A growing signup count can hide a shrinking audience. You keep attracting people, yet each month feels like starting over.
Before you spend more on acquisition, find out what happens to the users you’ve already won. Otherwise, the next campaign may leave you paying to replace people who never found a reason to return.
What is user retention?
User retention measures whether people continue using a product or service over time. A retention rate expresses the percentage of a defined group that returns and completes a chosen action during a specified period.
That action should reflect the value people get from your product. For an invoicing app, sending an invoice is usually more informative than opening the app. A login can show activity without showing that the user accomplished anything.
User retention vs. customer retention
User retention tracks people using the product. Customer retention tracks customers or accounts that continue the commercial relationship. Revenue retention tracks how much recurring revenue a starting customer group keeps generating.
These measures can move differently. A company might renew a software subscription while fewer employees use it. Another might retain its customers but lose revenue when they downgrade.
Even when the user and the payer are the same person, using the product and continuing to pay are different behaviors. Choose the measure that answers your business question.
Why user retention deserves attention
When users leave quickly, acquisition has to replace them before it can produce sustained business growth. Retaining paying customers gives you more opportunities to recover acquisition costs through repeat purchases or subscription payments.
That doesn’t make every retention tactic profitable. Discounts, support, and ongoing service all have costs. Judge a retention improvement alongside revenue and margin, especially if you’re paying to encourage users to stay.
Retention analysis also helps you decide where to invest. If users from one campaign keep returning while another campaign produces mostly one-time visitors, signup volume alone won’t tell you which performs better.
How to measure user retention
Agree on the definition before opening a dashboard. Otherwise, two people can report different retention rates and both be calculating their chosen metric correctly.
Define the group, action, and return window
A usable retention report needs four decisions.
Who you’re tracking: A cohort is a group with a shared starting condition, such as first use during the same week.
What counts as returning: Choose a meaningful action, such as completing another order or sending an invoice.
When a return counts: Specify an exact day, a fixed interval, or a return on or after a chosen day.
How users are identified: Use a consistent identifier and exclude test accounts. A person switching devices shouldn’t become two people in a report intended to measure individuals.
Match the interval to the task. A monthly reporting product might need a monthly measure; forcing a daily target could reward unnecessary activity.
Calculate retention from the original group
The basic cohort formula is:
User retention rate = (users from the original cohort who meet the return rule ÷ users in that cohort) × 100
Use a cohort whose full return window has elapsed. Suppose 1,000 users completed your chosen action in April, and 700 of those same users do it again in May.
Retention rate = (700 ÷ 1,000) × 100 = 70%
Another 200 new users complete that action in May. With no other users qualifying, that makes 900 active users for this report. The newcomers don’t belong in the April cohort’s retention calculation.
Count each qualifying person once for the interval. Someone who performs the action ten times is one retained user, not ten.
Check the limits of the customer retention formula
A common customer-retention formula is:
Retention rate = ((customers at the end − new customers during the period) ÷ customers at the start) × 100
For a subscription business, suppose there are 1,000 customers at the start and 900 at the end, including 200 newcomers. The calculation is ((900 − 200) ÷ 1,000) × 100 = 70%.
That shortcut only works when subtracting new customers leaves exactly the surviving members of the starting group.
If new customers join and leave before the period ends, subtracting every new signup understates retention. Reactivated customers who weren’t in the starting group can distort the result too.
For product usage, track the original users directly. Don’t assume that subtracting new signups from a monthly active-user total produces a valid retention rate.
Keep retention and churn definitions aligned
User churn measures the share of a starting group that meets your definition of having stopped using the product. Subscription churn usually concerns cancellations or non-renewals; a non-subscription product needs an explicit inactivity rule.
User churn rate = (users from the starting group who meet the churn rule ÷ users in that starting group) × 100
In the April-to-May example, 300 of the original 1,000 users didn’t return. If your rule classifies those users as churned for that month, churn is (300 ÷ 1,000) × 100 = 30%.
Retention and churn add to 100% only when they divide the same group into complementary outcomes using the same period and definitions. Missing a particular day doesn’t automatically mean a user has permanently left.
Choose the right retention method
Returning on a specific day is different from returning on that day or later. Those methods answer different questions, so a percentage without its method can mislead.
Exact-day retention measures whether users perform the return action on a particular day after their starting event. Day 7 retention counts qualifying activity on day 7. A user who returns on day 8 but skips day 7 doesn’t count for day 7.
Interval retention measures whether users return at least once during a defined window, such as days 7 through 13. This can suit products where people return weekly without following an exact seven-day schedule.
On-or-after retention measures whether users return on a specified day or any later day within the observation period. A user who returns on day 10 qualifies for day 7 on-or-after retention. Specify the observation cutoff because later returns can change the result.
Also check whether your tool uses calendar dates or elapsed 24-hour windows, and keep the timezone consistent. Compare results only after users have had enough time to reach the interval you’re measuring.
Use cohort analysis to see whether retention is improving
Compare signup cohorts at the same age to see whether newer users return more often. Looking at each group separately prevents a growing influx of signups from hiding weak retention among earlier users.
Here’s a hypothetical example. Each row contains a different signup group. Week 1 covers days 7–13 after signup, week 2 covers days 14–20, and week 4 covers days 28–34.
Signup group
Users
Week 1
Week 2
Week 4
Earlier group
200
45%
38%
30%
Later group
250
50%
44%
36%
Newest group
300
52%
46%
Not yet available
Each percentage is the share of the original signup group completing the return action in that window. The earlier group’s week 4 result means 60 of its 200 users returned during days 28–34. It doesn’t mean all 60 used the product every preceding week.
The later group has a stronger week 4 result, but that alone doesn’t prove an onboarding change caused it. Acquisition channels, customer needs, seasonality, and product changes may also differ between groups.
Compare equivalent groups and investigate the difference. Mark unfinished intervals as unavailable rather than zero. In this example, the newest group hasn’t completed its full week 4 observation window.
What is a good user retention rate?
There isn’t one useful target for every product. The return action, audience, business model, interval, and calculation method all affect the number.
Start with your own comparable cohorts. If a benchmark uses app opens on day 1, it won’t tell you whether your monthly invoice-sending retention is healthy. Look for improvement in the behavior that supports your product’s value and business economics.
Keep the underlying counts visible, especially with small groups. One additional returning user changes a ten-person cohort’s rate by ten percentage points.
Track supporting metrics without confusing them with retention
Retention tells you who came back. Other measures help explain their experience and its value to the business.
Activation rate tracks how many new users reach a meaningful early milestone. Define that milestone from your product’s value, then check whether people reaching it also return later. An association is a reason to investigate, not proof that forcing the action will improve retention.
Customer lifetime value (CLV or LTV) estimates the value of a customer relationship. Some models use revenue, while others account for margin or profit, so label yours clearly. Higher spending can raise LTV without improving the percentage of users who return.
Net Promoter Score (NPS) measures stated willingness to recommend. Calculate it by subtracting the percentage of detractors (scores of 0–6) from the percentage of promoters (scores of 9–10). Use the score to guide follow-up questions, then check respondents’ actual return behavior.
How to improve user retention
Choose an intervention based on where users stop getting value. A broad campaign to increase engagement won’t tell you whether the problem is setup, product reliability, poor fit, or an unnecessary step.
Help new users reach a useful result sooner
Onboarding should help someone accomplish the task they came for. Remove steps that can wait and make the next required action clear.
For an invoicing app, a worthwhile early result might be sending the first invoice. Requiring a full profile, team invitations, and a feature tour beforehand may delay that result. Check where users abandon the setup process, then test whether reducing that friction improves later retention.
Track both completion of the first useful action and subsequent returns. More people finishing a tutorial doesn’t necessarily mean more people will keep using the product.
Fix the obstacles users encounter repeatedly
Review support requests, failed actions, and user feedback alongside the retention report. Ask recent dropouts what they were trying to accomplish and what prevented them from continuing. Include people who left, rather than relying entirely on feedback from users who already like the product.
Prioritize problems that block the core task. After a fix, check whether the affected group can complete that task and whether its return rate improves.
Give people a relevant reason to return
Time reminders around the user’s task, unfinished work, or expected need. Let people control communication frequency and measure the return action that follows the message.
Use personalization when it reduces effort, such as restoring a saved workflow or surfacing a relevant next step. More messages and more customization aren’t useful goals by themselves.
If you test a loyalty offer, compare incremental repeat activity with the cost of the reward. Also check what happens when the incentive ends. An increase in discounted purchases doesn’t establish that users will return without the discount.
Check whether acquisition attracts the right users
Break retention down by acquisition channel or intended use case. A source of cheap signups may perform poorly once you measure meaningful returns.
Review the promise made before signup. If advertising creates an expectation the product can’t meet, more onboarding instructions won’t resolve the mismatch. Compare retained users and the cost of acquiring them before reallocating spending.
Test changes against a stable definition
Choose one problem, one intervention, and one primary retention measure. Record the cohort, action, interval, and baseline before making the change.
Where practical, randomly assign eligible users to the existing experience or the proposed improvement. If you compare before-and-after cohorts instead, account for other changes and treat the result as suggestive rather than conclusive.
Wait until both groups have completed the measurement window. Check retention alongside relevant costs, revenue, and complaints so an apparent improvement doesn’t conceal a worse customer experience.
Set up a report you can use consistently
For a small user base, a spreadsheet can work if you have reliable user identifiers, starting dates, and return-event records. Calculate each person’s eligibility and qualifying return before summarizing the results.
For websites and apps, Google Analytics 4 provides a Retention overview and a configurable Cohort exploration. Google notes that Cohort exploration uses device data and doesn’t consider User-ID when creating cohorts. Account for that limitation if you’re trying to measure people across devices.
Product analytics tools can support event-based retention, while billing records are needed to establish whether subscriptions remain paid. Choose the data source that matches the question. A website visit doesn’t establish a subscription renewal.
Write down the report’s definition so another person can reproduce it. Include the starting condition, return event, time window, identity method, and exclusions. Check event tracking before interpreting a sudden change as a change in user behavior.
Start with one retention problem
Choose a recent cohort whose observation window is complete. Identify the first point where users fail to return, then investigate what they were trying to do.
Pick one fix and set a date to review the same retention measure after the next comparable group completes that window. That gives you a concrete decision about what to improve before you spend more bringing people in.
Frequently asked questions
How do you calculate user retention rate?
Divide the number of users from a starting cohort who meet your return rule by the number of users in that cohort, then multiply by 100. If 700 of 1,000 users return, retention is 70%. Exclude new users outside the cohort and wait until the full measurement window has elapsed.
What is the difference between user retention and customer retention?
User retention tracks whether people continue using a product. Customer retention tracks whether customers or accounts continue the commercial relationship. A business can renew a subscription even when fewer employees use it, so usage retention and customer retention can move differently.
Do retention and churn always add up to 100%?
Only when they use the same starting group, time period, and complementary definitions. If 70% of a group returns and your churn rule classifies the remaining 30% as churned, they add to 100%. Missing a specific day does not automatically mean someone has permanently left.
What is a good user retention rate?
There is no universal target. A useful benchmark must match your product, audience, return action, time window, and calculation method. Compare equivalent cohorts within your own business and assess whether improvements support revenue and margin. Include user counts because small groups can produce large percentage swings.
Can Google Analytics 4 measure user retention?
Yes. Google Analytics 4 provides a Retention overview and Cohort exploration for websites and apps. Google notes that Cohort exploration uses device data and does not consider User-ID when creating cohorts. Account for that limitation when interpreting retention across devices.
Get new small business insights by email
Practical ideas and useful articles to help you make better business decisions.
Tech Help Canada Staff researches, writes, and reviews practical content for business owners and professionals. Our coverage spans business, marketing, SEO, technology, and the tools and systems people use to grow and operate online. We focus on clear, useful information backed by research, hands-on experience, and editorial review. Learn more about our team and editorial standards. Need help with something? Contact Us
