Inside The Prospect: How Outposts Get Found and Diamonds Get Ranked

I pulled 4 months of scraper data, roughly 68,800 listings across 300 broker sites, to test one hunch: that a whole tier of deals goes live every week on small websites and never reaches BizBuySell or any aggregator. The hunch held. 149 of those 300 sites qualified, and about 90% of the deals on them appeared nowhere else in the dataset.

That finding is the reason The Prospect exists, and it's the part of the methodology I most want to pull the back open and show you. The Prospect is a free weekly email. Every Thursday a script I wrote queries 7 days of fresh listings, runs them through a set of hard gates and a scoring formula, and drops a draft in front of me to review before it goes out.

Most "deal of the week" emails are a broker pushing whatever they want to move. This one starts from a database watching the whole market and works backward to a handful of deals worth your attention. Here's the machinery.

Two sections, on purpose

The email has two parts, and they run in a deliberate order.

Outposts come first. An Outpost is a small regional broker site carrying fewer than 20 listings that doesn't syndicate anywhere. The deal lives on the firm's own website, goes out to the firm's own email list, and that's it. It never reaches BizBuySell or any aggregator.

Back to that 4 months of data (the 68,800 listings). Of the 300 broker sites I watch, 149 qualified as Outposts. Together they generated 1,153 deals, and about 90% of those appeared nowhere else in the dataset. That 90% held whether I drew the line at 10, 20, or 30 listings per site, so it isn't an artifact of how I defined "small." Non-syndication is a structural property of these sites. It works out to roughly 60 businesses a week going live on exactly one small website.

The surprise: median asking price on Outpost listings runs about 50% higher than on the big aggregators. I expected micro-deals. I got mid-market businesses whose owners just chose not to syndicate.

So Outposts lead the email. A technically public listing that almost nobody is systematically watching is closer to a pocket listing than a BizBuySell post, so it sits up top where it belongs.

Diamonds come second. This section drops the off-market constraint and asks a simpler question: across every source I scrape, aggregators included, what are the best 3 deals in the country this week by the numbers?

How a deal survives the gates

Before anything gets scored, it has to clear a set of hard filters. These exist because the scraper is noisy and most listings are a no before you ever read them. A few of the gates:

• Scraper age has to be at least 7 days. When I onboard a new broker site, it backfills every existing listing at once, so the "new this week" count is a pile of false positives until the scraper has been running a while.
• Title blacklist. Anything tagged under contract, sold, pending, withdrawn, fire sale, cash only, distressed, or the various done for you / turnkey Amazon FBA flavors gets cut.
• Industry downweight. Gas stations, convenience stores, nail and hair salons, dispensaries, donut and ice cream shops. The categories I'd personally pass on without reading the CIM.
• Franchise-route penalty. FedEx routes, UPS Store, Matco, Snap-on. Commoditized, and you're really buying a job with a logo on it.
• Never-repeat dedup. Once a deal has been featured, its ID is excluded from every future run. I'm not recycling last month's picks on you.

There's one gate I'm a little proud of, because it came from getting it wrong. For Outposts specifically, before a deal gets called off-market, the system checks every other source for a fuzzy title match at a similar price (within about 20%). I added that after a Landscape Architecture listing slipped through that was sitting on both businessexits.com and BizBuySell with an identical title. It wasn't off-market at all. Now if a supposed Outpost deal turns up somewhere else, it gets moved to the Diamonds pool (where aggregators are fair game) instead of being sold to you as hidden inventory.

The Diamond score

Here's where I have to be precise, because the exact weights matter and a vague description would be worse than none.

Diamonds get a composite score. The single heaviest factor is the acquisition multiple, and it pushes the other way: a higher multiple drags the score down hard. Cheap relative to cash flow is the thing the formula rewards most. Margin stacks on top of that, so a higher cash-flow margin lifts the score. Then a deal earns points for a high quality signal, for being SBA pre-approved, for recurring revenue, for an absentee owner, with smaller bumps for a team already in place, multiple locations, or real estate included. Coming from one of my proprietary direct-broker scrapes (rather than an aggregator) earns a small edge too.

Then the penalties bite. Red flags carry the steepest deduction in the whole formula, heavier than almost any positive signal can offset. Industry downweight and the franchise-route penalty subtract on top of that.

I'll be honest about the state of it. The multiple coefficient and the margin coefficient are first-draft weights. I tune them based on which deals bubble to the top each week and whether they're the ones I'd flag as worth a real look. The formula is my own screening intuition written into math, and it moves.

The score isn't the final word, either. The top 10 by score go through an editorial pass where the descriptions get read and the junk that scores well on paper but stinks in the prose gets cut: partial buy-ins dressed up as acquisitions, hidden category risk, schemes the structured fields don't catch. The final 3 get picked from what's left with an eye toward industry and geographic spread, so you're not getting 3 HVAC companies in Florida.

I'll grant that the Diamond half is the less novel one. A lot of those signals are what an experienced searcher already carries in their head: cheap relative to cash flow, clean recurring revenue, no red flags. What the formula buys you is that same instinct applied to every listing the scraper touched this week, instead of the 20 sites you had time to open.

Why I built it this way

The acquisition funnel is brutal on volume. The benchmarks Athena Simpson at AcquiMatch shared on stage this spring put it at roughly 28,000 listings scanned to land a single close, with about 4 to 5% of what you scan even worth pulling into a pipeline. A committed searcher can realistically monitor 15 to 20 broker sites by hand. The other 280 go unwatched, the same way they always have.

The Prospect is one narrow attempt at that problem. It won't source your whole search (it's a few picks a week, not a feed). But it surfaces the tier most buyers never see, scored against signals I'd use myself, with the obvious garbage already stripped out.

A few of the picks have genuinely surprised me, and that's the part I like. The system finds deals I wouldn't have, on sites I'd never have thought to open.

The Prospect goes out every Thursday, and it's free.