What is SaaS sprawl and how to measure it

Finance calls the moment SaaS spending crosses the “don’t look” threshold differently than IT does. The CFO wants a baseline they can defend at the next operating review; IT wants a per-license posture they can patch on Monday. Both are right, and the reconciliation is the work — converting the feeling that “we have too many tools” into four numbers you can write back to the spend source.

WealthBuilt treats SaaS sprawl as the gap between contractually committed licenses and the licenses currently in use by an active identity. Below is the working definition finance and IT can sign off on together — and the four measurements that turn a sprawl claim into a quarterly reclaim.

A conventional spend audit reads the contract and the active seat list. The two reconcile within ten percent of annual spend, and the team walks away thinking there is nothing to reclaim. What the audit misses is the overhang: licenses contracted under one of three patterns that bill cleanly but never get exercised on a recurring basis.

First, duplicate licenses — two diagram suites, two whiteboarding products, two obvious targets of overlapping canvas primitives. Second, dormant licenses — seats that were active at renewal and slipped below a 30-day last-login threshold inside a quarter. Third, the AI-vendor tail — usage-priced tools whose per-token or per-call invoices have grown quarter-over-quarter without a corresponding usage-tier reset. Each pattern has its own negotiation shape and its own draft.

WealthBuilt posts one figure per quarter — the “named reclaim against a named contract, on a named date” — that finance can write back to the spend source and IT can defend in an architecture review. Underneath that one figure sit four numbers worth tracking month-over-month:

• Active licenses vs contracted licenses — the simplest sprawl proxy. Divide the active distinct users from identity by the contracted seat count per line item; anything below 70% is a reclaim candidate.
• 30-day inactive seats — distinct licenses with no login, no API call, and no sustained background activity in the trailing 30 days. The most defensible to renegotiate because the data is exact, not estimated.
• Per-vendor spend growth over four quarters — a line item that grew over 25% quarter-over-quarter for two consecutive quarters is suspect; usage-priced vendors are the worst offender.
• Duplicate-capability coverage — the count of products whose canvas, storage, or messaging primitives overlap at the capability level. Two overlapping products is a single consolidated line item; three or more is a renegotiation triage.

WealthBuilt pulls four signals over a one-week window: last-90-day daily-active from identity, last-30-day last-login per license, four-quarter invoice history per vendor, and a capability taxonomy against the active license list. The four signals collapse into one intake — a single connector chain — and emerge as four per-line metrics you can write back to the spend source.

The measurement window is short on purpose. A quarterly sprawl measurement is the cadence finance can defend (monthly costs the team too much reconciliation) and long enough that any reclaim candidate has at least one usage data point to argue from. The agent can also run a one-off measurement in the days before a renewal — that’s the cadence that catches the per-vendor growth number before it bills again.

Run this against your spend before the next renewal negotiation. Every checked row is a reclaim candidate with a named dollar and a named contract.

1. Connect identity (Okta, Azure AD, or Google Workspace) and one spend source (card, invoice inbox, or ERP export).
2. Reconcile active licenses against contracted licenses per line item; flag anything below 70% utilization.
3. Run a 30-day last-login heatmap; count seats with no login, no API call, no background activity.
4. Read four quarters of invoice history per vendor; flag line items with over 25% quarter-over-quarter growth for two consecutive quarters.
5. Compare the license list against a capability taxonomy; flag overlapping canvas, storage, or messaging primitives.
6. Write each flagged line item back to the spend source as a named reclaim against a named contract, with a target dollar.

A sprawl programme that lives in dashboards never produces reclaim. The audit must end at the contract and the spend source, with a named line item on a named date. That is the only shape that survives the next renewal call: the finance lead can point at the dollar, IT can point at the data, and the vendor sees both before they take a position.

WealthBuilt posts one figure per quarter — “the named reclaim against the named contract, on the named date” — the same posture that finishes the audit. It is the only measurement loop that converts a feeling about SaaS spend into a defended line item inside a quarter.