Driving installs after launch: the first 90 days

Launch day feels like the finish line, and then nothing happens. The listing is live, the integration works, you passed review, and the install counter reads a number so small you keep refreshing to see if it moved. This is the moment most marketplace efforts quietly stall, not because the app is bad, but because launching and growing are two different jobs and almost nobody plans the second one. Getting listed put your app on the shelf. The first 90 days decide whether anyone takes it off.

This is an independent playbook for those first 90 days, whatever marketplace you launched on. The specific mechanics differ across platforms, and each has its own review rules, listing fields, and analytics, so treat the platform details as things to confirm in that marketplace's current documentation. What does not change is the shape of the early work: seed a first burst of real installs from people who wanted the app, get those installs to activate before they churn, earn the first honest reviews, and use all of that to give the flywheel its first turn. A launch that skips this window ends up with a listing that technically exists and practically does not.

It pairs with our guide to how marketplaces rank apps, our guide to writing a marketplace listing that converts, and our SaaS marketplace strategy guide, so you can place the first 90 days inside the wider work of getting found, getting installed, and turning installs into revenue.

The 60-second version

• Launching and growing are different jobs. Passing review earns you a shelf, not installs. The first 90 days is a deliberate campaign, not a wait.
• Seed installs from people who already want the app. Your warmest installs are the customers who asked for the integration. Reach them by name the week you launch, not with a broadcast.
• Activation matters more than the install count. An install that never reaches the promised outcome churns and drags your signals down. Get the first customer to a visible result fast.
• Earn the first reviews on purpose. Satisfied customers rarely review unprompted while frustrated ones always do. Ask activated users at the right moment, and never buy or reward reviews.
• Concentrate the first burst. The same installs read as a flat line when spread over months and as momentum when clustered into a launch window.
• Give the flywheel its first turn. Installs feed reviews and usage, those feed ranking, ranking feeds more installs. The first 90 days is about manufacturing that first rotation honestly.
• Plan the window in three acts. Days 1 to 30 seed and activate, days 31 to 60 build the review and usage habit, days 61 to 90 amplify and measure what worked.

Why the first 90 days decide the outcome

A marketplace listing with zero installs and zero reviews is a cold start, and a cold start has nothing to compound. The signals marketplaces reward, install velocity, ratings, active usage, all begin at zero, so the listing sits wherever a brand-new app sits, which is usually far enough down the results that no organic shopper ever sees it. The uncomfortable part is that this state is stable. A listing that gets no help from you stays cold, because there is no traffic to convert into installs, no installs to convert into reviews, and no reviews to lift the rank that would generate traffic. Left alone, a launch does not slowly build. It sits.

The first 90 days is the deliberate campaign that breaks that stall. It is a finite window where you do the manual, unscalable work of manufacturing the first turn of the flywheel by hand, because the flywheel cannot turn itself yet. You reach specific customers, you watch specific installs activate, you ask specific people for reviews. None of it scales, and that is the point. You are not building a repeatable growth machine in the first 90 days. You are getting the machine to move once, so that it has enough momentum to keep moving with less of your hand on it later.

Framing the window this way changes how you spend it. If you think the launch was the work, you spend the 90 days waiting and checking the counter. If you think the 90 days is the work, you spend it on a short list of high-leverage moves: seeding, activating, and reviewing, in that order, then amplifying what worked. The rest of this playbook is that short list, broken into the three acts that fit a 90-day window.

Days 1 to 30: seed real installs and get them to activate

The first month has two jobs that run in parallel: put real installs on the board, and make sure those installs reach the outcome you promised. Skipping the second job is the most common way a strong launch turns into a weak one, because installs that never activate churn, and churn is a signal marketplaces read against you.

Seed installs from the people who already want the app. Your warmest installs are not strangers browsing the category. They are the customers who asked you to build the integration, the ones who mentioned they use both products, the prospects who said they would sign if you connected to their CRM. Reach these people individually the week you launch, by name, with a short message that says the thing they asked for now exists and here is the link. A broadcast email to your whole list is weaker than fifty personal messages to people who specifically wanted this, because the personal message converts and the broadcast mostly gets ignored. This concentrated first burst is also what creates install velocity, the momentum signal that a slow trickle never produces.

Then get each install to activate. An install is not an outcome. The riskiest moment in the whole 90 days is right after someone connects the app, because a customer who installs, hits a confusing setup, and cannot find the next step will uninstall within the week. Activation means reaching the result you promised in the listing's first line: the alert firing in the channel, the records syncing, the dashboard filling in. Design the first run so that result arrives in minutes, not after a support ticket. Nielsen Norman Group's research on onboarding tutorials describes the trap here well: onboarding that explains the interface instead of getting the user to their first success tends to be skipped and forgotten, so the first run should drive toward a visible result rather than a tour.

A useful way to hold both jobs at once is a short 30-day scoreboard:

The point of the scoreboard is not vanity. If installs are climbing but activation is low, the problem is your first run, not your seeding, and pouring more installs onto a broken first run just manufactures churn faster. Fix activation before you widen the funnel.

Days 31 to 60: build the review and usage habit

By the second month you have a base of activated customers. The job now is to turn that base into two durable signals, reviews and retained usage, and to do it as a habit rather than a one-time push, because both signals decay if you stop feeding them.

Earn the first reviews on purpose. Reviews do not happen on their own. A satisfied customer rarely thinks to leave one, while a frustrated customer almost always does, so a listing left to itself drifts toward a rating worse than the product deserves. The fix is to ask, at the right moment, the right people. After a customer has used the integration successfully for a couple of weeks, prompt them, with a short in-app message or a one-line email tied to a moment when the app just worked. Ask only your activated, satisfied users, because a review request sent to someone who never got the app working invites the bad review you were trying to avoid. And keep it clean: never buy reviews, never reward a positive one specifically, and never post as a customer yourself. Marketplaces are good at detecting all three, and the penalty for getting caught costs far more than the reviews were worth.

Protect and deepen usage. Activation gets a customer to the first result once. Retention is whether they keep getting value week after week, and it is the signal marketplaces trust most because it is the hardest to fake. Watch for the customers who activated but then went quiet, and reach them before they uninstall. A short check-in, a tip about a feature they have not used, or a fix for the thing that stalled them keeps a wavering customer in the active column. This is ordinary customer retention work applied to a marketplace install: the cheapest install to keep is the one you already have, and a retained customer is worth more to your rank than a fresh install that churns.

A few habits make the second month compound instead of stalling:

• Ask every activated customer, not a chosen few. A neutral prompt to all satisfied users at a natural moment is both more effective and safely within marketplace rules.
• Answer every review, especially the critical ones. A thoughtful public reply to a two-star review tells the next shopper you are present and you fix things, which often matters more than the star itself.
• Close the loop on churn signals. When a customer goes quiet, treat it as a question to answer, not a number to accept. The reason they stalled is usually fixable and usually shared by others.
• Log what you ship. A visible changelog and version notes tell both shoppers and the marketplace that the app is alive, which supports the freshness signal that quietly feeds rank.

Days 61 to 90: amplify and turn the wheel

The third month is where the manual work you did starts to compound, and where you add the amplifier. By now you have activated installs, a handful of honest reviews, and retained usage. That is a listing that actually converts, and only now is it worth pointing extra traffic at it.

Amplify with co-marketing, on a listing that already works. A joint launch, a partner newsletter mention, a webinar, or a co-authored post sends a burst of qualified traffic to the listing, which converts into a cluster of installs and reviews, which spins the wheel faster. The order matters. Co-marketing amplifies a listing that already converts; point a launch at a listing with a weak first line and a broken first run and you spend partner goodwill driving traffic that bounces. That is why this belongs in month three, not month one. The influenced versus sourced pipeline distinction is worth keeping in mind here, because a co-marketing burst mostly influences buyers who were already close, and you want to be able to see that effect rather than guess at it.

Measure what actually worked. Ninety days in, you have enough data to tell which of your moves moved the needle, and that answer shapes the next quarter. Which seeding channel produced installs that activated. Which review prompt got responses. Which first-run fix cut early uninstalls. The point is to stop doing the manual work that did not pay off and to systematize the parts that did, so month four is less hand-cranking and more of a loop that turns with lighter effort.

The three acts and the one turn of the wheel look like this:

Underneath all three acts is one loop. Installs feed reviews and usage, reviews and usage feed ranking, ranking feeds more installs, and each turn makes the next install cheaper because the rank is doing the finding for you. The first 90 days is the work of getting that loop to complete its first full rotation, by hand, so that it has the momentum to keep rotating after your attention moves on. The mechanics of the loop itself are covered in how marketplaces rank apps; the first 90 days is where you start it.

Common mistakes, and the fix

Treating launch as the finish line. The fix: treat the first 90 days as the actual campaign. Passing review earns a shelf, not installs, and a listing that gets no help from you after launch stays cold, because a cold start has nothing to compound.

Broadcasting instead of seeding. The fix: reach the specific customers who asked for the integration, individually, by name, the week you launch. Fifty personal messages to people who wanted the app convert far better than one broadcast to a whole list, and the concentrated burst is what creates the velocity signal.

Counting installs and ignoring activation. The fix: watch activation rate and time to first result, not just the install counter. An install that never reaches the promised outcome churns and drags your signals down, so fix the first run before you widen the funnel.

Waiting for reviews to appear. The fix: ask activated, satisfied customers at a natural moment, with a neutral prompt to all of them. Satisfied users rarely review unprompted while frustrated ones always do, so a passive listing drifts to a worse rating than it deserves. Never buy or reward reviews.

Amplifying too early. The fix: run co-marketing in month three, once the listing converts and the first run works. A launch pointed at a weak listing spends partner goodwill on traffic that bounces, and buries the conversion problem under more visits.

Not measuring what worked. The fix: at 90 days, identify which seeding channel, which review prompt, and which first-run fix actually paid off, then systematize those and drop the rest. The manual work of the first quarter is meant to teach you the loop, not to be repeated forever.

FAQ

How many installs should I expect in the first 90 days? There is no universal number, because it depends on your category, your existing customer base, and how concentrated your launch is, so treat any specific figure as illustrative rather than a promise. The more useful target than a raw install count is a high activation rate on the installs you do get, because activated installs feed reviews, usage, and rank while unactivated ones churn and work against you. A smaller number of activated, retained installs beats a larger number that never reached the promised outcome.

Where do my first installs come from? Your warmest installs are the customers who already asked for the integration or already use both products, so reach them individually the week you launch rather than sending one broadcast. Those people convert because they wanted the app, they become your first reviewers because they got value, and they create install velocity because they arrive in a cluster. Cold shoppers browsing the category come later, once reviews and rank make the listing visible to them.

What is activation and why does it matter more than the install count? Activation is the moment an installed customer reaches the outcome you promised in the listing, the alert firing, the records syncing, the dashboard filling in, rather than just connecting the app. It matters more than the raw install count because an install that never activates tends to churn, and churn hits the retention and usage signals marketplaces trust most. A converting listing with a broken first run just manufactures churn faster, so activation is the thing to fix before you drive more installs.

When and how should I ask for reviews? Ask after a customer has used the integration successfully for a couple of weeks, with a short in-app prompt or a one-line email tied to a moment when the app just worked, and send it to all your activated, satisfied users rather than a hand-picked few. Never buy reviews, never reward a positive one specifically, and never post as a customer, because marketplaces detect all three and the penalty is severe. Asking is expected and within the rules; incentivizing or faking is not.

Should I do co-marketing during launch week? Usually no. Co-marketing is an amplifier that works on a listing that already converts, so it belongs in month three, after the first run is solid and you have some reviews. A joint launch pointed at a weak listing spends partner goodwill driving traffic that bounces off a poor first line or a confusing setup. Get seeding, activation, and reviews working first, then use co-marketing to spin a wheel that is already turning.

How does the first 90 days connect to ranking and revenue? The first 90 days manufactures the first turn of the flywheel by hand, and ranking is what that flywheel produces: installs feed reviews and usage, those feed rank, and rank feeds more installs, as covered in how marketplaces rank apps. Revenue comes after, when those installs are tracked to opportunities and expansion. The first 90 days sits between a listing that converts and a rank that compounds, and its whole job is to get the loop moving once.

Further reading

• How marketplaces rank apps for the flywheel your first 90 days is meant to start.
• Writing a marketplace listing that converts for the conversion surface every seeded install lands on.
• SaaS marketplace strategy for where the first 90 days sits in the wider decision to build, list, and grow.
• Influenced versus sourced pipeline for reading the effect of a co-marketing burst honestly.
• Nielsen Norman Group on onboarding tutorials for why a first run should drive to a result rather than explain the interface.
• Nielsen Norman Group on conversion rates for how small friction in the first steps quietly costs you activations.
• Conversion funnel on Wikipedia for the funnel model that install-to-activation follows.
• Customer retention on Wikipedia for why keeping an activated install beats chasing a fresh one that churns.

The short version

Launching and growing are different jobs, and the first 90 days is the second one. Passing review earns a shelf, not installs, and a cold listing left alone stays cold because it has nothing to compound. So the first quarter is a deliberate, mostly manual campaign to turn the flywheel once by hand. In the first month, seed real installs from the customers who already wanted the integration by reaching them individually, and get each install to activate by making the first run reach the promised outcome in minutes. In the second month, build a habit of asking activated, satisfied customers for honest reviews and of protecting usage before it churns, since both signals decay if you stop feeding them. In the third month, amplify with co-marketing on a listing that already converts, and measure which moves actually worked so you can systematize them. Underneath all of it is one loop, installs to reviews and usage to ranking to more installs, and the whole job of the first 90 days is to get that loop to complete its first full turn.

If you want help planning the first 90 days, from seeding installs to designing the first run to earning the reviews that start the flywheel, a Partner Audit reviews your product, your listing, and your marketplace potential, then hands you a concrete plan for what to do in each of those three windows.