Definition
TTM Net Retention (Trailing Twelve Months Net Retention), at a specific point in time (say, at the end of June 2025), measures how revenue generated by customers who were active 12 months earlier (in our example, at the end of June 2024) has evolved over those 12 months.
More specifically, the calculation takes two points in time exactly 12 months apart (for example, June 2025 compared to June 2024) and computes the ratio:
TTM Net Retention = (BOP + Δ) / BOP
- BOP (Beginning of Period) = ARR produced 12 months ago from customers who were active 12 months ago
- Δ = Total ARR gained or lost by those same customers over the 12-month period
An example
Here’s how TTM Net Retention works with an example. You can also download the source file: TTM Net Retention Example.xlsx.
| Customer | June 2024 ARR | June 2025 ARR | ARR Change | Is ARR Change included in Δ? |
|---|
| A | $100 | $150 | +$50 upsell | ✅ Yes |
| B | $100 | $100 | $0 stable | - |
| C | $100 | $80 | -$20 downsell | ✅ Yes |
| D | $100 | $0 | -$100 churn | ✅ Yes |
| E | $0 (joined Jan 2025) | $120 | +$120 new | ❌ No |
| Totals | $400 | $450 | | |
BOP = $400
Δ = +$5 - $20 - $100 = -$70
TTM Net Retention in June 2025 = ($400 - $70) / $400 = $330 / $400 = 82.5%
Customer E’s +$120 variation is excluded — they weren’t present in June
2024, so they don’t factor into TTM Net Retention, which measures the
retention of June 2024’s customers in June 2025.
- 330$ comes from customers active in June 2024: this is June 2024’s 400$ ARR, multiplied by 82.5% (the TTM Net Retention Rate);
- 120$ comes from new customers, i.e. customers who were not producing revenue in June 2024: in our case, it’s customer E.
Note that, even if this is not the case in the provided example, TTM Net Retention may well be (and often is) above 100%: this happens whenever existing customer base, taken as a whole, is upsold across the 12-month window, and therefore increases its contribution to the revenues over time.
Approximate Method vs. Exact Method
Exact computation of TTM Retention requires looking at two columns, 12 months apart, of a Revenue Input Table recording customer-level data. Some analysts use a quicker approximate method to compute a TTM Net Retention rate out of ARR movements recorded in an ARR waterfall.
In practice, you compute Δ in the formula above by adding up
the violet rectangle in the picture below, which records all monthly ARR movements (except New ARR) that occurred during the 12-month window being considered.
BOP = $400
Δ = sum of violet cells = +$10
Approximate TTM Net Retention = (BOP + $10) / BOP = ($400 + $10) / $400 = 102.5%
When Approximate Method Works
We’ve seen that ARR movements from recent customers (more precisely, customers who were not active 12 months ago) explain why exact TTM Net Retention differs from the approximate method. The approximate method gives acceptably accurate results when the contribution of those recent customers to ARR movements during the 12-month window is negligible. In practice, this happens when new customer acquisition during the period is low, or when newly acquired customers have small movements compared to the other ones, so that most upsells, downsells, and churns observed during the 12-month window are attributable to customers who were already producing revenue at the beginning of such time window.
