Mainferret. ← Home

Methodology

How Mainferret works — and what it can't tell you.

We built Mainferret on a simple conviction: the best acquisition decisions come from evidence, not anecdotes. Here's exactly where our data comes from, how we score it, and — just as important — its limits. We'd rather you trust us because we showed our work.

The data

The record has no stable business ID, so we roll loans up to businesses by normalized borrower name, state, and ZIP. That yields 1,643,950 distinct businesses. Name changes and relocations will split or merge some of them — matching is probabilistic, and we say so.

The succession signal

The core insight: an owner's tenure is the strongest public predictor of seller-readiness. Owners are typically 35–50 at their first SBA loan, so someone who borrowed 15–30 years ago is plausibly 55–75 today — often having bought the business with that very loan. We compute tenure from loan vintage and rank owners by how likely they are to be ready for a conversation, long before they ever list with a broker.

Sell-readiness score (v0)

The current scorer is a transparent, weighted heuristic — version zero, and we label it that way. Six signals, all computed from the loan record, sum to a 0–100 score:

Hard exclusions: any distressed loan (charge-off, liquidation, delinquency, deferment), or a business first financed less than four years ago. A "prospect" is a surviving business scoring 90 or above — the bar where long tenure and a live debt-maturity window coincide. On the current build that's 133,874 businesses (8.1% of all businesses in the record), holding $77.2B of SBA-financed capital. None of them are listed for sale anywhere.

Known v0 limits, stated plainly: founding year is proxied by the first SBA loan (the business may be older); scheduled maturity ignores refis and prepayments except where a paid-in-full date exists; "jobs supported" is a lender estimate at approval; and none of this observes owner intent. The score tells you where to look. Verifying a specific owner is the concierge work.

Risk & survival intelligence

Because the record carries the final status of millions of loans, we can measure realized charge-off (default) rates by industry and business profile — so you can tell a durable trade from a landmine before you spend a dollar. These rates are computed from resolved loans in the same FOIA record, and every report states the cohort they come from.

The buy-vs-build numbers

The headline stat on the site (acquisitions charge off at 5.4% vs 14.6% for startups) is computed as follows.

What this is not

Sources

U.S. SBA 7(a) & 504 FOIA datasets (data.sba.gov), files as of March 31, 2026 — the source of every figure on this site. Enrichment for concierge work: Google Places API, state registration and licensing records, and publicly available web and court records. SBA refreshes the FOIA files quarterly and we rebuild on each refresh.