Founder Playbook · Sub Club by RevenueCat
11 tactics from David Barnard
Lessons from a Lackluster Launch — David Barnard, Weather Up
Watch the full episode“the big launch philosophy has been one of like the great mistakes of my career... for me it's definitely fear right is like oh I don't want to... I'm not going to show this to the world because I need this one little extra thing otherwise people are going to ignore me.”
Holding Features for a Big Launch Is Fear Dressed Up as Strategy
David spent four years bundling Weather Up 3.0 features — Apple Watch app, widgets, interactive widgets — waiting until it felt big enough to launch. What felt like strategic packaging was really delay rationalized as planning. During that time downloads dropped to 15 a day, the app fell out of Apple feature rotations, and he was funding it personally. The subscription model removed every incentive to hold features back; he just never updated his mental model to match.
“except for the getting attention there's no incentive to hold back features like launching a big 3.0 where you're doing a paid update where everybody's paying you for the features... I should have just launched them and kept moving.”
Subscriptions Make Every Small Update a Revenue Opportunity — Stop Saving Features
In the paid-upfront era, holding features for a major version made sense — the new paid SKU was the monetization event. Subscriptions changed everything: the user is already paying, so every shipped feature compounds the perceived value of the subscription being renewed. David explicitly names this as the cognitive update he failed to make: subscription business model, box-software launch mindset.
“that's a top tip sub Club listeners if you're going to launch your app don't do it on the day that Apple announces the dma... if you put all your eggs into one big launch basket you're sometimes just going to catch a stray.”
Don't Launch on the Same Day as a Major Platform or Industry News Event
Weather Up 3.0 launched the exact hour Apple announced DMA compliance — sucking all press attention out of the room. Several journalists who planned to cover it either published early (killing the coordinated rollout) or dropped the story entirely. Any single-bet launch is exposed to random news events, whereas frequent small launches statistically average out the bad-timing risk.
“this is the YC advice if you have a launch that goes poorly just launch again the definition of a bad launch is nobody sees it so nobody's going to see you do it again right.”
Launch Again Immediately After a Bad Launch — A Failed Launch Is Just Invisible
The counterintuitive cure for a bad launch is another launch. Because a failed launch is invisible by definition — most people never saw it — there is almost no reputational downside to relaunching within weeks. David committed to shipping one or two new features, bundling them as a mini-relaunch, putting it on sale, and starting again. Momentum compounds; dread of a re-do is the only real obstacle.
“my biggest regret is not the $40 a year is that I didn't do a sale on launch... all that attention I did get most of the people who wrote about it listed the price and a lot of people... saw $40 and were like no way but if they had seen $40 but 50% off I think we would have done better.”
Run a Launch Sale: Attention Without a Price Incentive Converts at a Fraction of the Rate
The press exposure was real: TechCrunch, 9to5Mac, MacRumors, Boy Genius Report. But everyone listed the $40/year price, and readers who might have subscribed with a discount impulse-purchased nothing instead. For any attention-grab launch, combining the moment with a time-limited offer (25-50% off) is the missing multiplier — it creates urgency where curiosity alone produces bookmarks.
“I priced based on things that I knew and based on my thinking of the value of weather up and my costs in servicing weather up not on the customer perceived value of weather up and I think that's a huge thing that Founders can get themselves tied up in knots about.”
Price to Perceived Value, Not Your Costs — Customers Don't Care What You Pay for API Data
David knew his weather-data API costs were high and set pricing to guarantee margin — a classic cost-plus mistake. Meanwhile, his target audience (weather nerds) already owned several weather app subscriptions, knew market prices, and immediately benchmarked Weather Up against Carrot Weather at $30/year. Value-based pricing is calibrated to what the customer believes it is worth, and for commodity categories, that means knowing competitor pricing cold.
“the headline is that realized LTV less than a month in... is 60% lower on the people who saw monthly... of the people who saw the annual as a default around 800 of them went with annual but almost 400 selected monthly.”
Pre-Selecting Annual Drove 10% More Conversions and 60% Higher Realized LTV Than Monthly Default
David ran his first-ever A/B test: annual pre-selected vs monthly pre-selected. Monthly default got 10% more total conversions (people clicked through more at $4 vs $40), but realized LTV at 30 days was 60% lower because most stayed monthly. Annual default lost that top-of-funnel edge but captured far more up-front revenue per subscriber. For most apps, defaulting to annual is the right call even knowing some users will switch to monthly.
“instead of hundreds of words I mean you just have so little time to capture somebody's attention so the main way I pitched it was like a DM on Twitter and like hey I'm working on some really cool interactivity for weather... and I got quite a few people who are like wow that looks really cool.”
Pitch Press With a Video Demo — Journalists Pass on Text Pitches They Would Act on After One Watch
David initially intended the feature demo video for the paywall, but it was so effective at communicating the value of interactive widgets that it became his entire press pitch. Journalists who replied they no longer cover apps changed their tune after watching the clip. In a text-heavy press inbox, a 30-second visual demo does the selling work that paragraphs cannot — especially for any feature that is inherently spatial or animated.
“of those maybe 30,000 people... all the headline oh another weather app I don't care even if they are potentially even my like ideal customer profile they already have three weather app subscriptions and they don't even get past the headline.”
Press Attention With Weak ICP Overlap Barely Converts — Don't Confuse Eyeballs With Buyers
Weather Up was covered by TechCrunch, 9to5Mac, MacRumors, and others — meaningful press by any indie standard. But David estimated the actual overlap between people who read those outlets, cared about weather apps, did not already have a preferred one, and were willing to pay $40 was tiny. For most consumer apps, press is a low-conversion channel unless the ICP and readership align tightly.
“things really dropped off the map like apple stopped featuring the app because it hadn't been updated in a while we stopped getting any kind of bumps from these new updates... downloads were at low double digits a day like 15 downloads a day.”
Not Shipping Is Compounding Harm — Stale Apps Fall Out of Features and Lose Organic Momentum
Between 2019 and Weather Up 3.0, David withheld updates for four years while working toward the big launch. The compounding cost was invisible until it was not: Apple editorial stopped featuring the app, organic discovery dried up, revenue fell below costs, and he was personally funding the deficit. Each month of non-shipping is not neutral — it is actively losing ground in editorial algorithms, press relationships, and user momentum.
“I didn't even have to talk to Brock to launch that second experiment I changed the price on App Store Connect I already had two SKUs and I set it all up in Revenue cat to run this subsequent experiment without touching code without updating the app.”
Run Paywall Experiments Without a Developer by Using Backend-Driven Paywall Tools
David is a non-technical founder who works with a developer partner. The old workflow for A/B testing a paywall meant involving the developer for every variant — days of work. By using a backend-driven paywall, he could change pricing, pre-selected plans, and run scheduled experiments entirely from the dashboard — no code change, no App Store update required. For any founder without a full-time developer, this unblocks a critical growth lever.