Skip to content
AppDues
Log in
Contact salesSign up

How to Launch an App: A Six-Stage Plan for Subscription Apps

A launch is not one event. It is a sequence of decisions, each of which can be tested cheaply before the next one gets expensive. This guide walks through six stages — concept and market research, a minimum viable product, pre-launch audience building, monetization choice, launch day, and post-launch retention — with the questions to answer at each gate. Examples from CardPointers, TalkingParents, Wavve Boating, Nomorobo, and Opal show how different teams answered them.

Stage 1: Concept and market research

Most app ideas start with a problem the founder has personally run into. Emmanuel Crouvisier built CardPointers after repeatedly fielding questions from his parents about which credit card to use for gas or groceries — a small annoyance that pointed to a broader gap in how people manage card rewards.

That personal origin is useful, but it is not the same as a value proposition. A unique value proposition (UVP) is a single clear statement of the problem you solve or the experience you offer that nobody else does. It is not a feature list. Crouvisier's UVP — simplifying credit card rewards — carried CardPointers from zero revenue to a full-time business in under a year.

With a UVP drafted, the next step is to test it against real people. "Market research" can sound grand, but in practice it often just means finding a way to talk to your intended audience. Two patterns from the source examples:

  • Survey the audience you already have. Before TalkingParents released its mobile app, the team surveyed users of its web product with plain questions: what do you want, what would you pay for, which features matter, how could this be better? The answers shaped the app. The team kept surveying afterward, using the results to decide which features to bundle when prices changed or a premium tier was introduced. If you have no existing users, survey the target market directly instead.
  • Prototype before you build. Adam Allore, founder of boat navigation app Wavve Boating, knew from marina work that people struggled to read nautical charts. He built a rough map with free GIS software and nautical data, hosted it on a webpage, displayed it on iPads at a boat show as if it were a finished app, and made a makeshift brand. Attendees responded with real interest and went looking for the app in stores. The booth cost was low; the validation was worth far more than the prototype itself.

Gate to pass before moving on: you can state your UVP in one sentence, and at least some members of your target audience have told you — in their own words — that the problem matters and they would use or pay for a solution.

Stage 2: Build a minimum viable product

Once the concept is validated, the temptation is to build the full app. Resist it. Spending heavily on a complete product before anyone has used it is the expensive way to learn the same lessons.

A minimum viable product (MVP) is the simplest version of your app that still delivers your UVP. The goal is not every feature — it is the right features for early adopters, plus meaningful feedback to guide what comes next.

CardPointers shipped its MVP as a Progressive Web App (PWA), which combines a website's reach with app-like features. That let Crouvisier get to market quickly and test his core hypothesis: that users wanted a straightforward way to maximize card rewards. Reaching 5,000 PWA users was the proof point that justified building the full iOS app.

Gate to pass before moving on: real users have used the MVP, and their behavior or feedback tells you whether the core hypothesis holds. If it does not, iterate on the MVP rather than scaling spend.

Stage 3: Pre-launch strategies

Interest can be built before the app is available. Nomorobo, a robocall blocker, is a useful model. Founder Aaron Foss entered the Federal Trade Commission's 2013 Robocall Challenge because the contest was low-risk and high-reward: he could apply his telephony background to a widespread problem with minimal financial outlay. He won.

With prize money secured and the app in development, Foss focused on three things:

  • Use the media attention. The contest win brought press coverage. Rather than let it fade, he used it to publicize the app and the problem it solved.
  • Build a waiting list. He collected more than 30,000 email addresses before launch — a measure of interest and a ready first audience.
  • Stay transparent. Open communication after the contest built a community of supporters invested in the app's success before it shipped.

The longer-term outcome reported at the time of this account: Nomorobo grew to roughly 2 million users and had blocked about 3.5 billion robocalls. Treat those as dated figures from that account rather than current totals.

Gate to pass before moving on: you have a channel to reach interested people at launch (a list, a community, or earned coverage) and evidence that the interest is real rather than assumed.

Stage 4: Choose a monetization model

The monetization model should fit your value proposition, your users' expectations, and market norms. Three common options for subscription apps:

  • Freemium — some high-value features and content sit behind a paywall while a basic level of access stays free.
  • Fully locked / hard paywall — users must subscribe to use the app. Limited free use may exist only to nudge people toward paying.
  • Hybrid — subscription revenue mixed with other streams such as non-renewable in-app purchases, ads, partnerships, e-commerce, or affiliate marketing.

A hard paywall from day one can feel counterintuitive, because at launch your app probably delivers the least value it ever will. Opal, a screen-time management app, went subscription-first anyway. Founder and CEO Kenneth Schlenker described phase one as proving the opportunity — not only that people would use better screen-time software, but that they would pay for it. Getting users to pay early, in that view, is a strong signal that the product is genuinely valuable. Opal reached $5 million in annual recurring revenue within about a year.

That is one path, not a rule. The trade-off to weigh: a locked app raises the bar for the first impression, while a freemium or hybrid model lowers it but may delay the signal on willingness to pay.

Gate to pass before moving on: you can name the model you chose, why it fits your UVP, and what early metric would tell you the choice was wrong.

Stage 5: Launch day — and launch again

Drop the idea of a single "big launch." Holding features back so they can ship together works against the subscription model, where steady small releases keep value visible over time. And any one launch may not go to plan for reasons outside your control.

What to do instead:

  • If a launch goes badly, few people likely saw it — so launch again.
  • Prefer regular small releases. Opal shipped a new version every week and kept the cadence up.
  • With the right narrative, the same improvements can be launched more than once, placed in new contexts and framed as new releases.

Gate to pass before moving on: you have a release cadence you can sustain, and a plan for what happens if launch day underperforms.

Stage 6: Post-launch growth and retention

After launch-day excitement fades, the work shifts to sustaining growth and keeping users. The number of possible next steps can be overwhelming; a simplified growth stack for small teams, adapted from Andy Carvell's Mobile Growth Stack by Thomas Petit, gives a practical order of operations.

  1. Engage and understand your audience. Go where users gather — Reddit, Discord — and use simple surveys. Direct emails or calls can surface insights no dashboard will. This grassroots habit matters as much after launch as before it.
  2. Refine the product continuously. Use feedback to iterate. Ask what makes the app unique and why users would choose it over alternatives, then focus on the features that drive retention. Opal's weekly release schedule is the clearest example: a steady cadence of experimentation and rapid improvement based on feedback and performance data.
  3. Use data for decisions. Set up event tracking and a data taxonomy early. Tools such as Firebase can provide behavioral analytics and crash analytics, so improvements target what actually affects retention and satisfaction.
  4. Optimize onboarding and activation. A smooth first encounter matters. Clear walkthroughs and early funnels let you see user behavior from the start, which shapes initial engagement.
  5. Monitor and adapt to churn. Early users churn at higher rates — expect it. Cohort analysis shows why people leave and what keeps them longer, and patterns emerge over time.

On churn specifically, investor Eric Stromberg's framing is worth keeping: a first-month churn rate compared with a mature company's is apples to oranges, because new customers are by definition early in their lifecycle. What matters is whether churn reaches an asymptote — 15% in month one, 10% in month two, 7.5% in month three, then leveling off — rather than falling in a straight line.

Gate to pass before moving on: you are tracking activation and churn by cohort, and you can point to at least one product change made in response to user evidence.

Launch readiness checklist

  • UVP written. One sentence that says how your app solves a problem no one else solves the same way.
  • Market validated. Surveys, interviews, or prototype testing show real demand from your target audience.
  • Prototype tested. A rough version has been shown to potential users and their feedback has shaped the build.
  • MVP scoped. The first release delivers the UVP with the right features, not all features.
  • Pre-launch audience built. A waiting list, community, or earned coverage gives you people to reach on day one.
  • Monetization chosen. Freemium, hard paywall, or hybrid — with a reason tied to your UVP and a metric that would flag a wrong choice.
  • Release cadence set. Small, regular launches rather than one big event, with a plan for a launch that underperforms.
  • Retention tracked. Activation and churn measured by cohort, with feedback loops feeding product changes.

Frequently asked questions

Do I need a finished app before I test demand?

No. Wavve Boating's founder tested a simplified map built with free GIS software, shown on iPads at a boat show as if it were a working app. The point was to see whether boaters wanted the concept before paying for full development.

Is a hard paywall from day one a bad idea?

Not automatically. Opal launched subscription-first to prove people would pay for better screen-time software, and the willingness to pay was treated as a product-market-fit signal. The trade-off is that a locked app at launch offers the least value it ever will, so the paywall has to be justified by the core promise.

What churn rate should I expect right after launch?

Early cohorts churn faster than mature ones, so a high first-month figure compared with established companies is not apples to apples. What matters is whether churn levels off over successive months rather than falling in a straight line.

Should I save features for one big launch?

Usually not. Small, regular releases keep momentum, and a launch that goes badly is often seen by few people — so you can launch again. The same improvements can be re-framed in new contexts.