What Percent of ARR Should Come From Expansion?

What Percent of ARR Should Come From Expansion?

Your pricing model sets the answer, and the useful question is whether expansion is adding to a stable base or replacing revenue that left. A seat-based product expands mechanically as customers hire. A usage-based product expands as customers ship volume. A flat-fee platform contract expands only when you release something new to sell.

Those three models produce different expansion shares from identical customer satisfaction, which makes a borrowed percentage useless as a target. Chasing someone else's expansion mix usually means repricing a product to match a business you do not run.

The test that survives across models is directional. Expansion that sits on top of high gross retention is growth. Expansion that sits on top of weak gross retention is a treadmill, and the faster it runs the harder it is to see the erosion underneath.

How Do You Calculate the Expansion Share Correctly?

Read it off a monthly ARR waterfall where each month's beginning ARR equals the prior month's ending ARR. The reconciling requirement is what makes the number trustworthy, since every dollar of movement lands in exactly one bucket and the endpoints have to tie. The distinction in the last three rows is where most reporting goes wrong. Rolling new customer ARR into an expansion figure produces a number that rises whenever new business has a good quarter, which tells you nothing about whether existing customers are buying more.

The second common error is netting. A customer who drops two products and adds one produces churned product ARR and new product ARR in the same month. Netting them to a single line erases the churn, inflates gross retention, and understates how much expansion work it took to stay flat.

When Does a High Expansion Share Become a Warning?

When it appears next to weak gross retention, because the two together mean upsells are funding replacement rather than growth. Gross retention is capped at 100% since it excludes expansion entirely. Net retention includes expansion and can exceed 100%. The gap between them is the amount of cover expansion is providing.

A company reporting 118% net revenue retention sounds healthy until you learn gross retention sits at 82%. That combination says nearly a fifth of the base is disappearing each year and expansion worth 36 points of the base is covering the loss and adding growth on top. The model depends on those accounts continuing to grow, and expansion concentrates in a small number of them.

Concentration is the specific risk. Run the cut before you trust the mix:

- Share of expansion ARR coming from your top ten accounts - Share coming from a single product - Share coming from seat growth versus deliberate upsell motion

If most expansion arrives through automatic seat growth in ten accounts, that is a market condition rather than a sales capability, and it reverses when those customers stop hiring.

Does Expansion Revenue Need Its Own Pipeline?

Yes, and most teams carry no coverage against it. Expansion opportunities have their own creation rate, conversion rate, and cycle length. Treating expansion as something that happens rather than something that gets sold leaves a quarter exposed with no way to detect the gap early.

The standard pipeline coverage discipline applies, with different inputs. Expansion deals usually convert at higher rates than new business, since the buyer already knows the product, which means the coverage ratio required is lower. They also stall in different places, generally around budget ownership rather than vendor selection.

Build the expansion pipeline explicitly. Every account with a known growth trigger, contract anniversary, or unused entitlement is an opportunity that should exist as a record with an owner and a date. Accounts flagged as retention risks should be excluded from expansion targets rather than counted twice, since an at-risk customer is not an upsell candidate.

The earliest read on which accounts belong in which bucket comes from support data. In ORM customer data, an account filing no support cases is at churn risk because nobody is using the product. An account filing seven or more cases in a year is also at risk. Three to five cases a year, typically tier two or tier three severity, is the healthy engaged pattern, and those accounts are the ones worth building expansion pipeline against.

How Do You Forecast Expansion Instead of Assuming It?

Model it on the same cadence as new business, with its own conversion and timing inputs, then reconcile the output against the waterfall. Expansion assumed as a percentage of the base is a plug, and plugs get discovered in the last week of a quarter.

Accuracy on new business plus expansion typically lands around 90% when a team commits real time and effort to producing the forecast. That version has a structural weakness: it is static, so it does not respond as conditions change, and rebuilding it every few weeks consumes the analyst capacity that should be spent on the business. ORM targets 95% accuracy without manual adjustments, holding from day one through day 90 of the quarter and updating as the quarter progresses.

The requirement that makes any of it work is a consistent definition of what counts as expansion. Teams routinely stall here, convinced their CRM data is too messy to model. Everyone has messy data. Garbage in does not have to mean garbage out, because a model can learn from inconsistent data as long as the inconsistency is consistent. What breaks a forecast is a definition that changes mid-year, not a field that was never clean.

Feed expansion into the same model that produces the revenue forecast. New business, expansion, and renewals compete for the same capacity and land in the same board number, and forecasting them in separate spreadsheets guarantees the three will not reconcile.