Issue No. 2 | July 28, 2026

Tools, AI, and workflow intelligence for private equity professionals.

THIS MONTH: THE METRICS UNDERNEATH NET REVENUE RETENTION

THE DATA POINT: 80-90%

An annual Customer Retention Rate between 80% and 90% is often considered a healthy benchmark for a recurring revenue business, a different measure from Net Revenue Retention, which tracks dollars rather than headcount. The two can diverge in ways that matter more than either number alone. High CRR paired with NRR below 100% means customers are staying but spending less, a pricing or engagement problem rather than a churn problem.

The reverse divergence carries a different kind of risk. Rising NRR alongside falling CRR can indicate that a relatively small group of expanding accounts is offsetting broader customer losses, a composition a single blended NRR figure will not reveal on its own. Reading the two metrics together, rather than either in isolation, is where a retention story starts to become diagnostic instead of descriptive.

TOOL SPOTLIGHT: READING EXPANSION AND CONTRACTION IN THE RETENTION ANALYSIS TOOL

The Retention Analysis Tool, covered in our first issue, classifies every account into one of seven lifecycle states each period: New, Retained (Incr), Retained (Flat), Retained (Decr), Lost, Returning, and Inactive. The two Retained subclasses that get the least attention in most diligence conversations, Retained (Incr) and Retained (Decr), are often where the real story lives.

Retained (Incr) accounts drive Expansion Revenue Rate, one of the clearest signals of whether revenue is growing within existing customer relationships. Retained (Decr) accounts drive Contraction Revenue Rate, and they are the accounts most retention reporting misses entirely, since a contracting account is still counted as retained under a standard persistency calculation. The revenue erosion has already happened before it becomes visible in headline retention metrics.

The tool’s Waterfall Visualizations plot the movement between all seven states period over period, making it possible to see whether an NRR figure was produced by broad-based expansion or by a handful of large accounts masking contraction across the broader customer base. The Detail Sheet then traces that composition back to individual accounts, and the same inactivity window and minimum activity threshold settings from issue one still apply, recalculating every classification and metric instantly when adjusted.

FROM THE BLOG

Customer Retention Metrics: Beyond the Headline Numbers

This month’s post walks through each of the six metrics, NRR, CRR, churn, expansion, contraction, and new customer revenue percentage, and the seven-state classification framework that produces them. It argues that no single retention figure can describe a customer base on its own. Blended reporting and cohort-level classification answer different questions from the same underlying data.

THE LAST METRIC: NEW CUSTOMER REVENUE %

New Customer Revenue %, new customer revenue divided by total revenue, is the metric most likely to be read the wrong way in isolation. A rising number can mean a business is acquiring faster than ever. It can also mean a business is running harder on acquisition to cover an installed base that is not holding.

The distinction becomes much clearer in the context provided by NRR. Strong NRR paired with a high New Customer Revenue % describes a business adding new accounts on top of a durable existing base, a genuinely healthy growth profile. Weak NRR paired with the same high percentage describes something different: acquisition substituting for a retention problem. The topline holds either way. Only one of those two businesses has a growth engine that is less dependent on continued acquisition spending.

For a hold period model, that distinction compounds. A business growing primarily through new acquisition has a customer base that looks meaningfully different in year two than it did at close, and whether those newer cohorts retain as well as the ones the business was built on is not a question blended metrics can answer. It requires the same cohort-level visibility, by acquisition period, that the Retention Analysis Tool is built to produce.

Until next month,

If a retention figure in a CIM you are reviewing right now does not explain how it was classified, reply here and tell us what you are looking at. We are glad to think through it with you.