Revenue / base / 6 to 12 months
For the customers a business already had at the start of a period, their recurring revenue is shrinking rather than holding or growing. AIFO looks only at that existing group and asks whether the money kept and expanded from them outweighs what was lost to downgrades and cancellations. Customers won during the period are deliberately excluded, because new logos can mask an existing base that is leaking. This is a labeled proxy: it can be measured only when revenue can be traced to individual customers over time, and where the books cannot do that, the signal reports not applicable rather than guessing.
A business can post healthy top-line growth while its existing customers quietly erode underneath, and only a measure that isolates the existing base exposes that leak. Bessemer Venture Partners and the KeyBanc private-company survey treat net retention as a core efficiency screen for recurring-revenue businesses, because keeping and expanding existing customers is far cheaper than constantly replacing them. When the existing base shrinks, the company has to win new customers just to stand still, which is an expensive and fragile way to grow.
New sales look fine, but a closer look shows existing accounts downgrading or leaving at a pace that offsets the wins. The revenue chart holds up while the makeup of it churns underneath. Renewals and expansions feel harder to come by than they used to.
Separate revenue from existing customers from revenue from new ones, so you can see whether your base is actually holding. Find whether the leak is customers leaving outright or staying and spending less, because those are different problems with different fixes, and focus retention effort on the accounts most at risk. Treat expanding and keeping existing customers as seriously as winning new ones, since a leaking base undermines every new sale.
Pinecrest SaaS, an illustrative company whose revenue can be traced to individual customers, started the year with about 1,000,000 of recurring revenue from its existing customer base.
Over the year that same starting group expanded their spending by about 90,000 but gave back about 210,000 through downgrades and cancellations, so the base retained roughly 880,000 of the original 1,000,000. New customers won during the year are set aside on purpose, because the question is whether the existing base held. Keeping only part of what it started with puts the base below the break-even line where expansion offsets losses.
Pinecrest concludes its existing base is leaking and that new sales are masking erosion it needs to address directly.
Exact thresholds, formulas, and severity bands are omitted from this public view. The full methodology, with every figure, is available in the authenticated app.
This is a labeled proxy: it needs recurring revenue attributed to individual customers over time, which a general ledger does not record. Where the ingestion cannot attribute revenue per customer, the signal is reported as not applicable rather than estimated, so the proxy label is not cosmetic.
The exact thresholds, the formula, and your own figures are in the authenticated app. See it against a real, simulated dataset for Integra Executive Services, our public demo company, or connect your own QuickBooks Online account.