Privacy AI · B2C · offline private chatbot
Test app icons first, including wildly different versions
Guest says icons move conversion most; no numbers
Unknownaso
What they did
Icon moves conversion more than screenshots or titles in their experience. For Privacy AI: icon signalling security plus the standard AI sparkle, run ~6 icon tests in 6 weeks. For the book-summary app: test an extreme alternative (cartoon/caricature generated via ChatGPT).
What happened
Guest: icon 'moves the needle' the most across their tests.
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- Fitbit to Apple Health Sync: Turn a $5.99 paid app into a free download with a subscriptionRevenue jumped from ~$200/day to ~$2,000/day overnight
- Bickster: Sell lifetime access instead of subscriptions unless the app has real running costsWorks well across their apps; subscriptions drew complaints on one
- Fitbit to Apple Health Sync: Offer $59.99 lifetime or $39.99/year with a 3-day trial on a closable paywall~6% of installs buy; best revenue per 1,000 users of the variants tested
- Fitbit to Apple Health Sync: Keep installs about half organic, half paid; fund only channels returning 2-3xOn track for ~110K MRR
- Fitbit to Apple Health Sync: Use Apple Search Ads only if the cost per install fits your numbersBest paid channel for this app; too pricey for many of their others
- Fitbit to Apple Health Sync: Run Google and Bing search ads and feed sales back to them every nightTwo of the three main paid channels; Bing small but profitable
- Bickster: Test a one-screen landing page against sending search clicks straight to the App StoreDepends on the app; no general rule
- Test lifetime prices from $4.99 up to $49.99 at fixed ad spend to find the ceiling$29.99 beat $49.99 on day-one revenue; a later $39.99 now beats $29.99
- Bickster: Build quick AI-coded versions of several app ideas; keep the one that sells
- Chatster: Spin off an AI Bible app for chat users who asked religious questionsSold at $29.99 lifetime with little effort; ads near break-even
- Launch with a hard paywall, one offer and no free trialOne offer gave the clearest pricing lessons
- Test with exact-keyword Apple Search Ads, then scale on MetaMeta converted even better than Apple Search Ads
- Bickster: Set Meta ads to find buyers, not installs, and test videos at ~$500/dayInstall-to-purchase rose from ~6% to ~18-20%
- Title the App Store page with plain search terms and put the brand last
- Bickster: Test dropping the onboarding screens entirelyNo onboarding won in several apps; even one welcome screen hurt
- Bickster: Use Adapty to change prices and paywalls without an app update
- Chatster: Show one offer on the paywall with a 'view all plans' link
- Bickster: Lay out paywall plans side by side instead of stacked~2 points higher conversion (guest unsure of the exact number)
- Bickster: Skip paid attribution tools; use Meta's own numbers, in-house tracking and Mixpanel
- Put a hard paywall with one lifetime price on an app ranking in search
- Run Meta ads at ~$150/day for a fitness app for people 65+Nobody bought after $3-5K; buyers came from App Store search