Multi-year contracts break the standard churn calculation. Divide lost ARR by total ARR and every account sitting in year two of a three year term counts in the denominator without ever having had the chance to leave. Reported churn falls, the retention slide improves, and none of the improvement reflects a customer decision.
Use available to renew, not total ARR
The renewal performance question is narrow. Of the revenue that reached a renewal decision this period, how much came back. That means the denominator is available to renew, which includes only contracts with an expiry date inside the window, plus any early renewal or renegotiation that was pulled into the period.
Total ARR still belongs in the reporting, just under a different label. Gross revenue retention measured against beginning ARR captures losses from accounts that were nowhere near an expiry date, and it is the number an investor will compare across companies. Publish both, label both, and never let the friendlier one stand in for the other.
The cliff is the real risk
A book weighted toward three year terms produces two or three quiet years followed by a year where a large share of the base decides at once. Two things go wrong in that year. The renewal team is staffed for the quiet run rate and suddenly faces several times its normal volume, and the annual retention number becomes hostage to a handful of large accounts.
The fix is scheduling rather than heroics. An available to renew schedule extended 24 to 36 months forward turns the cliff into a planning input. It tells you which quarter needs extra coverage, which accounts carry outsized weight in that quarter, and how early the renewal conversation has to open to leave room for a save.
Contract structure distorts the numerator too
Multi-year paper usually carries mechanics that look like retention events but are not.
- Contracted escalators raise ARR at the anniversary without anyone selling anything. - Ramped deals that start small and step up read as expansion in year two and three. - Co-terminated add-ons shorten the effective term of a product bought mid-contract. - Termination for convenience clauses allow a loss that no renewal schedule predicted.
Split the ARR waterfall so contracted movement and earned movement sit on separate lines. Otherwise a strong looking net revenue retention number turns out to be paperwork signed two years ago, and the customer success investment that appears to be working is doing less than the chart suggests.
What to report
Report churn on available to renew for renewal execution, gross revenue retention on beginning ARR for base erosion, and weighted average remaining contract term for exposure. The third number is the one most teams skip, and it is the one that tells you how much of the base is actually up for debate in the next twelve months. Feed all three into the revenue forecast rather than a single blended retention rate, because a blended rate applied to a book with lumpy expiry dates will be wrong in both directions in the same year.
Frequently Asked Questions
Do multi-year contracts lower churn or just delay it?
They delay it and they change its shape. A three year term removes two renewal decisions from the calendar, so reported churn falls while the contracts are running and then concentrates in the year the cohort comes due. The customer still decides, just less often and with more revenue riding on each decision.
What denominator should a multi-year book use?
Revenue available to renew in the period, for the renewal performance number. Report gross revenue retention against total beginning ARR alongside it, because that version captures mid-contract losses that the available to renew base never sees.
Should a contracted price uplift in year two count as expansion?
Decide once and apply it everywhere. A pre-negotiated escalator is contracted revenue rather than a new sale, so most teams book it separately from earned expansion. Mixing the two makes net revenue retention look like a customer success result when part of it was signed years earlier.
How do you spot a renewal cliff before it arrives?
Build an available to renew schedule that runs 24 to 36 months forward and read the ARR expiring in each quarter. A cohort of three year deals signed in one strong bookings year will show up as a single oversized bar, usually far enough ahead that the renewal motion can start early.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like churn rate for multi-year contracts into prescriptive action for your team.
Schedule a Demo