Skip to content
AppDues
Log in
Contact salesSign up

App growth takeaways from App Promotion Summit London 2024

Five practical lessons from App Promotion Summit London 2024 cover weekly pricing, AI-assisted content, web-to-app funnels, simple paywall tests, and onboarding that responds to how users feel. Each example below shows a specific change, the result reported at the event, and how to decide whether it fits your app.

1. Weekly subscriptions can reach a new paying segment

Many dating-app users never pay, not because the product has no value, but because a monthly commitment feels too large or they are unsure how long they will need it. A weekly option lowers that first decision barrier. Grindr used weekly subscriptions to reach people who had not converted to monthly plans. Monthly subscriptions did not fall, which pointed to a new segment rather than cannibalisation. The reported outcome was higher revenue per download than Tinder.

Use this when: your monthly conversion is low, users have short or uncertain needs, or you can serve a distinct price-sensitive segment without discounting the core plan.

Check before launching: confirm the weekly price still covers support and store costs, test copy that explains renewal clearly, and measure incremental subscribers separately from existing monthly buyers.

2. AI works best as an accelerator for a specific bottleneck

Memrise treated AI as a way to produce more, faster, rather than as a novelty feature. Three uses stood out: generating search-focused content for traffic acquisition, adding languages to widen the addressable market, and creating a more immersive learning experience. Its MemBot chatbot was presented as a lower-pressure way to practise speaking a foreign language than talking to a real person, alongside curated video in lessons.

The lesson is to assign AI to a bottleneck you can measure. Content volume, language coverage, and practice engagement each have a clear output. Start with one bottleneck, compare quality and conversion against the previous manual process, and expand only where the gain is repeatable.

3. Treat web-to-app as one continuous funnel

PlantIn reported that 65% of its revenue came through web-to-app funnels. The app store still offers a smooth purchase experience, but a web funnel gives more room to test, localise, and automate the route to install. High-performing landing pages were described as simple, clickable, matched to the user's problem, localised by geography, and built around a clear install call to action, with continuous testing.

Continuity matters as much as the landing page itself. The promise on the web page, the install step, and the first onboarding screens should feel like one journey aimed at the same goal. If the landing page speaks to one problem and onboarding asks about another, users lose momentum.

Checklist for a web-to-app test: one clear promise, one install action, geo-appropriate language, onboarding that mirrors the landing-page goal, and a testing loop that changes one element at a time.

4. Simple paywall changes can outperform complex ones

Microsoft 365 tested two small paywall changes. First, the more expensive family plan became the default. Second, the yearly price was anchored to its monthly equivalent. Both changes produced a significant revenue lift, and price anchoring was also reported to work well for the app Mojo.

Scale does not change the priority logic. Changes that reach the most users tend to matter most, and the fundamentals are often where apps lose money: paywall placement, onboarding, pricing, and packaging. Before building a complex optimisation, test plan order, the default choice, and how a longer-term price is framed against its monthly equivalent.

5. Empathy-based onboarding balances conversion and experience

Headway's onboarding was described as empathy-based: it tries to understand, acknowledge, and respond to user feelings instead of pushing straight to payment. The flow had four stages. First, identify feelings through user research and survey questions; Headway used 35 onboarding questions. Second, acknowledge those feelings with customised feedback screens. Third, offer a solution by showing relevant content and a personalised experience. Fourth, present the paywall after that preparation. If a user tries to dismiss it, the flow keeps offering solutions, including a free trial, a step reported to raise Headway's ARPU by 14%.

Length alone is not the point. A long flow earns its length only when each answer changes the feedback, content, or offer that follows. If answers are collected but ignored, shorten the flow.

How to choose which lesson to test first

A quick way to apply the five lessons:

  • If monthly conversion is weak, test a weekly plan for a distinct segment and watch for incremental revenue, not just total subscribers.
  • If content or language expansion is your bottleneck, pilot AI there first and measure output quality and engagement.
  • If paid acquisition is a priority, build the web page, install, and onboarding as one matched journey and test continuously.
  • If you have not tested basics recently, start with default plan, plan order, and yearly-to-monthly price framing.
  • If onboarding feels transactional, add feedback and personalisation that visibly use what users told you, before showing the paywall.

None of these ideas is universal. The common thread is a small, measurable change aimed at a clear user barrier, tested against the behaviour you actually want to improve.