stating that the supplement should be taken for at least 3 months for optimal results. No big deal, right?

Well, it’s still on their product page five years later.

That experiment changed how I think about CRO. We spend so much time optimizing the front door (the landing page, the checkout flow, the paywall) and almost no time thinking about what happens once people walk through it.

But here’s the thing: a 10% improvement in your conversion rate gives you 10% more customers.

A 10% improvement in 7-day retention means those customers stay longer, spend more, and refer others. The impact compounds over months.

I’ve written the full breakdown for Convert.

What you’ll learn about in this article

Your landing page already impacts activation, even before the purchase. What you communicate pre-purchase (expectations, timelines, outcomes) shapes how customers behave in those first crucial days.

Beyond the landing page, your post-purchase emails are probably focused on the wrong thing.

Most companies send a confirmation and jump straight to upselling. Those first emails should be about activation, not revenue.

I also share a practical approach to improving your first 7 days:

• How to figure out what “activated” actually means for your product (it’s usually not what you think)

• Why you should audit your post-purchase experience as if you were a brand-new customer

• Where the biggest drops happen and what to experiment on first

• The metrics that actually tell you if your first 7 days are working

Plus, there are some great examples from Mission Teas and La Doux showing how even physical products can reinforce activation through packaging.

Read the full article

Recommendation

Convert sponsors this week’s newsletter.

If you’re running experiments on your website, they’re a great testing platform with support that actually responds (their median first response time is seven minutes, which is four times faster than the industry standard; talk about optimizing for the first 7 days).

If activation is on your mind, my previous Convert article on using Jobs to Be Done to drive growth through your e-commerce funnel pairs well with this one. Understanding your customer’s JTBD helps you figure out what to say before they buy and what to reinforce after.

Most CRO teams are optimizing the front door. The real opportunity is the first 7 days after someone walks through it.

Till next time,
Daphne

I have freelancer guilt. And I’ve been hiding it from everyone — including you.

I didn’t mention the Ecuador trip in my last newsletter. Working for myself, I’ve found it hard to fully own taking time off — especially two trips so close together — even though I know they ended up nearer in timing because of life circumstances, not poor planning.

If I took five weeks of leave across March and May in an in-house role, it probably wouldn’t raise too many eyebrows (assuming that’s all my holiday for the year). But working for myself, I felt like I had to keep quiet about it.

I always promise to be real in these recaps, and hiding my holiday isn’t that. So here’s a photo from an unforgettable trip.

YES. I SWAM WITH PENGUINS AND SEA LIONS. It was incredible. Also snorkelled with sharks and did some beautiful hikes. I’m in love with the Galapagos.

I spent two weeks resetting, enjoying nature (I repeat, there were penguins!?), and on a more serious note, reflecting on my company and where I want to go with it.

And around that trip: articles were written, AI tools were built and more. So maybe I don’t need to feel all too guilty after all.

What I’ve been writing

As you can imagine, I’ve been writing very little content except for Instagram captions. Meanwhile, four articles went live this month:

In-house vs Freelance vs Agency for CRO (Convert). Takeaway: There’s no one-size-fits-all. It’s about finding the right resources for your stage and challenge.

Post-purchase screens for subscription apps (RevenueCat). Takeaway: Some apps completely underutilise these screens. Here are 13 examples that show the range of approaches and how to level up yours.

Pre-mortems: a 30-minute exercise to plan for success (RevenueCat). Takeaway: Imagining a launch failing and asking why it would is a powerful way to surface risks before they become problems.

20% of your churned users will come back – but are you ready? (RevenueCat). Takeaway: Don’t assume you should focus on Reactivation; the impact varies hugely from app to app.

I also built a free Reactivation Calculator for that last article (had to get on the vibe coding train, of course), luckily, Ethan Garr helped guide me through it.

What I’ve been presenting

Experimentation Elite — I attended with Women in Experimentation and, in one of the workshops, built my first AI agent with Iqbal Ali. Very cool to go from “I should learn this” to actually doing it.

Women in Experimentation dinner — We hosted a lovely evening with Convert on being (or becoming) a female leader. The challenges, the lessons, and how we support each other. The thing that stuck with me most: how many women in the room had quietly talked themselves out of opportunities before anyone else got the chance to. We’re working on that.

What’s been happening with clients

1. My client did a SubClub interview about their growth!

Here it is: How Removing the Free Trial Grew Monthly Subs 2000% – Nancy Anderson, Natal

Definitely check it out. Here are some of the interesting learnings:

• Socials should be focused on building trust and adding value, and at most, a soft sell when relevant

• Trust is such an important factor, even if hard to measure

• The importance of a deeper one coherent app vs multiple niche apps

Really proud of her and the work we are doing.

2. Built my first email flow through an AI agent in customer.io.

Pros: much faster to set up, generated good ideas and different approaches I might not have thought of.

Cons: the copy got repetitive and shallow, so I rewrote quite a bit and improved my prompt for next time. Still, the speed was great.

3. Also used Replo AI for the first time to build landing pages.

They describe it as “Lovable for Shopify” — funny, because Replo existed long before Lovable did. But it’s a fair shorthand. Much faster than their old block editor, and the output was solid enough that I wasn’t fighting it the whole time.

The theme of this month, other than freelancer guilt? AI is helping me deliver more value, faster — and freeing me up to focus on where I add the most: strategy.

This month’s flip-flops

Freelancer guilt. I’ve already shared this one. It’s a mindset thing I’m working on. Progress, not perfection.

The WiE Summit speaker call exploded. I posted on LinkedIn looking for speakers for the Women in Experimentation Summit (coming in September!) on a busy day... and my DMs went from manageable to 400+. Thankfully, Marcella Sullivan (WiE co-founder) stepped in and handled a lot of the responses. I dread to think what it would have looked like without her.

Excited to go through all the applications soon, though! We got around 100 applicants, which was amazing.

Pretty sure the era of LinkedIn Carousels is over. I love carousels, as you probably know, but a recent update makes them all hard to read (you see in the preview 2 and a bit of slides vs 1).

So my carousel days may be over… since the update, reach seems to have decreased dramatically. Why LinkedIn? Why?

My little plants died / never grew. I tried to plant things this year, again. I spent hours on it. And I even roped my husband into watering them on a strict schedule while I was away.

Maybe it was the heatwave (or me, given my track record), but nearly nothing grew, and what did died. So much for my year of homegrown fruit and veg.

I pitched an experiment that didn’t land. Something felt off as I was working it out, but I couldn’t articulate it until I talked it through with someone. The plan's complexity didn’t match the audience.

It reminded me of a question I wasn’t asking often enough — and one that’s easy to skip when you’re in execution mode:

Is our priority learning or short-term impact?

Different answers lead to very different experiments. Ironic, given I literally wrote a whole article on learning roadmaps. Oh well, I’m a work in progress too.

What I’ve been enjoying

Ecuador! Swimming with penguins and sea lions in the Galápagos was a true bucket list trip. The hikes through the Amazon and around Laguna Quilotoa, a volcanic crater lake, were stunning too.

Finished reading Company of One by Paul Jarvis. A good reminder that growth for growth’s sake isn’t the goal, and that staying intentionally small can be a strategic choice, not a failure. Timely, given how I’ve been thinking about my own business.

Building out my AI workflows. Between the agent-building mini unplanned hackathon at Experimentation Elite and experimenting with tools for client work, I’ve been investing in my “second brain.” The goal: reduce cognitive load and protect my energy for the work that actually needs me.

Recording a podcast about my journey. I did an interview this month about my path into growth, dealing with ADHD, and more. Talked about things I don’t usually say out loud. Will share when it’s live.

If you missed my newsletters this month

Looking ahead to June

Continuing to build my AI skills — both for client delivery and for my own workflows. It’s genuinely changing how I work. I’ve invested in an accelerator course, so I'm excited to dig into that.

Sub Club Live with Alice Muir — now happening July 2nd, so technically not June, but I’ll start preparing. Between the two of us, we’ll cover everything from “Should you even focus on reactivation?” to the tactics that actually work.

No more holidays planned for a while, so you’re stuck with me for the coming weeks. Let’s get to business.

Till next time,
Daphne