How we estimate installs and revenue

Shopify does not publish install counts or developer revenue. Every "Est. Installs" and "Est. Lifetime Revenue" figure on this site is the output of our own model, built from public signals on the App Store.

Treat the estimates as order-of-magnitude guidance, not accounting. They are most useful for comparing apps and niches against each other, and least reliable for any single app.

What goes into the model

• Reviews. Only a small share of merchants who install an app ever review it. Review volume is the strongest public signal of how many stores are using an app.
• Rating. How willing people are to leave a review depends on how they feel about the app, so the rating changes how we read the review count.
• Pricing. The plans an app offers, whether it has a free tier, and how it bills (monthly, annual or one-time) shape how many installs turn into paying customers and what they pay.
• Age. When an app launched limits how long it can have been earning, and how much of its early customer base is still around.

Installs

We estimate how many stores each review represents and scale up from there. That ratio is not fixed: it varies with the app's rating, and we keep calibrating it. Apps with no reviews yet still get a small baseline, because a listed app almost always has some users.

Lifetime revenue

From the estimated installs we work out how many are likely to be paying, what a typical paying merchant spends, and for how long. The result is an estimate of what the app has earned over its life so far, not a monthly figure. We lean conservative throughout: when in doubt, the model assumes merchants sit on cheaper plans and churn sooner. Apps with no paid plan are estimated at $0.

Known limitations

• Review rates vary by category. Apps merchants interact with daily (marketing, reviews, upsells) get reviewed far more than background utilities, so utilities are probably under-estimated.
• Usage-based and per-order pricing is hard to model from a listing, so those apps carry wider error bars.
• Enterprise or custom pricing is invisible to us and ignored.
• Very new apps and apps with only a handful of reviews have the widest error bars.
• We keep refining the model as better signals become available, and changes apply to every app at the next nightly recalculation.

When are the numbers refreshed?

Estimates are recalculated every time an app page is re-scraped and again in a nightly pass over the whole database, so a change in review count or pricing shows up within a day.