Reading the Bridge: Auditing an Account-Level MRR Waterfall

Marketing Academy · Field Work●Growth Marketing

All projects·From: Net Revenue Retention (NRR): The Metric That Decides If Growth Compounds

MiniAudit· 25 minutes

Adyen

Objective: Given a real-shaped 20-account MRR waterfall (starting MRR, expansion, contraction, churn), calculate NRR and GRR and diagnose whether expansion is masking a churn problem.

You're a marketing analyst at Adyen reviewing last year's mid-market account cohort to see whether the reported topline growth is hiding account-level churn.

Use the lesson's NRR/GRR formulas to compute both from a real account waterfall and flag if GRR reveals a problem NRR hides.

Before you start

What you'll need

Free path (everything below is enough to finish)

Google Sheets

FreeCompute NRR/GRR and segment accounts by ACV tier

Free, handles the waterfall formula and segmentation without a BI tool

Google Analytics 4(optional)

FreeCross-check which accounts' usage data explains the churn/contraction rows

Free product-usage view helps validate why specific accounts contracted

No access? Skip and rely on the provided waterfall export alone

Paid upgrades (optional, faster/deeper)

Google Sheets fully covers a 20-account audit; a paid tier only pays off once you're tracking hundreds of accounts continuously.

Mixpanel(optional)

FreemiumBuild a live account-health dashboard segmented by ACV tier

Automates the segment-vs-benchmark comparison at renewal-review scale

The process

2 steps

Step 01 of 02

What NRR Actually Measures

NRR = (Starting MRR + Expansion - Contraction - Churn) / Starting MRR. GRR is the same but excludes expansion.

Starting MRR $500,000, Expansion $95,000, Contraction $18,000, Churn $42,000: what are NRR and GRR?

Google Sheets— Enter the four waterfall inputs and compute both formulas in adjacent cells.

Procedure

Enter starting MRR, expansion, contraction, and churn in separate columns
Compute NRR = (Start + Expansion - Contraction - Churn) / Start
Compute GRR = (Start - Contraction - Churn) / Start

Sample output

Starting MRR: $500,000
Expansion: +$95,000
Contraction: -$18,000
Churn: -$42,000

NRR = (500,000+95,000-18,000-42,000)/500,000 = 107.0%
GRR = (500,000-18,000-42,000)/500,000 = 88.0%

Healthy

NRR and GRR sit close together, meaning expansion isn't doing heavy lifting to cover up churn.

Unhealthy

NRR reads a healthy 107%, but GRR at 88% is right at the lesson's 'watch this' threshold: expansion revenue is masking real churn.

What this means

The 19-point gap between NRR and GRR is the tell: without expansion, this account base would be shrinking, not growing.

So what do I do about it?

Symptom	Action	Effort
NRR alone looks fine but GRR is near or below 90%	Escalate the churn/contraction accounts to customer success for a root-cause review before the next board update	half day

YouYou can do this yourself, no engineering access required.

Step 02 of 02

Current Benchmarks: Where Companies Actually Land in 2026

2026 benchmarks: enterprise SaaS (over $100K ACV) NRR near 118%, SMB (under $25K ACV) near 97%.

This account cohort's average ACV is $180,000. Is a 107% NRR strong, weak, or roughly on-benchmark for its segment?

Google Sheets— Compare the computed 107% NRR against the segment-specific benchmark, not the blended average.

Procedure

Tag each account by ACV tier (enterprise vs. SMB)
Filter to the enterprise-tier accounts only
Compare their NRR against the 118% enterprise benchmark, not the blended 101-106% figure

Sample output

Segment check
  Blended benchmark: 101-106%   (wrong comparison for this cohort)
  Enterprise benchmark: ~118%
  This cohort's NRR: 107%   <- 11 points below its real benchmark

Healthy

The team compares against the enterprise benchmark and correctly flags 107% as underperforming for this segment, not 'above average.'

Unhealthy

Comparing 107% against the blended 101-106% median and concluding the account base is performing well, when against its real segment it's 11 points behind.

What this means

A blended benchmark can make a genuinely underperforming enterprise cohort look fine; segment-matching the benchmark is what makes the number actionable.

So what do I do about it?

Symptom	Action	Effort
NRR is being benchmarked against the blended industry average	Re-segment NRR reporting by ACV tier before the next board deck	30 min

YouYou can do this yourself, no engineering access required.

Final deliverable

A one-page NRR/GRR audit memo: both computed figures, the segment-correct benchmark comparison, and a flag on whether expansion is masking churn.

See a reference example

Sample output

Coinbase, enterprise cohort NRR/GRR audit (excerpt)

NRR: 121%   GRR: 94%
Segment benchmark (enterprise, >$100K ACV): ~118%
Verdict: on-benchmark, and the 27-point NRR/GRR gap is expansion-driven growth, not churn masked by upsells

Success criteria

You're done when you can:

Correctly computes both NRR and GRR from the four waterfall inputs
Uses the segment-specific benchmark (enterprise/SMB), not the blended median
Flags a wide NRR/GRR gap as a potential churn-masking signal