How do I measure success after launching my MVP?

You launch your MVP and the first thing everyone asks is: how did it go? Downloads arrive, a few reviews appear, your dashboard fills with numbers, and it becomes very easy to mistake all of that activity for proof that something is working. The honest answer is that most of what founders measure in the first few weeks tells them almost nothing useful about whether they have built something people genuinely want.

The question founders need to ask is: are you solving a real problem people will pay for, or are you in love with the solution?

The problem is knowing which data matters and what it is actually telling you. Download counts go up while the product quietly loses people after day three. Session length looks healthy while users are actually confused. Investor enthusiasm arrives before real retention has been established. These are patterns we see across the products we work on, and they explain why so many MVPs that look successful on paper end up stalling at the next stage.

Measuring success after an MVP launch means getting honest about what your numbers are actually describing. This piece works through what to track, what to ignore, and how to turn what you find into decisions that move the product forward.

Define What Success Means Before You Launch

One of the most common mistakes founders make is launching first and then deciding what success looks like. By the time the numbers arrive, they are already being read through the lens of whatever the founder hoped to see. That is confirmation-seeking.

Success for an MVP is specific and depends entirely on what question you were trying to answer. Were you testing whether users would complete onboarding? Whether they would return after day one? Whether they would pay at a particular price point? Each of those requires a different number to watch, and if you have not named the question in advance, you will end up cherry-picking whichever number looks best after the fact.

Write the Benchmark Before You Open

Before launch, write down a single primary question your MVP is designed to answer. Then write the number that would represent a clear yes. If you are watching day-one retention, decide in advance what threshold would concern you and what threshold would give you confidence to move forward. Simon's benchmark is explicit: anything below 50% retention after day one is a warning sign, and the goal is to push that figure as high as possible and then hold it through day thirty and beyond, particularly if you are spending on user acquisition and need that spend to be justified.

Secondary Signals to Name in Advance

Alongside the primary question, name two or three secondary signals, things like completion of a core action, repeat visits within the first week, or a specific number of unprompted shares. Having these written before launch gives you a frame to read your data against, rather than a blank page you fill in retrospectively.

The Metrics That Actually Matter for an MVP

Not all metrics carry equal weight at the MVP stage, and some of the most visible ones carry the least. The numbers that actually matter are the ones that tell you whether users found real value in the product, not whether they arrived or how long their session lasted.

The metrics worth tracking at MVP stage fall into a clear order of priority. Retention at day one, day three, day seven, and day thirty is the closest thing to a ground truth the product has. If users are not coming back, the product has not yet earned a place in their life, regardless of how many downloaded it. Alongside retention, track the completion rate of your core value action, the single thing your product exists to help users do. If most users never reach it, that is where the problem sits.

• Day-one, day-three, and day-seven retention rates
• Completion rate of the primary value action
• Return visit rate within the first two weeks
• Drop-off point in the onboarding flow

What does not belong at the top of your dashboard at this stage is download volume, total session time, daily active users, or monthly active users as standalone figures. These are useful context, but as Simon puts it, they are vanity metrics, numbers that can look impressive while the product is quietly failing the people who tried it. Amplitude's 2025 Product Benchmark Report, covering over 2,600 companies, found that 98% or more of users churn within two weeks if they have not experienced real value. That is the urgency behind tracking the right things early.

Why Download Numbers Lie

Downloads feel like proof. They arrive quickly, they are easy to read, and they tend to go up in the first few days after a launch as curiosity drives installs. The problem is that a download tells you only that someone found your product interesting enough to try. It says nothing about whether they found anything worth staying for.

We have seen this on products where download numbers were rising week on week, which created a real sense of momentum, while the retention numbers told a completely different story. Users were arriving and leaving within the first few days, sometimes within hours, without ever reaching the moment where the product becomes useful. The growth line went up. The product was not growing.

Seeing download numbers rising can mask the fact that most of those users are already gone by day three.

According to research cited by Andreessen Horowitz's Andrew Chen, most mobile apps lose 77% of their daily active users within three days of install. That figure sits alongside a median day-one retention rate of roughly 25% across mobile apps, according to Adjust. Downloads can compound those numbers into something that looks healthy on a chart while the underlying behaviour tells you the product is not holding people.

The honest read of your download number is this: it is the size of your experiment, not a result. Every person who downloaded is someone the product had a chance to retain. What happened next is the actual data.

How to Read Retention as a Signal of Product Health

Retention is the most direct signal you have that the product is doing what it is supposed to do. If users come back, the product gave them a reason to. If they do not, it did not, and the download count, the rating, and the session length all become secondary to that single fact.

Reading retention well means looking at it across multiple time points, not just the day-one number. Day one tells you whether onboarding worked. Day three tells you whether there was enough value in the first session to justify a second visit. Day seven tells you whether a habit is forming. Day thirty tells you whether the product has become part of someone's life. Each of these can drop sharply and the previous ones can still look reasonable, which is why looking at only one misses the picture.

The Stickiness Question

Session length is a number that deserves particular scepticism. A user spending a long time in the product sounds positive, but the question worth asking is whether that session length reflects genuine value or whether it reflects confusion, or a gamification mechanic that is keeping them present without giving them anything real. Those are very different situations and they look identical in the data.

What Good Retention Looks Like

A day-one retention rate below 50% is a warning sign worth taking seriously. the higher you can push that figure, the better, and the goal is to hold retention past day one and through to day thirty and beyond. Bain and Company research suggests that increasing retention by just 5% can drive up profits by as much as 95%, which gives a sense of how much leverage sits in this number compared to continued acquisition spend.

What User Behaviour Tells You That Users Never Will

Users are optimistic about themselves. When asked whether they would use an app, 60 to 80% typically say yes. Actual usage once installed often sits between 10 and 20%, according to AppTweak. This gap between what people say and what they do is one of the most reliable patterns in product research, and it is why behavioural data consistently tells a truer story than survey responses or user interviews alone.

What users do inside the product, where they stop, which actions they repeat, which screens they revisit, carries information they either cannot articulate or would not think to share. A user who churns on day three rarely leaves a review explaining why. A user who returns every day rarely explains what brought them back. The behaviour is the signal. The words are a partial account of it at best.

Map your drop-off points by looking at where users exit the product, not just where they spend time. The exit point is usually where the friction lives, and that is where a small change has the most leverage.

This is why tracking at the flow level matters as much as tracking aggregate numbers. Which step in onboarding loses the most people? Which screen do users visit once and never return to? Which feature drives the highest return rate? These questions have answers in your event data, and those answers are more reliable than asking users to explain their own behaviour after the fact.

How Small UX Changes Can Shift Your Numbers Significantly

The changes that move retention and completion rates are rarely the large ones. They tend to be small adjustments to how the product communicates with users, particularly at the moments of highest friction, which are usually the beginning of a flow, a decision point, or a moment where the user does not know how much further they have to go.

On a fitness product we worked on, one of the most significant shifts came from a single change to how the onboarding flow opened. Users were being taken through a series of orientation questions, and a significant proportion were abandoning before they finished. The change was to tell users upfront how long the process would take, nothing technical, just a clear expectation set before they began. Before that change, drop-off rates through the flow were around 80 to 85%, typically happening within the first three or four questions. After introducing upfront priming, completion rates rose to approximately 95%.

Before investing in new features to improve retention, audit the existing onboarding flow for moments where users lack context. Setting expectations before a process starts consistently outperforms adding progress indicators inside the flow.

That result matters because it shows where the leverage often sits. The problem was not the questions themselves. It was the uncertainty around what was being asked of the user before they committed to answering them. Telling people what to expect, and when, put them in a more prepared state and that changed their behaviour significantly. Small, clear, and grounded in how people actually process new experiences.

Turning Data Into Documented Product Decisions

Gathering data and acting on it are two different things, and the gap between them is where most MVPs lose their momentum. You can have excellent retention tracking, clear behavioural data, and a well-mapped drop-off funnel, and still not translate any of it into a product change if there is no process for doing so.

Research by Nielsen Norman Group suggests that in lower-maturity organisations, including startups and smaller companies, fewer than 20% of research findings result in a documented product change. Even in more established organisations, that figure does not reach 50%. The data is not the problem. The translation from insight to decision is where things stall.

Make the Decision Visible

Every change to the product that comes from data should be written down before it is built. Not a long document, but a short record of what the data showed, what decision it led to, and what you expect to change as a result. This does this two things. It forces the team to be specific about what they are testing and why, and it gives you something to compare against once the change is live.

Close the Loop

After the change goes live, return to the same metric that prompted it. Did the drop-off rate improve? Did retention shift? If it did, you have a decision you can trust and replicate. If it did not, you have a finding worth recording. Either way, the loop is closed and the next decision starts from a better place than the last one.

Keep a simple decision log, one row per product change, with the metric that triggered it, the change made, and the outcome measured. This becomes your single most useful document when presenting to investors or planning the next build phase.

When Investor Confidence Becomes a Proxy for Product Readiness

Early-stage funding is a real signal of something, but it is not a signal that your product retention is healthy or that your users are finding genuine value. Investor confidence reflects a belief in the opportunity, the team, and the potential. It does not reflect what is happening at day three or day seven for the people using the product right now.

Between funding rounds, our team made an explicit decision to put in much better tracking of user retention and to start asking users proactively how things were going. The reason was that investor confidence had been functioning as a kind of reassurance that the product was on track, when the actual retention health of the product had not been properly understood. The funding said the opportunity was real. The retention data needed to say the product was working. Those are different things and they require different evidence.

This is a pattern worth naming clearly because it affects decision-making. When investors are interested and conversations are progressing, it becomes easy to deprioritise the harder work of understanding whether users are staying. The numbers that impress in a pitch deck, downloads, DAUs, session length, are often not the same numbers that tell you whether the product is ready for the next stage. Keeping those two sets of questions separate is one of the more important habits a founding team can build.

How to Know Whether Your MVP Has Worked

An MVP has worked when it answers the question it was designed to answer, clearly and with enough evidence to make a confident next decision. That is it. The question is not whether the product is good yet, whether users love every part of it, or whether the metrics match what a scaled product would produce. It is whether you know more now than you did before launch, and whether that knowledge points somewhere specific.

The table below maps the core questions an MVP is typically trying to answer against the signals that give you a usable answer.

The question Simon frames as the most useful one for any founder at this stage is direct: are you solving a real problem that people will pay for, or are you in love with the solution? That distinction is what separates a product that has earned its next phase of investment from one that has been built primarily to satisfy the founder's own vision. The MVP data is what makes the answer visible.

Conclusion

Measuring success after an MVP launch is not complicated in principle. It gets complicated because the numbers that are easiest to see, downloads, session length, star ratings, are often the least useful ones, and because the numbers that actually matter, retention at day three, day seven, day thirty, completion of the core value action, are the ones that require more honesty to look at.

The pattern we come back to on every product we work on is this: build a specific question before you launch, track the metric that answers it, and then close the loop by documenting what the data led you to change. That sequence, question, measurement, decision, record, is what separates founders who learn from their MVP from those who collect data without ever quite trusting it enough to act on it.

Consumer apps are a long-term investment. Retention compounds, as does the understanding you build by watching how real users behave inside a real product. The MVP is the point where the evidence starts to be worth something, if you know what to do with it.

If you are trying to make sense of your MVP data and work out what it means for the next stage of the product, let's talk about your product.

Frequently Asked Questions