BYTETOOLS

Cohort Retention Calculator

Paste a cohort grid and get the retention triangle, average retention curve, month-1 and month-12 retention and the implied average customer lifetime.

58.0%
month 1 retention
40.3%
month 3 retention
—
month 12 retention
2.4
Implied lifetime (months)
6
Cohorts
8,650
Total users tracked
86.3%
Steady-state retention
7.3
Steady-state lifetime (months)
CohortSizePeriod 0Period 1Period 2Period 3Period 4Period 5
2024-011,200100.0%57.0%45.0%39.0%36.0%34.0%
2024-021,450100.0%58.0%46.6%41.0%38.0%
2024-031,310100.0%58.0%47.0%41.0%
2024-041,580100.0%58.0%47.0%
2024-051,490100.0%59.0%
2024-061,620100.0%
Average—100.0%58.0%46.4%40.3%37.0%34.0%
monthAverage retentionWeighted retentionCohorts with dataUsers still active
0100.0%100.0%68,650
158.0%58.0%54,080
246.4%46.4%42,573
340.3%40.4%31,599
437.0%37.1%2983
534.0%34.0%1408

retention(period n) = active users in period n ÷ cohort size  ·  implied lifetime = 1 ÷ (1 − retention)

Worked example: a cohort of 1,200 signups with 684 still active in month 1 retains 684 ÷ 1,200 = 57.0%. Holding that rate every month implies an average lifetime of 1 ÷ (1 − 0.57) = 2.3 months.

Blank trailing cells are treated as “not measured yet” and excluded from the averages, so a young cohort never drags the curve down. A cohort whose size is zero is skipped rather than dividing by zero. The 1 ÷ (1 − r) lifetime assumes a constant retention rate forever, which almost never holds — real curves flatten, so the steady-state figure computed from later periods is usually the more honest one. Everything runs locally in your browser; your cohort data is never uploaded.

What is the Cohort Retention Calculator?

A cohort retention table is the clearest picture of whether a product keeps the people it acquires. Paste a grid of active users — one row per cohort, one column per period — and this tool turns the raw counts into retention percentages, draws the retention triangle with a colour ramp and derives the average curve underneath it.

  • Accepts comma, semicolon or tab separated grids pasted from a spreadsheet
  • Colour-ramped retention triangle with optional raw user counts
  • Average and user-weighted retention curve per period
  • Period-1, period-3 and period-12 retention plus implied lifetime
  • Steady-state retention taken from period-over-period ratios
  • One-click CSV download and copy of the calculated percentages

How to use the Cohort Retention Calculator

  1. 1

    Paste your cohort grid: a label in the first column, then the active users in each period.

  2. 2

    Leave the first number column as the cohort size — retention is measured against it.

  3. 3

    Choose whether the columns are months, weeks or days, and which period drives the lifetime figure.

  4. 4

    Read the retention triangle, the colour ramp showing where the curve flattens, and the average row beneath it.

  5. 5

    Download or copy the retention CSV to share the percentages with your team.

About the Cohort Retention Calculator

A cohort retention table is the clearest picture of whether a product keeps the people it acquires. Paste a grid of active users — one row per cohort, one column per period — and this tool turns the raw counts into retention percentages, draws the retention triangle with a colour ramp and derives the average curve underneath it.

It also does the arithmetic people usually reach for next: retention at period 1, 3 and 12, a steady-state rate taken from the later periods, and the implied average customer lifetime as 1 ÷ (1 − retention). Blank cells count as not yet measured and stay out of the averages, so a young cohort never drags the curve down.

The grid is parsed entirely in your browser. Nothing is uploaded, stored or sent to a server, so you can paste real user counts straight out of a spreadsheet, and the calculated percentages can be downloaded as CSV or copied back into one. A cohort whose size is zero is skipped rather than divided by, and these are planning estimates rather than financial advice.

Frequently asked questions

How do you calculate cohort retention?

Divide the number of users from a cohort still active in a given period by the size of that cohort at the start. If 1,200 people signed up in January and 684 were still active a month later, month-1 retention is 684 ÷ 1,200 = 57%.

What format should the cohort grid be in?

One row per cohort. The first column is the cohort name or month, and the columns after it are the active users in period 0, 1, 2 and so on. Period 0 is the cohort size. A header row is detected and skipped automatically, and trailing blanks are fine.

How do you turn retention into an average customer lifetime?

Average lifetime is 1 ÷ (1 − retention rate) periods. At 90% monthly retention that is 1 ÷ 0.10 = 10 months. It assumes the rate stays constant forever, which it rarely does — real curves flatten, so the steady-state figure from later periods is usually more honest.

Why is one of my rows being skipped?

A row is skipped when its first number column is missing, zero or not a number, because retention would then be a division by zero. Check that the cohort size sits immediately after the label and that stray text has not crept into the column.

What is a good month-1 retention rate?

It depends entirely on the category. Consumer mobile apps often lose two thirds of a cohort in the first month, while B2B software sold annually can hold above 95%. The shape matters more than the level: a curve that flattens has a sustainable business behind it, one that keeps falling does not.

Related tools