How do you evaluate a manufacturing business before you write an LOI?
Buying a manufacturing business starts with a written buybox. Evaluation is the test of whether a specific plant matches that box. How it works is the process spine. This page is the kill-sheet order.
Write the buybox before you request a teaser
A usable buybox is a one-page filter, not a wish list:
- Process you will actually run (CNC, fab, molding, EMS, aero, medical, food — pick one)
- Earnings definition and a number (SDE or EBITDA, not “profitable”)
- Radius you will live in, and a yes/no on moving iron
- Certificates that must be live on day one
- Highest top-account share you will underwrite without extra structure
- Union / open-shop preference
- Dirt: lease only, buy the plant, or either
- Processes you will refuse (plating, ITAR, automotive releases are common refusals)
HedgeStone will not burn a seller’s time by circulating every plant inside a 400-mile circle. If the filter is blank, you are shopping buildings. You are not evaluating a company.
The eight-item first pass
- Three earnings stories. Tax, internal P&L, and job-cost. When they conflict, budget a quality of earnings before you exclusive. That fee is part of evaluation.
- Who pays the rent. A single account above roughly a quarter to a third of sales is a structure problem. A “backlog” that is only a forecast tab is a sales slide.
- Iron and liens. Match serials and controls. Ask what crashed and what was rebuilt. Sanity-check asking EV against orderly liquidation before you argue growth.
- The cage. Separate raw, WIP, and finished. Age the lots. Confirm who owns tools and consigned steel.
- The four names. Who programs, who quotes, who runs nights, who holds the top accounts. A missing cross-training sheet is a finding.
- The wall hanging. Certificate expiry, last certification-body visit, open majors, and what the customer portal actually shows.
- Dirt and chemistry. Order a Phase I when you take title or when solvents, tanks, or finishing are in the history. ASTM E1527-21 is the AAI standard (40 C.F.R. Part 312).
- Paper that may not travel. FAR 42.12 novation on government work. ITAR registration on defense. Set-aside status that can die on sale.
The full checklist expands each room. Evaluation is whether you keep reading.
| Signal on the first pass | Keep reading | Stop or restructure |
|---|---|---|
| Earnings basis | Tax, P&L, and job-cost reconcile within a story you can test | Three ledgers, three religions |
| Multiple vs size band | SDE shop in 2.0x–3.6x; LMM file in the matching EBITDA cell | 6.2x pasted onto a $400k-SDE job shop |
| Financing | Project fits $5M 7(a) or a named above-cap stack | "We'll figure out the loan after LOI" |
| Peg | Method written: AR + inventory − AP after E&O | "Customary NWC" with no WIP rule |
| People | Second chair named | Owner is estimator, salesperson, and programmer |
| Access | Live certificate, last audit in the room | Expired ISO sold as paperwork |
Main Street manufacturing in BizBuySell's 2021–2025 set cleared at 2.75x median SDE (IQR 2.04x–3.59x) on about $268k SDE (Source: BizBuySell via Iconic). Stanford's $16 million / 6.2x / ~25% margin median is a search-fund profile (Source: Stanford GSB). Evaluation includes refusing to mix those cells.
Price, peg, and whether the stack exists
SBA 7(a) maximum is $5 million. SOP 50 10 8 requires a 10% injection; a seller note is equity only on full standby for the life of the 7(a) and only up to half the injection (Source: SBA). If the project does not fit, evaluation includes saying so before you fall in love with the crane.
The working-capital peg is part of price. Manufacturing hides dollars in unbilled WIP and percentage-of-completion. If the seller will not discuss method, you cannot evaluate enterprise value. You can only evaluate a rumor.
When OLV of the iron plus cleaned net working capital approaches asking EV, you are evaluating an asset floor, not a growth story. That is still a purchase. It is a different purchase.
Climate stays in the margin, not in the teaser
ISM Manufacturing PMI was 54.6 in August 2026 with a Prices Index of 71.1 (Source: ISM). Fed G.17 put manufacturing capacity utilization at 76.0% in July 2026 (Source: Fed G.17). Those series tell you whether pass-through and capacity are live issues. They do not underwrite the target's top customer.
Deloitte / Manufacturing Institute: as many as 3.8 million employees needed 2024–2033, 1.9 million potentially unfilled (Source: Deloitte/MI). Evaluation of labor is names and a hiring plan, not a national headline.
IBBA Q1 2026: 83% of deals over $5 million had three or more offers (Source: IBBA). Competitive tension is a reason to be ready, not a reason to skip the first pass.
After the paper survives
Walk the plant only after the paper survives. The plant-visit protocol is the etiquette page. On this evaluation pass you are checking that the money cell is on the floor, the crane actually covers that bay, the CMM still holds a current cert, and the cage you counted on paper matches what you can see. Do not pull operators aside without the seller’s script.
Red flags that should stop a multiple argument: a perpetual that cannot produce a physical, a certificate that lapsed last spring, a lien column that is blank, “we have always handled waste this way,” and government work treated as if it assigns itself.
BizBuySell 2025 put manufacturing at a 223-day median close (Source: BizBuySell). Evaluation includes whether you can live on that calendar.
How to score a teaser in fifteen minutes
A teaser that omits size band, sub-sector, and whether real estate is in is not a teaser. It is bait. Write four numbers on a pad: stated earnings and whether they are SDE or EBITDA; top-customer share if shown; asking EV; and whether the stack can be 7(a), 504, or above-cap. If asking EV implies a 6.2x on a file that cannot be a ~25% margin search-fund median, you already know the answer (Source: Stanford GSB). If it implies 2.5x SDE on a certified medical cell, you may have a process — after paper.
Teasers that say "highly profitable manufacturing company" and nothing else waste seller goodwill when you request the room. HedgeStone will not send every plant in a 400-mile radius. Your fifteen-minute score is how you stay off that list of tourists.
Geography is an evaluation item. You will rarely relocate a plant. If the buybox said 150 miles and the teaser is 400, kill it unless the capability is scarce. Census CBP 2023's state establishment counts (California 35,300; Texas 21,034; and the rest) describe where shops live (Source: Census CBP). They do not make a distant file local.
When to spend money on the next report
After the eight-item pass survives, spend in this order: UCC and corporate good-standing; machine-list gap analysis; inventory sample; then QoE / appraisal / Phase I as the LOI scopes them. Do not buy a Phase II because you are nervous. Buy it because the Phase I gave you a REC. Do not buy a PE-style QoE on a $220k SDE file unless add-backs are the whole story.
SBA 7(a) still caps at $5 million with a 10% injection (Source: SBA). Evaluation includes whether the project, fees, and peg fit. Combined 7(a)+504 can reach $10 million (Source: SBA). Occupancy still applies. If the teaser needs $12 million of seller proceeds on a thin check, evaluation is "no" — not "we'll engineer it."
IBBA Q1 2026: 83% of deals over $5 million had three or more offers (Source: IBBA). Evaluation speed matters in that band. Speed without the eight items is how you join a process you should have skipped. Main Street 2.75x SDE files (Source: BizBuySell via Iconic) still deserve the same eight items. Small is not a waiver.
ISM PMI 54.6 and Prices 71.1 (Source: ISM) belong in the margin comment, not in the kill decision, unless the shop has no pass-through history. Fed utilization 76.0% (Source: Fed G.17) belongs in the capacity comment, not as a reason to buy.
Keep / kill / structure — three boxes only
Keep: paper reconciles enough to test, multiple in the right size-band cell, stack exists, second chair named or stay fundable, certs live or recert funded. Kill: three ledgers, 6.2x on a job shop, no-cash SOP folklore, expired access cert, environmental shrug, missing money-cell serial. Structure: 25–30%+ customer, deferred capex you can see, peg method fight you can freeze.
SBA 7(a) $5 million / 10% injection (Source: SBA) sits in the stack box. Combined 7(a)+504 can reach $10 million (Source: SBA). Stanford $16 million / 6.2x / ~25% margin (Source: Stanford GSB) sits in the kill box when misapplied — it is never a job-shop multiple. Main Street 2.75x SDE (Source: BizBuySell via Iconic) sits in the keep box when the SDE is real. IBBA 83% three-offer (Source: IBBA) sits in the speed box, not the waiver box. ISM 54.6 / 71.1 (Source: ISM) and Fed 76.0% (Source: Fed G.17) sit in comments.
Jason Taken is a business broker. Educational information only. Confirm legal, tax, environmental, and lending questions with licensed people. No outcome is guaranteed.
Questions a teaser must answer in fifteen minutes or die
What is the earnings basis. What size band. What sub-sector. Is dirt in. What implied multiple. Does a stack exist. If the teaser cannot answer those, do not request the room. You will waste seller goodwill and your own calendar.
Geography is an evaluation item. You will rarely relocate a plant. If the buybox said 150 miles and the teaser is 400, kill it unless the capability is scarce. Spend money on reports in order: UCC, machine-list gaps, inventory sample, then QoE / appraisal / Phase I as the LOI scopes them. Do not buy a Phase II because you are nervous.
Keep, kill, or structure. Those are the only three boxes. Enthusiasm is not a fourth box.