Are newer cohorts more or less valuable than older ones?
Are the customers you acquired this year as valuable as the ones you acquired two years ago? Cohort analysis answers this question. Here's how to run one and what it tells you about the direction of your business.
The short answer
Group customers by the quarter they first purchased, then compare revenue per customer and retention rates across cohorts. If Q1 2025 customers generated $4,800 in their first 6 months but Q1 2026 customers are on pace for $3,200 in the same window, your newer cohorts are less valuable. That means your acquisition quality, pricing, or product-market fit may be deteriorating.
Why cohort analysis reveals what aggregate metrics hide
Your overall LTV might look stable at $18,000. But if customers acquired in the last 6 months have a projected LTV of $12,000 while customers from two years ago had $24,000, your aggregate number is being propped up by legacy customers. The trend line is downward.
This happens more often than you would expect. Businesses scale marketing, reach a broader audience, and the new customers are less targeted, less loyal, or less willing to pay. The aggregate stays fine because old customers are still spending. But when those old customers eventually churn, the aggregate catches up to reality.
Cohort analysis is the early warning system. It shows you what is happening to customer quality at the point of acquisition, not after it has already washed through your financials.
What cohort analysis actually measures
A cohort is a group of customers who share a common start date (usually the quarter or month of their first purchase). For each cohort, you track:
Comparing these metrics across cohorts tells you whether customer quality is improving, stable, or deteriorating. If Q3 2025 customers are retaining at 60% after 6 months but Q3 2024 retained at 75% at the same point, something changed in your acquisition, product, or competitive landscape.
How to run a basic cohort analysis from QuickBooks Online
- 1Export Sales by Customer Detail for all time
Go to Reports → Sales by Customer Detail. Set to “All Dates.” Export to CSV. This gives you every transaction for every customer with dates and amounts.
- 2Identify each customer's first purchase date
In your spreadsheet, use MINIFS to find the earliest transaction date for each customer. Group these into quarterly cohorts (e.g., Q1 2025, Q2 2025).
- 3Calculate revenue per cohort at fixed intervals
For each cohort, sum the revenue from months 1-3, months 4-6, months 7-12, etc. Divide by the number of customers in the cohort for a per-customer figure.
- 4Calculate retention at each interval
For each cohort at each interval, count how many customers had at least one transaction. Divide by the starting cohort size.
- 5Compare cohorts side by side
Build a table with cohorts as rows and time intervals as columns. Compare: is the Q1 2026 cohort tracking ahead of or behind Q1 2025 at the same age?
Total time: 60-90 minutes. This is the most analytically complex metric in this guide. It requires a large data export, date-based grouping, interval calculations, and multi-dimensional comparison.
How to run a cohort analysis from Xero
- 1Export all invoices from Business → Invoices
Set to all time. Export to CSV. You need the full invoice history with customer names, dates, and amounts.
- 2Assign cohorts and calculate intervals
Same spreadsheet process as QuickBooks: find first purchase date per customer, assign to quarterly cohorts, then calculate revenue and retention at each interval.
- 3Build and compare the cohort table
Rows = cohorts, columns = time intervals. Compare revenue per customer and retention rates across cohorts at equivalent ages.
Total time: 60-90 minutes. The analytical complexity is identical to QuickBooks. Xero's invoice export is the starting point instead of Sales by Customer Detail.
Why cohort analysis is the most underused metric in small business
- 60-90 minutes of advanced spreadsheet work. This is not a report you pull. It is a multi-step analysis that requires pivot tables, date arithmetic, and structured comparison.
- It requires your complete transaction history. You cannot do cohort analysis with one month of data. You need at least 12-18 months of history, ideally more.
- The payoff is enormous when you get it right. No other metric tells you with such clarity whether your business is improving or deteriorating at the customer level. Companies that track cohorts make better acquisition, pricing, and retention decisions.
Or get cohort analysis done automatically
Bottomline groups your customers into quarterly cohorts automatically, tracks revenue per customer and retention rates at each interval, and shows you the comparison so you can see whether newer customers are matching older ones.
The trend is clear: newer cohorts are less valuable. Customer count is increasing (which feels like growth) but value per customer is declining. This is the kind of insight that aggregate metrics cannot show and that manual analysis rarely catches because of the time investment required.