SaaS Retention Analysis: Metrics, Cohorts and a Dashboard

SaaS retention analysis answers one question: of the customers and revenue you had at the start of a period, how much do you still have at the end, and why? You need three things to answer it. First, a handful of metrics (customer retention, gross revenue retention, net revenue retention). Second, cohort tables that show when people leave. Third, a dashboard that puts both in front of the team every week, so the numbers get acted on and not just calculated.

We build dashrendr, an analytics platform for SaaS teams, so we look at this from the tooling side as well as the maths. This guide covers both.

dashrendrapp.dashrendr.comMake retention metrics easier to trackBuild custom dashboards for internal metrics, embed analytics for your customers, or schedule retention reports to arrive by email.Try the dashboard playground

TL;DR

• Retention analysis measures how many customers and how much recurring revenue you keep over a set period, then digs into the patterns behind that number. Churn is the inverse of retention.
• Track four numbers: customer (logo) retention, gross revenue retention (GRR), net revenue retention (NRR) and N-day retention for product usage.
• Cohort tables and retention curves tell you when customers leave. A single blended churn rate can’t.
• Benchmarks vary a lot by contract size. Compare yourself with companies at a similar stage and ACV, not a headline average.
• A retention dashboard that refreshes itself and lands in inboxes on a schedule beats a spreadsheet someone updates monthly.

Table of Contents

• What retention analysis measures (and how it differs from churn analysis)
• The four metrics worth tracking
• What a good number looks like
• Cohort analysis: finding out when customers leave
• From analysis to action: reducing churn
• What a retention dashboard should show
• Getting the data in one place
• Promo: build the dashboard without building a BI stack
• Perspective
• Five mistakes that make retention analysis useless
• FAQ
• Sources

What retention analysis measures (and how it differs from churn analysis)

Retention analysis looks at how many customers, users and how much revenue a company keeps over a set period, and then works out the patterns behind that figure, as Gainsight’s guide to retention analysis describes it. Customer retention rate counts the accounts you kept. Churn rate is simply its inverse: if you keep 95% of customers, churn is 5%.

So are they the same thing? Mathematically, nearly. In practice they point your attention in different directions:

• Churn reporting tells you what you lost. It’s a scoreboard.
• Retention analysis asks who stayed, who left, when, and what they had in common. It’s a diagnosis.

A founder who only watches a monthly churn percentage knows something is wrong. A founder who segments retention by plan, signup month and first-week behaviour knows where to look.

The four metrics worth tracking

You don’t need thirty KPIs. These four cover most of the ground.

Customer (logo) retention rate

The formula, per Gainsight: (customers at end of period minus new customers acquired in the period) ÷ customers at start of period × 100.

Worked example: you start the quarter with 200 customers, win 30 and finish with 218. Customers kept from the start: 218 − 30 = 188. Retention is 188 ÷ 200 = 94%, so churn is 6%.

This counts every account equally, so it’s the right metric when you want to know how many relationships you’re keeping. It hides the fact that losing one enterprise customer hurts more than losing five on the cheapest plan.

Gross revenue retention (GRR)

GRR measures the recurring revenue you keep from existing accounts, excluding any expansion. It can’t go above 100%, which makes it a clean read on how leaky the bucket is. If you want the recurring revenue basics first, our MRR glossary entry covers how to count it consistently.

Net revenue retention (NRR)

NRR includes upsells and cross-sells on top of what you kept. SaaS Capital’s retention research defines it over a twelve-month window: take the MRR today from customers who were already customers twelve months ago, and divide it by their total MRR back then. New logos are excluded.

NRR over 100% means your existing base grows even if you sign nobody new. For the formula, edge cases and how to read it, see our NRR guide.

Read GRR and NRR together. A healthy-looking NRR can sit on top of a weak GRR if a few large accounts expand while many small ones quietly leave.

N-day retention

The first three metrics are about contracts. N-day retention is about behaviour: of the users who signed up on a given day, what share came back on day 1, day 7, day 30? It’s your earliest warning, because usage drops weeks or months before a cancellation. It pairs naturally with activation rate, since people who never reach their first moment of value rarely come back.

Metric	Counts	Includes expansion?	Best for
Customer retention	Accounts	No	Overall health of the base
GRR	Recurring revenue	No	How leaky the bucket is
NRR	Recurring revenue	Yes	Growth from the existing base
N-day retention	Active users	Not applicable	Early warning on engagement

What a good number looks like

Benchmarks are useful for orientation and dangerous for comfort. In SaaS Capital’s 2025 survey of private B2B SaaS companies, median NRR was 101% and median GRR 91%. The same research shows NRR rising with contract size: a median of 104% for companies with an ACV of $100K to $250K, and 106% above $250K.

What that means for you:

• If you sell $15 a month plans to small teams, an enterprise NRR benchmark isn’t your yardstick. Self-serve products typically live with more logo churn and lean on GRR and customer retention.
• If you sell larger contracts, expansion is part of the story and NRR deserves a prominent spot.
• Your own trend matters more than any survey median. A GRR that slipped three points over two quarters is a finding, whatever the benchmark says.

Cohort analysis: finding out when customers leave

A blended churn rate averages old, loyal customers with brand-new ones. That blur is exactly what cohort analysis removes. You group customers by when they started (usually the signup or first-payment month) and follow each group over time. Our cohort analysis guide for SaaS walks through the set-up in detail.

Build the table

Rows are cohorts (January signups, February signups and so on). Columns are months since signup: month 0, 1, 2, 3 and on. Each cell is the share of the cohort’s customers, or revenue, still active. Build one table for logos and one for revenue. The gap between them shows whether the customers who stay are spending more or less.

Read the shape of the curve

Plot one cohort as a line and the shape tells you where to focus. As one cohort-analysis explainer puts it, a curve that drops early and then flattens means the remaining customers are sticking, while one that keeps sloping down hasn’t found a floor.

• Steep drop in months 0 to 2, then flat: an onboarding and activation problem. The product works for people who get going, and too many don’t.
• Steady decline with no floor: the product isn’t becoming part of anyone’s routine. Look at feature adoption and the use case itself.
• Cliff at month 12: annual renewals. Look at what happens in the 90 days before renewal.
• Curve that rises (revenue view): expansion is outpacing churn within that cohort.

Compare cohorts with each other

The other half of the value is down the columns. Is the March cohort retaining better at month 3 than the January cohort did? If a change to onboarding, pricing or the product shipped in between, the table is your before-and-after test. Cut cohorts by plan, acquisition channel or company size too. You’ll often find one segment holding up the whole average.

A word of caution on small numbers: a cohort of 12 customers will swing wildly on one cancellation. Wait for enough volume before drawing a conclusion, or lump cohorts into quarters.

From analysis to action: reducing churn

Metrics only matter if they change what the team does. A workable loop looks like this:

1. Find the leak. Use the cohort table to locate the stage where customers drop off.
2. Define risk signals. Pick the usage signals that precede churn in your own data: logins falling, key feature unused, fewer active seats.
3. Score accounts. Predictive health scores built on product-engagement signals can flag risk 30 to 90 days ahead, according to Pepper Effect’s customer success guide.
4. Route at-risk accounts to a playbook: a check-in call, a training session, a nudge email.
5. Recalibrate. The same source suggests checking scores against accounts that actually churned and stayed over the last 12 to 24 months, and recalibrating quarterly.

If your product already has dashboards inside it, those dashboards generate engagement data you can use. We cover that in reducing churn with dashboard engagement data.

What a retention dashboard should show

A retention dashboard is a small number of charts that answer are we keeping customers, and where are we losing them? at a glance. For most SaaS teams, a good first version has:

• Headline tiles: customer retention, GRR and NRR for the current period, with the change on the previous one.
• MRR movement: new, expansion, contraction and churned MRR by month.
• Cohort heatmap: the retention table, coloured by percentage.
• Retention curves: a line per recent cohort, so shape changes are visible.
• Segment cuts: retention by plan and by customer size.
• Engagement leading indicators: N-day retention and weekly active accounts.
• At-risk list: accounts whose usage has dropped, with owner and last contact.

Resist adding more. Every extra chart is one more thing nobody checks.

Three audiences, one set of data

• Your internal team wants the full picture: cohorts, at-risk accounts, MRR movement. See how internal dashboards for operations and support teams can be set up.
• Leadership and investors want a weekly or monthly snapshot of the same headline numbers, delivered without anyone asking.
• Your customers can benefit from a retention view of their own: usage, adoption and results in your product. Customers who can see the value they get are easier to keep. Our guide to customer-facing analytics for SaaS covers what to expose, and five metrics to show customers is a good shortlist.

Getting the data in one place

Retention analysis usually needs data from several systems: billing for revenue, your application database for usage, product analytics for events, a spreadsheet someone keeps for renewals. Before you build anything, settle three definitions and write them down:

• What counts as an active customer (paid, trialling, paused)?
• What counts as churn (cancellation date, or end of the paid period)?
• When does a customer join a cohort (signup, first payment)?

Teams skip this and then argue about numbers for months. Pick definitions, stick with them, and note them on the dashboard itself.

Promo: build the dashboard without building a BI stack

dashrendr is an analytics platform for SaaS teams. You connect your data sources (Google Analytics, BigQuery, PostgreSQL, MySQL, MongoDB, Google Sheets) or push your own data through the REST API. Then you build dashboards in a drag-and-drop editor, with no code needed for the visual side.

From there, a retention dashboard can go three places: embedded in your product so customers see their own data, on your team’s screens for internal tracking, or emailed on a schedule as a daily, weekly or monthly report. Our automated reports feature handles that last part, and exports are available as PDF, CSV and PNG.

Plans start at $6/month. There’s a 14-day free trial and no credit card is required. See dashrendr’s pricing for the Hobby, Starter, Grow and Premium plans, or start from the pre-built templates to get a first dashboard up quickly.

Perspective

Most retention problems we see aren’t maths problems. The formulas fit on a sticky note. The real problem is that the numbers live in a spreadsheet that one person updates when they have time, so by the time anyone sees a churn spike, the customers are already gone.

Our view is that retention reporting should be boring and automatic. Connect the data once, build the dashboard once, and let it refresh and arrive in inboxes on a schedule. That’s the middle ground we built dashrendr for: more structure than a chart library where you build layouts, multi-tenancy and exports yourself, and far less cost and set-up effort than tools that run to $1,000+ a month.

The other thing we’d push is showing retention-related data to customers. If your customer can open a dashboard in your product and see what your product is doing for them, that’s a retention tool in itself. It’s also why we built embedding and multi-tenant security in from the start.

Paloma

Five mistakes that make retention analysis useless

1. Using one blended churn rate. It hides the cohort that’s actually leaking.
2. Mixing logo and revenue views. Report both, label them clearly, and don’t swap one for the other mid-year.
3. Counting new customers in NRR. NRR looks only at customers who were already there at the start of the window.
4. Comparing to the wrong benchmark. Match on contract size and stage.
5. Never recalibrating. Health scores and risk definitions drift. Check them against actual churn every quarter.

FAQ

What is the difference between retention analysis and churn analysis?

They use the same underlying data. Churn rate is the inverse of customer retention rate, so 94% retention is 6% churn. Churn analysis focuses on what you lost, while retention analysis also looks at who stayed, when the others left and what the patterns suggest.

Which retention metrics should an early-stage SaaS team track first?

Start with customer retention, GRR and N-day retention. Add NRR once you have meaningful expansion revenue to measure. Early on, the most useful single view is a cohort table by signup month, because it shows where new customers drop off.

How often should I review retention metrics?

Weekly for leading indicators like N-day retention and at-risk accounts, monthly for GRR, NRR and cohort tables. A scheduled email report helps here, since the numbers arrive whether or not anyone remembers to open a dashboard.

Can I show retention dashboards to my customers?

Yes. With dashrendr, dashboards can be embedded in your product, with multi-tenant security so each customer only sees their own data. You can also email scheduled reports to customers on a daily, weekly or monthly basis.

Sources

• The Essential Guide to Retention Analysis (Gainsight)
• 2025 B2B SaaS Retention Benchmarks (SaaS Capital)
• SaaS Cohort Retention Analysis (Livmo)
• Customer Success Metrics: The KPIs That Predict Retention (Pepper Effect)

Recommended

• Stop Losing Users: Cohort Analysis for SaaS
• Cohort Analysis: Definition, Formula & How to Track It
• 5 Metrics SaaS Founders Must Show Customers
• Dashboard Design Best Practices for SaaS