Strategic Brand Positioning, Category Creation & Product MarketingPlaybook3 min readUpdated September 2026

Building a Cohort LTV Table That Tells You Something Useful

A single lifetime value number for your whole customer base hides more than it reveals. Customers who signed up eight months ago behave differently than customers who signed up last month, and averaging them together makes it impossible to tell whether the business is actually getting better or worse at keeping and growing customers.

Vendors Covered in this Article

Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.

How do you set up a cohort LTV grid?

Build a table with signup month down the rows and months since signup across the columns. Each cell holds one number: the average cumulative revenue per customer from that signup month, measured that many months later. This is the standard cohort layout, and it works whether you are tracking ten customers or ten thousand. The value is not in the format, it is in what the grid reveals once it is filled in.

Fill in Retained Revenue First, Then Add Expansion

Start with the simplest version: what percentage of a cohort's starting revenue is still active in each later month. This alone usually surfaces something useful, such as a sharp drop-off around month three that a blended average would never show. Once that is working, add expansion revenue, upgrades, add-ons, and cross-sells, layered on top of the base retained amount, so the table shows not just what stuck around but what grew.

Keep the two layers visually distinct, even if they end up in the same cell as a combined total. A cohort whose total revenue looks flat could be quietly losing customers while the remaining ones expand enough to mask the churn, and that is a very different situation from one that is genuinely stable across the board.

Why compare cohorts to each other instead of to a target?

The real value of the grid shows up when you read across a diagonal: comparing the January cohort at month three to the April cohort at month three. If the more recent cohort is retaining and expanding better than the older one, something the team changed, in onboarding, in pricing, in the product itself, is working. If it is worse, that is worth investigating before it shows up as a company-wide trend a year from now.

Resist the temptation to compare against an arbitrary target number pulled from somewhere else, such as a figure a competitor's press release mentioned or a round number that felt reasonable in a planning meeting. Your own recent cohorts are a better benchmark than an outside number, since they reflect the same product, the same pricing, and the same customer base the newest cohort actually came from.

Keep the Review on a Fixed Schedule

A cohort table that gets built once and never revisited is not much better than the single blended number it was meant to replace. Put a recurring review on the calendar, monthly is usually enough, where someone actually looks at the newest diagonal and flags anything that looks off. A tracker in a tool like ClickUp works well for logging what was noticed each month and who is following up on it, so the review produces action items instead of just an interesting chart nobody acts on.

Build and maintain the table in this order:

  1. Put signup month down the rows and months since signup across the columns.
  2. Fill each cell with average cumulative revenue per customer from that signup month, measured at that point.
  3. Start with the share of each cohort's starting revenue still active, then layer on upgrades, add-ons and cross-sells.
  4. Read across the diagonals to compare cohorts at the same age, such as January and April at month three.
  5. Review the newest diagonal on a fixed monthly schedule and log anything that looks off.

Turn It Into Something Leadership Actually Reads

A raw spreadsheet grid is useful for the person building it and hard for anyone else to parse quickly. Once the table is in good shape, turning the key diagonal comparisons into a simple visual for a leadership update makes the finding land faster than handing over the underlying spreadsheet. A tool such as Gamma can turn that comparison into a clean one-page summary for a monthly or quarterly review without redesigning a deck from scratch each time.

Watch for a Cohort That Looks Great Too Early

A brand-new cohort can look unusually strong at month one simply because early enthusiasm has not had time to fade yet. Avoid drawing conclusions from a cohort's earliest months alone, and wait until it has enough history to show whether that early strength actually held up. A cohort that looks like your best one ever at month one and looks average by month four is a reminder that the diagonal only means something once it has enough columns filled in to compare fairly.

Executive Capability Standard

What Good Looks Like

A good cohort LTV table tracks retained and expansion revenue by signup month, gets compared diagonal to diagonal against recent cohorts, and gets reviewed on a fixed schedule rather than built once and forgotten.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Pull your last six months of signups and revenue data to see whether you currently have what you would need to build even a basic cohort grid.
2. Do Manually:Build the first version of the grid by hand in a spreadsheet, starting with retained revenue only, before adding expansion revenue on top.
3. Delegate:Assign an analyst or ops hire to own the monthly review and flag anything unusual in the newest diagonal before it becomes a company-wide trend.
4. Automate:Set up a recurring data pull that refreshes the cohort table automatically each month instead of rebuilding it by hand every time.
5. Buy:Bring in a data or finance specialist to build a more sophisticated version once the basic table is proving useful and the business can justify the investment.

How to Get Started

Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.

Frequently Asked Questions

How many months of data do we need before this is useful?

You can start reading signal after a handful of cohorts, even with only two or three months of history each. The table becomes more valuable over time as more diagonals fill in, but waiting for a full year of data before starting means losing months of early warning you could have had instead.

Should we track cohorts by signup month or by something else?

Signup month is the standard starting point and works for most businesses. If you have a strong reason to believe a different grouping matters more, such as acquisition channel or plan tier at signup, build a second table around that grouping once the basic monthly cohort table is working.

What counts as a concerning drop between cohorts?

There is no universal cutoff. Compare each cohort's diagonal to the two or three cohorts before it. A drop that shows up once might be noise, but the same drop repeating across consecutive cohorts is a pattern worth investigating rather than dismissing.

About the numbers

This guide doesn't quote a sourced benchmark. Figures in it are estimates or general guidance, so check them against your own numbers.

Related Guides