Cohort Analysis for Series A: Investor-Grade Templates
Cohort retention is the single most inspected Series A metric because it is hardest to fake. Build it right and partners underwrite the business. Build it wrong and diligence unwinds everything around it.
Institutional Series A diligence probes cohort analysis more rigorously than any other quantitative dimension of the business. Revenue is a point. Growth is a line. Cohort retention is a structure — and the structure either holds or it does not. Soft cohort curves are the single most common reason a warm Series A process goes cold. Strong curves that are properly constructed are the single strongest signal of operational quality at the stage.
Key Takeaway: Partners do not trust cohort analysis they did not see constructed. Your own cohort data, presented with full construction visibility, is worth more than a third-party analytics tool's export.
The three cohort views that matter
1. Revenue retention by acquisition cohort
Group customers by the month or quarter they were acquired. Track revenue from each cohort month-over-month. The picture that emerges tells the partner whether revenue holds, expands, or decays. Three variants: gross revenue retention (GRR — what happened if no churn), net revenue retention (NRR — after churn and expansion), and cohort LTV (cumulative revenue over the cohort's life).
2. Logo retention by acquisition cohort
The same view, but counting customers rather than revenue. Logo and revenue retention can diverge — a cohort may show strong revenue retention because a small number of customers expanded aggressively while many smaller customers churned. The divergence is a diligence signal in its own right.
3. Cohort behaviour by segment
The first two views done separately for each major customer segment — enterprise vs mid-market, industry vertical, customer size tier. Blended cohort retention often masks segment-level weakness. Partners probe the segment cut specifically to find the weakness.
Construction rules
Start from the first transaction, not the first contract. Cohort month is the month the customer first paid. Contract-date construction understates early churn.
Include everyone. Do not exclude "test accounts", "reference customers", or "strategic accounts" from the cohort numerator. Partners discover exclusions in diligence and treat them as construction flags.
Show cohorts at least 12 months old. Younger cohorts are noisy. Partners trust cohorts with enough age to show the curve shape.
Expose the denominator. Revenue retention at Month 12 means what, divided by what? Be explicit.
Reconcile to the management accounts. Total cohort revenue at each month should tie to the revenue line in the financials for the same period. Numbers that do not reconcile create a diligence hole.
Common cohort construction errors
Blending after acquisition channel shifts
If your acquisition mix has changed materially — say, outbound-heavy in Year 1, product-led in Year 2 — then a blended cohort view is misleading. Show the mix and, where possible, segment by acquisition channel.
Ignoring the pricing change
A price increase mid-cohort materially affects NRR without reflecting anything about actual retention. Flag pricing changes explicitly on the cohort chart and, where possible, show cohort retention on a constant-price basis.
Mixing contract types
Monthly and annual subscriptions have different cohort dynamics. Show them separately where both exist. Partners would rather see two cleaner curves than one muddled blended curve.
Presenting only the best cohort
Pitch decks sometimes feature the strongest cohort as if it were representative. Diligence teams always ask to see every cohort. Partners who later discover the "representative" cohort was cherry-picked lose trust in every other number presented.
Warning: The single fastest way to damage a Series A process is to present a cohort curve that does not survive recalculation. Partners build the cohort themselves from the management accounts during diligence; any divergence from what you presented is interrogated in detail.
What good looks like at Series A
Institutional Series A partners are looking for specific shapes. Median expectations by sector differ; directional guidance below reflects the pattern visible across recent transaction data.
SaaS: Gross revenue retention above 85%, net revenue retention above 105%, with top-quartile NRR above 120%. Logo retention above 80% at 12 months.
Consumer: Repeat-purchase rate with a clear shape. For subscription consumer, similar SaaS metrics apply. For transactional consumer, partners want evidence that the second and third purchases have compounding value.
B2B transactional: Revenue retention varies widely. Partners look for expansion motion evidence — customers who started small and grew, or customers who added product lines.
These are directional ranges. The specific number matters less than whether the trajectory is improving, the construction is rigorous, and the segment view holds up.
The connection to intangible assets
Cohort retention is not just a metric. It is the evidence of customer capital — the intangible asset that retention measures. The Opagio 12 surfaces customer capital as one of the 12 drivers partners underwrite, and the cohort analysis is what populates the evidence for that driver. Strong cohorts mean deep customer capital; weak cohorts mean the asset needs work.
For the register-grounded view of customer capital and how it maps into institutional diligence, see Series A readiness. For the full metrics construction approach, see what metrics you actually need.
The Bottom Line
Cohort analysis is the evidence layer for customer capital. Build it from the first-transaction date, include everyone, segment where it matters, and reconcile it to the financials. The curve that results either holds or doesn't — and partners can read the answer in thirty seconds.
Put the cohort story in the register
The diagnostic maps your customer capital to what Series A partners actually underwrite — cohort retention as intangible asset evidence.