“Am I overpaying?” is the question that sits under every deal review. You can read the CIM twice, model the debt service, and grill the broker, and still have no real answer to whether the ask itself is sane.
Take an HVAC company listed at $1.8M on $410k of cash flow. That's 4.4x. Fair? High? I've been searching for about a year and until recently I couldn't have told you with any confidence.
Asking-price gravity
Here's why the question is so hard to shake. The asking price is the first number you see, and it anchors everything that comes after it. Offer 10% under the ask and you feel like a shark. Offer 30% under and you feel unreasonable, even when the math says you're being generous.
Sellers know this. Brokers definitely know this. The ask is a negotiating position dressed up as a fact, and most buyers negotiate down from it because they have nothing to negotiate up from.
The comp data you can buy today
The standard answer is a sold-comp database. DealStats runs $1,050 a year. PeerComps runs $600. They're built from closed transactions, which sounds like the gold standard until you filter down to your industry and size band and see how old and thin the sample gets. (BizBuySell will also sell you a $59.95 report, which is about as deep as $59.95 suggests.)
Meanwhile, Searcher OS watches hundreds of broker sites every day. That work exists to feed deal flow, but a side effect is a comp pool most buyers never see: roughly 45,000 live and recent listings that each carry a real asking price and real reported cash flow. We throw out every listing where either number was estimated, and every duplicate. What's left is asking-side data, and asking-side data honestly labeled turns out to be a weapon.
What a Pricing Report does
Every report compares your deal against that pool through a tiered comp lens. For the HVAC deal above, the primary lens is “HVAC and mechanical services, $1M to $2M,” which currently holds 139 comps. You get the quartiles, and you get a histogram showing exactly where the ask sits in the distribution. If your deal is at the 88th percentile of its peer group, you see that bar sticking out to the right.
The report opens with a plain-English executive summary: where the ask sits, the top flags, and the number to anchor a negotiation on. The full detail is underneath if you want to dig.
The real estate finding
The thing that surprised me when I pulled the corpus apart: whether real estate is included in the asking price moves the multiple more than any industry distinction does. RE-included listings in a segment and size band run around 6x cash flow. The same segment and size without the building runs around 3.7x.
That gap is bigger than the difference between an HVAC shop and a plumbing shop. I suspect most buyers have never seen the split because nobody's comp set is big enough to compute it. So every comp set in the report is calculated both ways, and your deal gets matched to its own side. A comp set that mixes the two hands you a number that looks precise and means nothing.
The rest of the report
Live named comps. The 10 to 15 closest lookalikes, with title, state, ask, cash flow, multiple, and a link to the actual listing. Sold-comp databases structurally can't show you this because the transactions are confidential. Our comps are public listings, so we can point right at them.
Days on market and delisting velocity. How long lookalike deals sit before they leave the market. A delisting can mean sold, and it can also mean the seller gave up, so the report labels it as velocity and leaves it at that.
An SBA financeability read. The report computes the maximum price a bank would realistically finance: the highest number that clears a 1.25 debt service coverage ratio at 10% down, 10.5% interest, over 10 years. When that lands below the ask, you've found the most defensible anchor in the whole negotiation. “The math a lender will run says this business supports $1.5M of price” is hard for a broker to argue with.
Triangulation against published sold-side data. The report quotes IBBA Market Pulse, the BizBuySell Insight report, and SBA 7(a) loan data next to our asking-side numbers, with each figure labeled as asking or sold. Asking multiples run higher than sold multiples, and that spread is itself information, so we print both and say which is which.
Why you can trust the numbers
Every figure in a Pricing Report is computed from the data with boring, deterministic SQL. The AI writes the commentary and only the commentary. Then a validator reads the finished narrative and checks every dollar figure and every multiple against the computed stats. If it finds a number it can't trace, the report fails and regenerates.
I built it that way because I've read too many AI-written analyses that state invented numbers with total confidence. An untraceable number is worse than no number, because it anchors you anyway.
The paragraph you send the broker
The practical payoff is negotiating ammunition. Here's the shape it takes:
“Across 139 comparable HVAC and mechanical listings between $1M and $2M, the median ask is 3.5x cash flow. This listing is asking 4.4x. I'm interested, and I'd like to understand what supports the premium.”
Nobody has to get emotional. The data does the arguing, and the broker either has a good answer or starts moving toward the median. (Sometimes they do have a good answer. That's useful too.)
Where to find it
Reports live at /reports, and every listing page has a Get Pricing Report button. Trial accounts get 1 report a month, Searcher plans get 3, and Deal Team gets 10. Free accounts get 1 to try.
Run one on the deal you're closest to. The histogram answers the 11pm question in about 5 seconds, and sometimes the answer is that the ask is fine. Worth knowing either way.