Your Net Dollar Retention Is a Lie. Here's the Metric That Actually Predicts Churn.
Tomasz Tunguz analyzed 374 quarterly NDR observations from 25 public software companies. The trend is clear: NDR is declining everywhere. But the real problem isn't the decline — it's that NDR was always a vanity metric masking the health signal that actually matters.
By Nina Okafor, Marketing Ops · Oct 14, 2025
NDR is declining across 25 public software companies. Tomasz Tunguz's data shows the metric was always a vanity signal. Here's what actually predicts churn and how to measure it.
Frequently Asked Questions
What is Net Dollar Retention (NDR)?
Net Dollar Retention measures how much revenue existing customers generate over time compared to the previous period. An NDR of 120% means existing customers are spending 20% more than they did a year ago. An NDR below 100% means existing customers are spending less — through downgrades, seat reductions, or cancellations. Historically, 'best-in-class' SaaS companies maintained NDR above 130%. As of 2026, NDR is declining across the industry, with many companies falling below 110%.
Why is NDR declining across SaaS companies?
NDR is declining for three structural reasons: (1) AI is reducing seat counts — companies need fewer human employees for tasks that software supported, which means fewer seats purchased, (2) platform consolidation — companies are consolidating from multiple point solutions to fewer platforms, reducing spend per vendor, (3) procurement sophistication — enterprise procurement teams are actively auditing and renegotiating software contracts, eliminating unused licenses and downgrading plans.
What metric should replace NDR?
Workflow Dependency Depth (WDD) measures how many daily operational decisions flow through your product. Unlike NDR, which is a lagging financial indicator, WDD is a leading indicator of retention because it measures how embedded your product is in the customer's actual work. A product with high WDD is practically impossible to remove, regardless of seat count changes. Products with low WDD — tools that are used occasionally or for non-critical tasks — are the first to be cut.
How do you calculate Workflow Dependency Depth?
WDD is calculated by measuring: (1) the number of unique daily active workflows that touch your product, (2) the percentage of those workflows where your product is the system of record (data originates in your product), (3) the number of downstream systems that depend on data from your product. A high WDD score means the product is deeply embedded in daily operations with multiple downstream dependencies. The metric can be implemented through product analytics by tracking workflow initiation events, data export events, and API integration usage.
Is NDR still a useful metric for SaaS companies?
NDR remains useful as a financial reporting metric — it accurately describes revenue trends from existing customers. But it should not be used as a health indicator or predictive metric for retention. The problem: NDR is a lagging indicator that tells you what already happened. By the time NDR declines, the underlying causes (reduced usage, workflow displacement, seat compression) have been building for months. Leading indicators like Workflow Dependency Depth, daily active workflow count, and integration density provide earlier warning signals.
Related Articles
Topics: SaaS Metrics, Product Management, Data Analysis, Churn
Browse all articles | About Signal