What should I spend on getting new app users?

Spend enough time watching app budgets get allocated and a pattern emerges. A founder or product lead decides they need more users, so they increase acquisition spend, download numbers rise, and then six weeks later the numbers look roughly the same as before. The new users came in and left almost as quickly as the money went out. The budget did not solve the problem. It revealed one.

Download numbers rising while retention quietly falls is one of the most expensive illusions in mobile.

The question of what to spend on acquiring new app users is genuinely worth answering carefully, because the number you arrive at is not fixed. It depends on what happens after someone downloads your app, and most products are losing that battle quietly before they ever think to ask.

This piece works through the full picture: what retention actually costs you, how to estimate what a user is worth before you spend to find them, and how your onboarding shapes every acquisition decision that follows. Getting this right does not require a larger budget. It usually requires a clearer-eyed view of what the budget is actually buying.

Why downloads are not the metric you think they are

Downloads feel like growth. The number goes up, the chart looks healthy, and it is easy to keep spending in the direction the chart is pointing. But downloads measure interest, not commitment. They record the moment someone thought your app might be worth trying, which is a very different thing from whether it actually was.

We have seen this pattern play out directly. Teams watch their download numbers grow and treat that as validation, while quietly not looking at retention on day three, day five, or day seven. The download graph and the retention graph are often moving in opposite directions, and the download graph is the one getting shown in meetings.

The deeper problem is that downloads have a clean cost attached to them. You can see exactly what you paid per install. Retention has no obvious line item, so it does not get the same scrutiny. But the cost of poor retention is real and accumulates fast. According to Andrew Chen, around 77% of apps lose their daily active users within three days of install. If that describes your product, you are refilling a bucket with a significant hole in it each time you run an acquisition campaign.

The shift worth making is treating downloads as a leading indicator of nothing in particular, and treating day-seven retention as the number that tells you whether acquisition spend is justified at all.

What retention actually looks like in mobile apps

Retention is a shape. The pattern is familiar: a steep drop on day one, a further fall by day three, and then a slower decline that continues for weeks. What matters is where that curve levels off, because the floor of your retention curve is essentially your real audience size.

A day-one retention rate below 50% is a warning sign worth taking seriously. The goal is to push retention as high as possible through to day thirty and beyond, particularly when you are paying to bring users in. On a water tracking app we developed, we observed a drop-off pattern similar to what the top personal finance apps experience: losing around 71% of daily active users between day one and day thirty. The silence from departing users felt like approval at first. The retention data said otherwise.

Even products with strong onboarding typically see a 40 to 50% retention drop after three days. That range is genuinely different from the average 77%, and the gap between those two figures represents the financial value of getting your first few days right. A product at 50% three-day retention needs roughly half as many downloads as a product at 25% to reach the same active user base.

Track retention at day one, day three, day seven, and day thirty as a dashboard habit, not just when something feels wrong. The pattern across those four points tells you far more than any single figure does.

Retention also interacts with quality in ways that show up in the numbers over time. Products that treat users well retain them better: ethical products tend to see around 23% higher retention rates than those using manipulative engagement tactics, with higher brand affinity as a longer-term benefit too.

The silent churn problem: when users leave without telling you

Around 25% of mobile apps are used once and never opened again, according to Statista. Those users do not complain. They do not cancel. They do not submit a support ticket explaining what went wrong. They simply stop. And in the absence of any signal, it is easy to read their silence as satisfaction.

With mobile apps, the assumption that no news is good news tends to work against you. Users abandon apps without submitting complaints, and their silence is not approval. It often masks a poor experience or a product that has stopped feeling relevant. Without proactively tracking retention, teams can mistake the quiet for calm.

Users who leave never tell you why, the only way to hear them is through retention data.

The session length and daily active user figures that get reviewed most often in product meetings do not help here. Those metrics tell you that a user was present, but nothing about why. They cannot distinguish between someone who is genuinely getting value and someone who is confused or being held in place by a notification loop. Relying on them to diagnose churn is like checking that a seat is occupied without asking whether the person in it is happy to be there.

Identifying silent churn requires looking at behaviour, not just presence. Users who open the app but navigate nowhere useful, who log one session and disappear, or who engage only when prompted by a push notification are candidates for churn who have not yet uninstalled.

Set up cohort analysis by acquisition channel so you can see whether users from different sources churn at different rates. A channel that looks cheap per install often turns out expensive per retained user once you track it properly.

What lifetime value is and how to estimate it before you spend

Lifetime value (LTV) is the total revenue a user generates over the full time they stay active in your app. It is the figure your acquisition cost needs to sit comfortably below, and estimating it before you spend is the only way to know whether a given channel is worth using at all.

A useful target is a customer acquisition cost to LTV ratio of 1:3, meaning for every pound spent acquiring a user, that user should return three pounds over their lifetime, according to Userpilot. That ratio gives you room to cover operational costs and still run a sustainable product. A ratio closer to 1:1 means you are breaking even at best, and any retention problem immediately tips it into loss.

How to build a rough LTV estimate

You do not need a year of data to start. Take your average revenue per active user per month, multiply it by your average number of active months before churn, and you have a working estimate. If you are pre-revenue, use the pricing you intend to charge and apply a conservative retention assumption: assume most users will churn by month three, because most do.

Consumer app revenue compounds over time through volume and retention, not through pricing that reflects production costs. Expecting to recover development spend quickly through subscription fees usually sets prices too high, which in turn drives churn, which collapses the LTV the pricing was meant to protect.

How your onboarding shapes every acquisition cost that follows

Every user you pay to acquire arrives at your onboarding. What happens there determines whether the acquisition cost was worth paying. A poor onboarding does not just lose one user. It raises your effective cost per retained user across every campaign you run, because you are paying full acquisition price for users who will not stay.

Abandonment in the first sixty to one hundred and twenty seconds is almost always an onboarding problem: too many screens, forced registration before any value has been shown, permissions requests without explanation. Forced early registration alone can cause 15 to 20% of users to drop off before they have seen what they signed up to see.

Users who experience friction in their first session are 2.7 times less likely to return by day seven, according to AppsFlyer. That multiplier is not recoverable through re-engagement campaigns. The damage is done in the first interaction, and paying to bring someone back after a bad first session is more expensive and less reliable than getting the first session right.

What good early onboarding does

The goal is to get a user to a moment of genuine value as fast as possible, with as little friction as standing between them and it. That means showing the product working before asking for anything, requesting permissions only when the context makes the reason obvious, and keeping the path from download to first useful action short. Products that do this well do not need to spend as much on reactivation, because they are not generating the same volume of lapsed users in the first place.

Map the steps between download and first meaningful action and count them. If the number is above five before any value has been delivered, the onboarding is contributing to churn and inflating your effective acquisition cost.

Designing acquisition loops that compound in your favour

Paid acquisition is a linear equation: spend a pound, get a user. Stop spending, stop growing. Loops work differently. They convert existing users into a source of new users, which means the growth you have already paid for starts doing some of the work for the growth you need next.

On a travel booking product we worked on, aimed at younger adults booking group trips, we tried the standard acquisition toolkit: referral discounts, push notifications to re-engage lapsed users, social sharing prompts, email campaigns. What actually moved the numbers was a loop built into the booking flow itself. When someone organised a group trip, the app prompted each individual traveller to download the app to handle communication and submit passport details. One person booking a trip for ten people generated nine new users automatically, each of whom could then trigger the same loop for their own trips.

That loop worked because it was structural, not incentive-based. It did not depend on users wanting to recommend the app. It created a natural reason for new people to join that was directly tied to getting value from the product. The booking could not happen effectively without it.

Referral schemes and social sharing can layer on top of a loop like this, but they rarely replace it. A loop that requires no active effort from the user compounds reliably. One that depends on goodwill is subject to how the user is feeling that day.

When retention drops, more acquisition spend makes it worse

There is a tempting response when growth stalls: spend more to bring more users in and hope the numbers average out. This tends to accelerate the problem rather than solve it, because a product with poor retention needs to understand why the users it already has are leaving.

On an art-based auction game we worked on, with real money prizes, we saw this play out in detail. Initial retention sat in the mid-sixties to low seventies. As the volume of available games declined, usage became sporadic and retention fell to around 30 to 35%. The right response was to address the content supply problem. Instead, the client pushed for new features. Retention continued to fall. It was only when the product refocused on the core user journey, specifically making sure enough games were available, that retention climbed back to the mid-sixties to low seventies. But rebuilding the trust that had eroded during that period took considerably longer than fixing the product did.

Pouring acquisition spend into a product during a retention drop means paying to acquire users who are statistically more likely to churn than the users you already lost. Each of those new users enters the product in its current state, which is the state that was not retaining people. The spend compounds the problem, and the retention data gets noisier because the new cohorts mask where the original drop came from.

1. Identify when in the user journey churn is happening (day one, day three, or later)
2. Diagnose the cause before adjusting any acquisition budget
3. Fix the retention problem first, then scale spend against an improved baseline

What a sustainable acquisition budget actually looks like

A sustainable acquisition budget is one derived from your retention and LTV figures, not from what a competitor is spending or what a channel salesperson has told you is the going rate. The budget is the output of the model, not the starting assumption.

Start with your current day-thirty retention rate. If it is below 30%, your acquisition spend is working against a significant structural problem and should be limited until that figure improves. If it is above 50%, you have a product that is holding users, and the economics of acquisition start to look reasonable.

A working budget framework

Your maximum sustainable cost per acquired user is roughly your LTV divided by three, using the 1:3 ratio as a guide. From there, the question is which channels can deliver users at or below that cost, consistently and at scale. Channels that are cheap per install but deliver users who churn in three days are never cheap per retained user, and that is the figure that actually matters.

Acquisition spend should also grow proportionally to retention improvement, not ahead of it. A product with rising retention can justify increasing acquisition spend because each pound is buying a better asset than it was before. A product with flat or declining retention has no basis to scale spend, regardless of how healthy the download numbers look.

Building the product as a long-term asset, rather than trying to recover production costs through early revenue, changes how these decisions feel. The brands that sustain growth in lifestyle categories are the ones that compound retention over time and treat acquisition as something that rewards a working product, rather than something that creates one.

Conclusion

The question of what to spend on new app users does not have a fixed answer, but it does have a clear method. You start with how well your product retains the users it already has, work out what a retained user is actually worth, and set your acquisition ceiling from there. Any other starting point produces a number that looks like a budget but functions as a guess.

The products that spend well on acquisition tend to be the ones that have done the harder prior work: onboarding that gets users to value quickly, a retention curve that levels off above the category average, and growth loops that reduce dependence on paid channels over time. When those pieces are in place, acquisition spend produces compounding returns rather than a leaky cycle of spend and churn.

The art-based auction game, the group travel product, the water tracking app we built, each of them taught a version of the same thing. Retention is the condition under which acquisition makes sense at all. Get the retention right, and the acquisition question becomes much easier to answer with confidence.

If you are working through these questions for your own product and want a second pair of eyes on the numbers, let's talk about your app's retention and acquisition strategy.

Frequently Asked Questions