Skip to content
AppDues
Log in
Contact salesSign up

How an AI App Studio Uses Revenue Data to Decide What to Scale

A studio that ships several consumer apps at once faces a different question from a single-product team.

It is not only whether one app converts; it is which app, channel, and pricing model deserve the next round of budget and engineering time. This guide lays out the operating logic behind that decision, using the reported experience of KRNL, a Milan-based AI app studio, as a worked example. The studio's own account describes making a centralized monetization platform the source of truth across seven apps and using it to guide investment decisions.

The real cost of unclear revenue data

In a portfolio business, muddy revenue numbers can push the team toward the wrong app, the wrong campaign, or the wrong pricing model. KRNL encountered this in the early Loopsie period, when the team was trying to understand revenue across platforms, track trial conversion, and optimize paid acquisition. Without a clean monetization layer, even simple questions were hard to answer with confidence.

The stakes were high because the studio was bootstrapped and moving quickly. Revenue data shaped how much the team could spend, how fast it could scale, and whether to keep pushing a product or move on. The studio's CEO framed the failure mode bluntly: if revenue metrics are wrong, the team is effectively flying blind and might as well ignore the numbers and wait for the monthly wire transfer.

A useful distinction: unclear revenue data is an instrumentation problem, while a weak product signal is a demand problem. The first can be fixed with better event tracking and a centralized source of truth; the second requires product or positioning changes. Confusing the two leads studios to kill healthy products or over-invest in weak ones.

A monetization foundation for every new bet

KRNL adopted a centralized monetization platform early and made it a standard step in the app launch process. The studio reports using it across all seven apps. The launch sequence described is: build the app, add the monetization layer so the most important metrics, the revenue ones, are tracked from the start, then continue with the project.

The foundation also removes work the studio does not want to own on every new product. The team avoids rebuilding the same purchase infrastructure, maintaining Apple server notifications, connecting Stripe webhooks from scratch, and keeping up with platform changes around proceeds, taxes, and payloads. That frees engineering to focus on what differentiates each app: user experience, creative output, pricing model, and growth strategy.

Turning monetization data into operating decisions

The value showed up first in acquisition. In the early days, the team sent trial events as purchases because it lacked a good server-to-server way to track trial conversion. Once the monetization layer was in place, KRNL could send the right events with the right numbers. The studio's CEO recalls that this coincided with changes in Facebook campaigns and was the moment the team became convinced of the value.

The same principle still applies. Revenue events, webhooks, and charts help the studio understand cohorts, subscription retention, payback time, pricing, cash flow, and overall LTV. Those metrics influence how spend is allocated, which products get attention, and which apps should be wound down. The CEO's account is that this data made it possible to judge which app was doing better, which made sense to keep working on, and which made sense to kill.

Scaling with trusted revenue signals: the One More Shot AI example

One More Shot AI gave the studio a clear signal that something was different. Compared with earlier apps, the CEO saw stronger retention, stickier usage, and a clearer user persona around music artists creating promotional content.

The product also monetizes beyond a single subscription. Users subscribe for access and can buy additional packages when they want to keep creating more videos, covers, images, or edits. The important question is not only whether users subscribe, but whether they keep using the product enough to come back and spend again.

In the revenue curve, the studio looks for a pattern where revenue drops after the first month and then rises again as engaged users return, create more, and buy additional packages. The CEO describes this as the point where the team understands it is building not just a subscription generator but a product people want to keep using: the more they use it, the more useful they find it, and the more they spend.

A worked reading of that pattern: a dip after month one is normal as trial and first-month subscribers lapse. The recovery is the signal. If the curve rises again, the product is creating repeat demand and additional package purchases; if it stays flat or keeps falling, the studio has a retention or value problem rather than a pricing problem. That distinction is what makes the revenue curve useful for deciding where to keep building.

ConfigureTargetTest

Field and control map for portfolio revenue decisions

The following map connects the signals described in the studio's account to the operating question each one answers. It is a reading aid, not a schema.

  • Revenue events and webhooks: feed acquisition and attribution systems with correct trial and purchase signals. Question: are campaigns optimizing toward real conversions?
  • Cohorts: group users by start period to compare behavior over time. Question: does retention hold after the first month?
  • Subscription retention: measures whether subscribers keep their plans. Question: is the product creating durable value?
  • Payback time: the period before acquisition cost is recovered. Question: can the business fund the wait?
  • Pricing: the structure of subscription and additional packages. Question: does the model match how users actually consume?
  • Cash flow: the timing of money in and out. Question: does reported revenue match money actually received?
  • LTV: the expected value of a user over time. Question: is value rising with usage or only with spend?
  • Revenue curve: the shape of revenue after the first month. Question: does the product generate repeat demand?

About the studio in this example

KRNL describes itself as an AI app studio building consumer apps such as One More Shot AI, Loopsie, and Sharp AI, used by tens of millions of people. Its website is krnl.ai. The reported outcomes in this guide come from the studio's own account and are presented as such, not as independently verified results.

Frequently asked questions

Why centralize revenue data across App Store, Google Play, and Stripe?

Because a portfolio studio needs comparable numbers to decide which app, channel, or pricing model deserves more investment. When revenue is spread across separate systems, even simple questions about trial conversion or total revenue become hard to answer with confidence, and the team risks optimizing toward the wrong signal.

Why does it matter whether trial events are sent as trials rather than purchases?

Sending trial events as purchases distorts acquisition data. Campaigns then optimize toward the wrong conversion, and the team cannot tell whether a trial actually converted. Correct event tracking gives the acquisition system accurate signals to work with.

What does a revenue dip after the first month followed by a rise indicate?

A dip is normal as early subscribers lapse. A subsequent rise suggests engaged users are returning, creating more, and buying additional packages. That pattern points to repeat demand rather than a one-time subscription spike. A flat or falling curve points to a retention or value problem.

How do studios decide when to wind down an app?

The reported approach is to compare apps on cohorts, subscription retention, payback time, pricing, cash flow, and LTV, and to use those numbers to judge which app is doing better, which is worth continuing, and which should be wound down. The decision is a comparison across the portfolio, not a single metric in isolation.