Why Net Retention Is Not a Strategy
Net Revenue Retention has become the comfort blanket of subscription businesses.
If your NRR is 120%, investors are happy, the board is relaxed, and leadership feels justified in carrying on as usual. A single percentage tells everyone the business is compounding without needing many slides to explain it.
On its own, it tells you nothing about how you got there, what it cost, or whether the pattern is sustainable. It can hide cannibalisation, promo dependence, unnoticed product downgrades, and price decay. Treated carelessly, it can become a flattering mirror that blinds you to real risk.
This is not an argument against NRR. It is an argument against treating NRR as the plan rather than the scoreboard.
What NRR Actually Tells You
Stripped right back, NRR answers one question:
If you ignored all new customers this year, did the revenue from your existing customers grow or shrink?
Start with $1m in recurring revenue from last year’s customer base. End the year with $1.1m from those same customers after churn, downgrades, and upgrades. Your NRR is 110%. The existing base grew by 10%.
Executives like NRR for good reasons. It focuses attention on the quality of your customer relationships rather than just raw acquisition, and it is relatively hard to manipulate without someone noticing. If you are valuing a subscription business, NRR is one of the first numbers you should ask for.
The problem comes when that number is treated as a strategy in its own right, or when its components are poorly defined.
The Illusion Of A “Good” NRR
A healthy NRR can coexist with deeply unhealthy patterns.
You can maintain 115% NRR while actually training your sales team to give away margin in the form of discounts. You can report 110% NRR while your flagship product slowly erodes as customers trade down into cheaper bundles. You can hit 105% NRR whilst over-serving one segment and losing another.
On paper, the number looks fine. Underneath, you may be eroding the very economics that number is supposed to protect.
Two forces create this illusion.
First, NRR is an average. A single percentage blends together segments, products, and behaviours that are often moving in very different directions.
Second, NRR depends entirely on how you define “existing customers”. If your cohort definition is sloppy, the metric can end up inflated by revenue that does not belong there at all.
Three Companies, Same NRR, Very Different Futures
Consider three subscription businesses, each reporting 110% NRR. To an investor scanning a dashboard they look equally healthy, but let’s look at the reality.
Company A: The Hidden Segment Problem
Company A has:
Enterprise customers at 130% NRR
Mid-market at 108%
SMB at 85%
Weighted together, that still lands at 110%, so it appears that retention is in good shape.
Yet the SMB segment is bleeding unnoticed. Those customers churn early, require support, and often damage reputation when they leave. The enterprise segment is doing so well that it masks the damage.
Unless you break NRR down by segment, you are effectively letting your best customers hide problems with your worst.
Company B: The Promo-Dependent Mirage
Company B has an aggressive renewal playbook. The rules are simple: no renewal should be lost on price.
Sales teams are given broad authority to discount renewals, add “free” modules, and extend trial terms inside multi-year agreements. Almost every at-risk customer is rescued with some kind of sweetener.
On paper, NRR is 110%. The existing base grows each year.
Look a little deeper and you’ll see the effective price per unit is declining. Average discount levels are climbing. Expansion ARR is inflated by temporary uplift that has to be renewed, discounted, or replaced in the next cycle.
On paper, they look the same as Company A. In reality, one is paying for its NRR with future margin, and the other is quietly compounding at high quality.
The company has not built a retention strategy. It has built a discount strategy and is calling it retention. The NRR looks strong, but the unit economics deteriorate every year.
Company C: The Subtle Compounding Machine
In Company C churn is low and boring. Expansion comes mostly from measured, value-led upgrades rather than heavy discounting. There is limited cannibalisation between products. Cross sell is strong, yet customers do not need large promotions to stay.
Identical NRR to the other companies, but clearly a different strategy.
How Cohorts And Calculations Can Distort NRR
NRR depends on a simple idea: compare revenue from the same cohort of customers across two periods. Execution is rarely that clean though.
A common pitfall of NRR is the way “existing customers” are defined.
Imagine you start the year with $1m of ARR from your January customer base. Over the year, you sign new clients, some of whom renew or expand mid-term. If your NRR logic simply sums all expansions from customers who currently exist in the system, you can end up counting:
Cross sell to customers who were not in the starting cohort
Mid-term upsells to customers whose initial deal never featured in the baseline
In that situation the line between true retention and growth from new logos has started to blur. Now NRR is answering a looser question: “did revenue from anyone who happens to be a customer today grow?”.
The same distortion happens when NRR is calculated on bookings rather than recurring revenue, or when the contract start and renewal dates are misaligned. Noise from new sales can leak into a metric that is supposed to isolate the health of the existing base.
Executives rarely see this level of detail in a board pack. They see the headline number and interpret it as a direct proxy for customer health.
How Executives Misinterpret NRR
Most executive missteps with NRR fall into a few patterns.
NRR is treated as a target in itself rather than an outcome of deliberate choices. Leadership sets an NRR goal of, say, 115% and asks teams to “hit the number”. Sales and customer success respond with the tools they control most directly: discounting, bundling, and promotion.
NRR is interpreted as a single, universal truth about the customer base. Segment differences, product cannibalisation, and pricing effects are glossed over because the top line looks acceptable.
Cohort quality is taken for granted. Few boards ask exactly which customers are in the NRR denominator, or how cross sell to new clients is being treated.
The result is a metric that looks precise but is only loosely connected to the questions you actually care about:
Are we building durable, profitable relationships, or paying for retention with margin?
Are some segments thriving while others slowly die?
Are we training customers to expect a deal every time they come up for renewal?
Using NRR As A Scoreboard, Not A Strategy
NRR becomes truly useful when executives treat it as the headline of a story rather than the whole plot.
The starting point is simple. Continue to track NRR, and treat a consistently high number as a genuine positive. Just resist the temptation to stop there.
Break it down into its components. Separate:
Churned revenue: customers who left entirely
Contraction: downgrades, reduced seats, or reduced usage
Expansion: upgrades, cross sell, and price increases
Then look at these components by segment, product, and geography. An overall 110% NRR with minimal churn and healthy expansion in enterprise looks very different from 110% propped up by steep promotions in SMB.
Next, examine your reliance on deal sweeteners. If your renewal playbook depends heavily on “save” deals, free months, or large one-off discounts, you are effectively borrowing NRR from the future.
Finally, tighten your cohort definition. Ensure that:
Only customers who existed at the start of the period are in the NRR base
Cross sell to customers acquired mid-year is clearly separated as growth, not retention
One-off non-recurring items are excluded or transparently flagged
These adjustments do not make the number prettier. They make it more honest, which is what leadership actually needs.
What Leaders Should Really Ask Of NRR
When an executive team reviews NRR, the most valuable questions are often not “is this good?” but “what sits underneath this?”.
For example:
If we stopped all discounts tomorrow, what would happen to NRR by segment?
If we isolated our top 50 customers, what does their NRR look like versus the rest?
If we removed cross sell to customers acquired this year, how does NRR change?
Those questions turn NRR from a comfort blanket into a diagnostic.
Used this way, NRR becomes a powerful lens rather than a vanity metric. It helps you decide where to invest, which segments to reshape, and where the real strategic risk lies.
NRR Is The Score, Not The Playbook
Net Revenue Retention remains one of the best single numbers for evaluating a subscription business. It is difficult to fake convincingly. It rewards genuine customer success. It correlates strongly with long-term value.
Just don’t mistake it for strategy. Strategy lives in the choices you make about product, pricing, segment focus, and how you earn the right to expand with customers over time. NRR simply tells you whether those choices are compounding or eroding value within the existing base.
Treat NRR as your scoreboard. Use it to highlight where you should pay attention. Then go beneath the headline number to understand what is really driving it, and how to keep it that way.