Key Takeaways
- Forecast app downloads from the promotions you’ll actually run: how many people each email, text, and site placement sends to your App Store and Google Play listings, and how many of those visitors download.
- Group shoppers by the month they first shop in the app, then apply return rates to each group. Monthly active shoppers are the new group plus everyone from earlier months who comes back.
- Monthly app sales equal active shoppers × shopping sessions per shopper × orders per session × average order value, measured the same way as your store’s sales reports.
- After launch, compare each step with what actually happened, fix the first assumption that missed, and update the months still ahead.
An app launch plan usually comes with a sales target. Someone in the business wants to know what the app will bring in, and the answer usually starts as a single number: a download goal, or a share of mobile revenue.
A single number is hard to plan around. It doesn’t tell you how many emails to send, which placements matter, or whether a slow first month is a problem. A forecast built month by month does, because every figure in it traces back to something your team will do or something you can measure.
This guide builds that forecast step by step, following one hypothetical store through its first three months. It starts with the promotions that bring people to your app listings, then moves through first-time downloads and returning shoppers to orders and sales.
If you only need a quick sense of scale first, our guide to how much revenue an ecommerce app can generate uses a simpler model with a single monthly active rate.
How Many Downloads Can You Expect at Launch?
Your app’s launch downloads depend on two numbers: how many people your promotions send to the App Store and Google Play listings, and how many of those visitors download. Both numbers change from month to month, because your promotion calendar does.
A launch email produces one big month, and the site banner and order confirmation page add a steadier flow after it. So forecast downloads by month and by placement, rather than as one launch total. That way, a quiet second month is something you planned for, and a weak placement shows up on its own line.
Who Will See Your App Promotion?
Start with the people you can reach in each month: your email and SMS subscribers, returning visitors to your mobile site, and customers who place an order. Then estimate how many of them will tap through to your app listing.
The best guide to that tap-through rate is how your own audience responded to comparable campaigns. If you use Klaviyo, its campaign performance report shows unique clicks, “the number of unique recipients who clicked your message.”
A loyalty announcement or product launch gives you a starting point. Then adjust it down, because downloading an app asks more of a customer than reading a product page.
Say a coffee store, whose customers reorder every few weeks, plans this launch:
| Placement | Month 1 | Month 2 | Month 3 |
|---|---|---|---|
| Launch email (40,000 subscribers × 3%) | 1,200 | ||
| SMS (15,000 subscribers × 6%) | 900 | ||
| Reminder email (40,000 subscribers × 1.5%) | 600 | ||
| Mobile site banner (returning visitors × 2%) | 600 | 550 | 500 |
| Order confirmation page (6,000 orders × 5%) | 300 | 300 | 300 |
| Listing visits | 2,100 | 1,750 | 1,400 |
The banner line falls each month because the store expects 30,000, then 27,500, then 25,000 returning mobile visitors who don’t have the app yet. Shoppers who download stop seeing the banner.
The same people sit on several of these lists, so count unique people, too. Most of the 15,000 SMS subscribers are probably on the email list as well.
Total downloads can’t exceed the number of different people you reach. So check the plan against that count, as well as against the size of each list.
Promoting your app through other channels, such as packaging or stores, works the same way: give each placement its own line, its own audience, and its own response rate.
How Many Listing Visitors Will Download?
Not everyone who reaches your listing downloads the app. Forecast that rate separately for iPhone and Android, because each store counts listing visits and downloads differently, and your shoppers may not split evenly between them.
The coffee store gets about 57% of its returning mobile traffic from iPhones, so it splits listing visits the same way. It assumes 40% of iPhone visitors and 30% of Android visitors will download:
| Platform | Month 1 | Month 2 | Month 3 |
|---|---|---|---|
| iPhone listing visits | 1,200 | 1,000 | 800 |
| iPhone first-time downloads (40%) | 480 | 400 | 320 |
| Android listing visits | 900 | 750 | 600 |
| Android first-time downloads (30%) | 270 | 225 | 180 |
| First-time downloads | 750 | 625 | 500 |
Count first-time downloads only. App Store Connect separates First Time Downloads from redownloads, and its Total Downloads is “the sum of First Time Downloads and Redownloads.” A customer who reinstalls the app isn’t a new user, so including them would inflate the forecast.
The two percentages are the store’s own assumptions, not benchmarks. Don’t test them against Apple’s headline Conversion Rate, because it measures something different.
Apple calculates that rate as “total downloads and pre-orders divided by unique device impressions,” and impressions include people who only saw your app while browsing or searching the App Store. Your rate compares downloads with listing visits from your own promotions, so build it from those two counts.
Shoppers who find the app by searching the App Store or Google Play belong on a separate line. Leave that line at zero until you can measure it, and treat it as upside rather than part of the plan.
How Many Shoppers Will Use the App Each Month?
A download only matters for sales once the customer starts shopping in the app. So before you forecast active users, decide what counts as active for this forecast.
Analytics tools define the term broadly. For example, Google Analytics counts active users as “the number of unique users who engaged with your site or app in the specified date range.” That includes someone who opens the app to check a delivery date.
For a sales forecast, count shoppers instead: people who view a product, a collection, or their cart in the app during the month.
First Opens and First Shopping Sessions
Some people download the app and never open it. Others open it once to sign in and come back weeks later to shop. So track two milestones after the download: the first open and the first shopping session.
Group each shopper by the month of their first shopping session. Each group is a cohort, and following cohorts separately shows whether the shoppers you won in a strong launch month behave differently from the ones who arrived later.
The coffee store assumes 80% of first-time downloads lead to a first shopping session in the same month:
| Month 1 | Month 2 | Month 3 | |
|---|---|---|---|
| First-time downloads | 750 | 625 | 500 |
| New shoppers (80%) | 600 | 500 | 400 |
Those 600, 500, and 400 shoppers are the store’s first three cohorts.
How Many Shoppers Will Come Back?
Apply a return rate to each cohort for every month after it starts, then add the cohorts together. The coffee store assumes 50% of a cohort shops again the month after it starts, and 35% two months after. That second rate counts everyone active in that month, including shoppers who skipped the month in between.
| Cohort | Month 1 | Month 2 | Month 3 |
|---|---|---|---|
| Started in Month 1 (600) | 600 | 300 | 210 |
| Started in Month 2 (500) | 500 | 250 | |
| Started in Month 3 (400) | 400 | ||
| Active shoppers | 600 | 800 | 860 |
New downloads fall each month in this plan, but active shoppers keep rising. By Month 3, more than half of them (460 of 860) are returning shoppers from earlier cohorts. A month-by-month forecast makes this pattern visible: repeat use builds up across cohorts, while launch downloads arrive once.
Choose return rates that match how often your customers buy. A store whose customers reorder coffee every few weeks can expect some of them back the following month. A furniture store can’t, and its forecast should show smaller, later returns.
Your retention data from email or the website is the best starting point before the app has its own.
From Active Shoppers to Orders and Revenue
To turn active shoppers into sales, multiply four numbers:
Monthly app sales = active shoppers × shopping sessions per shopper
× orders per shopping session × average order value
Before launch, take the last three from your returning customers on the mobile website, since they’re the closest match to the people who’ll download your app. The coffee store’s returning mobile shoppers average three shopping sessions a month, place an order in 5% of those sessions, and spend $60 per order:
| Month 1 | Month 2 | Month 3 | |
|---|---|---|---|
| Active shoppers | 600 | 800 | 860 |
| Shopping sessions (× 3) | 1,800 | 2,400 | 2,580 |
| Orders (× 5%) | 90 | 120 | 129 |
| App sales (× $60) | $5,400 | $7,200 | $7,740 |
That’s $20,340 in app sales over the first three months. Starting with mobile web figures keeps the model conservative. If the app converts better than your mobile site, you’ll see it in the actuals and can raise the rate for later months.
Measure order value the same way your store reports sales. Shopify, for example, defines net sales as “gross sales - discounts - sales reversals.” If your $60 comes from net sales, compare the forecast with net sales from app orders, not gross.
The $20,340 is the value of orders placed in your app. Some of those orders would have happened on your website anyway, so it isn’t the same as the sales the app adds.
Working out the app’s incremental revenue means comparing these shoppers’ spending across every channel with what they’d have spent without the app.
What Changes the Forecast Most?
The two biggest uncertainties are reach and repeat use, so build a conservative, an expected, and a strong case for each. For the coffee store, the conservative case has 20% fewer listing visits than planned and lower return rates. The strong case has 20% more visits and higher return rates:
| Conservative | Expected | Strong | |
|---|---|---|---|
| Listing visits | 20% fewer | As planned | 20% more |
| Return rate after one month | 40% | 50% | 60% |
| Return rate after two months | 25% | 35% | 50% |
| Active shoppers in Month 3 | 600 | 860 | 1,200 |
| App sales in Month 3 | $5,400 | $7,740 | $10,800 |
A change in reach moves every cohort up or down once. Return rates work differently, because they apply to every earlier cohort in every later month. In Month 3, they apply to two cohorts. In Month 12, they’ll apply to eleven. So the further ahead you forecast, the more your return rates decide the result.
Timing changes the shape of the forecast, too. If the app launches during a sale, the first cohort’s order rate may be higher than it will be later, so don’t carry launch-month rates into quiet months.
If customers reorder every six to eight weeks, fewer may return one month after they start than two months after. Set each rate month by month to match your buying cycle, rather than assuming a smooth decline.
The conservative case should still make the launch worth doing. If the plan only works in the strong case, it needs more promotion or a stronger reason to download, and it’s better to find that out before launch than after. For the costs side of that decision, see how to build an app ROI case.
How Do You Update the Forecast After Launch?
To update the forecast after launch, compare each step with its actual result, starting at the top. The first step that missed explains most of the gap, so fix that assumption rather than adjusting a later one to make the totals match.
Each step has its own source, and each source has its own definition:
| Forecast step | Where to find the actual |
|---|---|
| Listing visits | Unique clicks in your email and SMS reports, and banner taps in your site analytics |
| First-time downloads | First Time Downloads in App Store Connect, and New users in the Google Play Console |
| First shopping sessions and returns | A cohort report in your app analytics, built on your shopping event |
| Orders and sales | Your store’s sales report, filtered to app orders |
A few of those sources need care:
- Apple’s campaign data starts late. App Store Connect can tie listing views and downloads to a campaign link, but “you can’t generate a campaign link until your app has data.” Track launch sends with your own links, and add campaign links once they’re available.
- Google Play’s listing report counts install taps. Since July 2026, its store listing reports focus on “the moment a user expresses intent to install.” For completed first-time installs, use New users, “users who installed your app for the first time.”
- App Store Connect usage data is partial. Its sessions and active devices are “based on users who have opted in to share their analytics with you,” so count active shoppers in your own app analytics.
- Cohort reports count active users per period. In Google Analytics’ cohort exploration, the standard calculation includes everyone who meets the return criteria in a period, “regardless of what they do in other periods.” That matches the return rates used here.
Once you know which step missed, change it for the months still ahead and keep the actual figures for the months already past.
The coffee store’s Month 1 cohort will have three real months of returns by the end of Month 4, which is a much better guide to Months 5 and 6 than the original assumption. Each month, the forecast relies a little less on assumptions and a little more on your own shoppers.
Final Thoughts
A useful app forecast is a launch plan with numbers attached. Each download figure traces back to an email, a text, or a placement your team has agreed to run, and each sales figure traces back to shoppers who started in a particular month and came back.
A forecast like this is worth building before launch, because it tells you which promotions have to happen for the targets to hold, and which assumptions to check first.
Start with the listing conversion rates in the first few weeks and the first cohort’s return rate a month later, because those are the guesses with the least of your own data behind them.
Frequently Asked Questions
How Do You Forecast App Downloads With No App Data?
You can forecast app downloads before launch from data you already have: subscriber counts, returning mobile visitors, monthly orders, and click rates from comparable email and SMS campaigns. The listing conversion rates and the share of downloads that start shopping are the biggest guesses, so keep them modest and replace them after the first month.
Should You Forecast iOS and Android Separately?
Forecasting iOS and Android separately is worth it, because the two stores count listing visits and downloads differently and your customers may not split evenly. Use your website’s own mix of iPhone and Android visitors to divide listing visits, then apply a separate conversion rate to each.
How Long Should You Wait Before Trusting the First Cohorts?
The first cohorts become reliable once they’ve passed at least one full buying cycle, and ideally two. For a store whose customers reorder every few weeks, that’s two to three months. For products bought less often, it takes longer, and early return rates will understate repeat use.
What’s the Difference Between App Downloads and Active Users?
App downloads count installs, while active users count people who use the app in a given period. A download happens once, but a customer only adds to sales in the months they come back and shop. That’s why the forecast tracks first shopping sessions and returns, not downloads alone.
Is App Revenue the Same as Incremental Revenue?
App revenue is not the same as incremental revenue. App revenue counts every order placed in the app, including orders the customer would have placed on your website anyway. Incremental revenue counts only the extra spending the app creates, which takes a comparison of these customers’ total spending with and without the app.



