Field note · November 2025
Expansion metrics without vanity MRR
Net revenue retention is allowed to be boring. When it becomes a costume for price changes, Product-Led Growth Measurement has left the building.
A regional SaaS team came to Observatory with a glowing NRR tile. Half of the lift was a list-price change that sales had applied at renewal. The product had not earned another seat. Feature adoption after conversion was flat. Nobody had lied exactly — they had allowed two different stories to share a denominator.
Split the story on purpose
We chart three siblings, never one blended number:
- Seat depth — paid seats in use, not licences purchased and idle.
- Post-conversion adoption — a small set of features that indicate the product is doing the job for the new commercial relationship.
- Price movement — list or discount changes, labelled as commercial, not as product expansion.
If leadership wants a single NRR figure for investors, fine. Put the siblings underneath it. The moment you hide price movement inside “expansion,” you train the company to raise prices instead of earning seats.
Lag is part of the metric
Expansion after a PLG conversion often arrives weeks later. A champion invites two colleagues, then procurement adds seats next quarter. If you celebrate week-one NRR, you are measuring invoices, not product-led behaviour. In Expansion Instrumentation Desk we force a lag note onto every readout: “Do not claim expansion before day 45 unless seat invites are both sent and accepted.”
What we refuse to count
One-time professional services bundled into recurring. Credits that expire. “Success packages” that are really onboarding fees. Currency swings presented as expansion for Thai baht contracts invoiced in USD. Each of these has appeared in student dashboards. Each of them failed critique.
A mild operational cost
Sibling charts annoy sales for one cycle. They then become useful in renewals because the feature adoption view is something a customer success manager can actually act on. That pattern showed up in our regional SaaS case study. The analytics lead still wanted a vendor SQL clinic; we still declined. Expansion measurement is a contract problem first.