App Growth 2026六大工作坊：语言库、AI广告与营销漏斗实操

At App Growth Annual 2026, over 5,000 people signed up to watch the day online and take part in six virtual workshops. Now, we’re opening the virtual vault: all six of these previously-exclusive workshops are now free to watch, complete with the presenter’s full slide deck, handouts, and our curated further reading to help put the workshop’s lessons into practice.

Missed the rest of the day?

Catch up on the keynotes from Babbel, Burner, Tolan, and Ravi Mehta (ex-Klarna) — plus five experiments they inspired — in our full App Growth Annual 2026 recap.

Lisa Kennelly (Klarna) — Copy that converts: the language-market fit teardown

AI-written onboarding and paywall copy often reads fine, but it fails at conversion. Because it doesn't use your customers' language. Lisa Kennelly (Growth and Product Market Strategist, Klarna) gave a great workshop on how to write copy with language-market fit: copy that sounds the way customers already speak.

Lisa walked through her process for mining App Store reviews, competitor reviews, cancellation reasons, support tickets, Reddit thread, and interviews, where she looks for five things:

• The outcome users want
• The trigger that made them download
• The alternatives they tried (including doing nothing)
• Their anxieties
• The exact words they use

Everything goes verbatim into a single ‘language bank’ that every draft starts from. Before shipping, every line faces a five-point check — starting with whether it uses a phrase customers actually repeat.

Watch on YouTube

Resources:

Download Lisa's workshop slides Download Lisa’s handouts: prompt pack + language bank template

Further reading:

• 📕 From review to roadmap: how mining reviews can inform your app's features by Daphne Tideman
• 🎧 Sub Club podcast: How to Scale Your App Marketing Team from Scratch ft. Lisa Kennelly
• 📗 Start here: a paywalls study guide

Bria Sullivan (Honey B Games) — The playbook behind 23 million organic downloads

Bria Sullivan's game Boba Story has 18 million organic downloads, and Focus Friend — the gamified focus timer built with Hank Green — passed 5 million, most of them arriving long after Hank stopped posting about it. Her secret is a simple, two-pronged repeatable research funnel:

• Short surveys to find people who already use apps to solve your problem
• Customer interviews to understand them
• Disciplined pattern-finding in the feedback

Surveys ask what; interviews ask why and how, starting with the user's relationship to the problem before you show them anything. Bria’s workshop walks through her research process, including how she tests her ideas, the questions she asks, and how many interviews to aim for.

Watch on YouTube

Resources:

Download Bria's workshop slides

Further reading:

• 🎧 Launched podcast: The Hail Mary pitch that launched a #1 App Store hit ft. Bria Sullivan
• 📕 How user interviews can inform your ad creatives by Nathan Hudson
• 📗 Don't just solve a problem — solve it emotionally by Daphne Tideman

Sven Jürgens — From competitor ad to finished creative, in one hour

Your competitors have already paid to learn what works in your category — and thanks to the impression filter in Meta's Ad Library, their best performers are public. At App Growth Annual, growth consultant Sven Jürgens shared how he uses these competitor ads to inform his own creative, with AI-generated video fast-tracking the process.

Sven’s session shows his nine-step AI workflow in real time, with the audience voting on the app and actor. About 10 minutes and roughly $3 later, out came a lip-synced UGC-style Cal AI ad. With the heavy lifting done by AI, Sven spent the rest of the workshop on what actually makes a good output:

• Research into your users' pains and awareness stage
• Rich context (brand guidelines, user insights, real app footage) rather than clever prompts
• Ads built from swappable ‘Lego blocks’ (hook, body, call to action) so winners are fast to iterate

The full workshop covers Sven’s AI ad process, his golden rule for reporting, copywriting guidelines, brand examples, and how to structure your documentation.

Watch on YouTube

Resources:

Download Sven's workshop slides Download Sven’s nine-step diagram

Further reading:

• 📕 AI-generated ads: balancing attention and trust in user-generated content by Anthony Morcillo
• 🎧 Sub Club podcast: How to Go Viral on TikTok (and Profit From It) ft. Joseph Choi
• 📗 Forget chasing viral ads — build a creative testing system that scales by David Vargas

Yuliya Lennox (Genesis Tech) — Nothing works until this does: the app marketing sequence

Most apps don't stall because of bad tactics; they stall because they skipped a step. In her workshop, growth marketer Yuliya Lennox (BetterMe, Replika, Palta) presents app marketing as four stages, each one a prerequisite for the next:

• Prove there's demand
• Build a journey that converts to paid
• Make the economics work
• Scale

Her advice for stage one: charge from day one (free-user retention tells you nothing about willingness to pay), test demand with a simple funnel before building the full product, and test genuinely different creative concepts rather than 20 variations of one slogan. Once something works, she pushes teams to count every cost — store fees, payment fees, AI infrastructure, refunds — and aim for payback as close to day one as possible.

Watch on YouTube

Resources: Download Yuliya's workshop slides

Further reading:

• 📕 Finding product-market fit: a guide for subscription apps by Daphne Tideman
• 🎧 Sub Club podcast: Make Ugly Ads to Grow Your App ft. Yuliya Lennox
• 📗 Stop chasing growth hacks — fix your funnel first by Hannah Parvaz

Marcus Burke (Ctrl Shift) — 6+1 AI workflows to run and scale Meta ads from your phone

Checking whether one Meta ad is profitable can mean hopping between Ads Manager, Events Manager, your MMP, RevenueCat, product analytics, and a spreadsheet. Meta ads advisor Marcus Burke now has AI do the tab-hopping for him.

His setup is deliberately simple: Notion as a central, model-agnostic ‘brain’ holding his SOPs, client context, and a change log, with Claude connected to it, and to his data tools via MCPs. On top sit the workflows he runs across client accounts, including a regular ‘account patrol’, age and placement breakdowns that predict trial-to-paid before the data matures, and value rules. Others cover format research across ad libraries and TikTok, voice-of-customer copywriting, and an ad–onboarding fit check that flags when a scaling ad promises something your funnel never pays off.

The new workflows have cut Marcus’s time dramatically, from two hours to 10 minutes in some places. Check out his full workshop for a detailed walkthrough of his AI setup and each workflow.

Watch on YouTube

Resources:

Download Marcus's workshop slides Meta Ads account x-ray

Further reading:

• 🎧 Sub Club podcast: Selling Your Subscription App with Meta Ads ft. Marcus Burke
• 📕 7 Meta ad testing frameworks for subscription apps by Daphne Tideman
• 📺 Sub Club Live: Meta Ads for Apps: Creative, Signal & Scaling in 2026 ft. Marcus Burke

Julien Daver (SplitMetrics) — Make AI-generated ads that actually convert

In our last workshop, Julien Daver (Head of UA at SplitMetrics) shares the AI creative framework his entire team uses to find winning ad creatives.

It starts with a research agent that continuously collects your own data (brand material, audience insights, onboarding answers, campaign results) alongside external sources like competitor ads, Reddit, app reviews, and YouTube comments. A second agent turns that into concepts, angles, and hooks, kept separate so you can trace what actually moved performance — helped by naming every ad after its hook, angle, and concept.