What rules govern these examples?
- Name the process. CNC is not medical.
- Name the earnings base. SDE versus EBITDA is the fork.
- Apply only the matching cell from multiples by sub-sector or the calculator table.
- Move inside the cell with concentration, certificates, backlog, owner dependence, and deferred capex.
- Compare to OLV plus clean NWC. If the floor wins, say so.
BizBuySell’s 2021–2025 manufacturing recap — 2.75x median SDE, 2.04x–3.59x interquartile — is the Main Street climate (Source: Iconic recap). GF Data’s 7.2x average and 5.9x under $25 million EV is a larger-EV climate (Source: Iconic GF Data recap). Stanford’s 6.2x search median sits on a ~25% margin professional-operator file (Source: Stanford GSB). Each example stays in one room.
Example: owner-operated CNC shop under $1M SDE
Label: machining, sde-under-1m, general job shop, no specialty cert.
Facts (illustrative): trailing SDE $380,000. Owner is the estimator. A $140,000 replacement-supervisor add-back is refused because the likely buyer is an owner-operator. Top customer 22%. Iron is mixed 2008–2019 CNC. No written OLV yet; a decent cell list might appraise near $1.05 million OLV. Inventory is clean after a physical.
Cell: general CNC / job shop SDE under $1M = 2.0x–3.6x SDE (BizBuySell manufacturing quartile habit).
Illustrative earnings value: $760,000–$1,368,000 before peg and excess cash. A mid-cell 2.5x–3.0x band is $950,000–$1,140,000.
Floor test: if a written OLV lands at $1.05 million and NWC is normalized, you are in asset-floor territory. Do not invent 5x SDE because someone heard a PE podcast.
Buyer set: individual / SBA, maybe a searcher who will pass once the GM is subtracted. SBA 7(a) capacity is a later structure question, not a reason to change the cell.
Example: medical molder at $3–10M EBITDA
Label: plastics / medical components, ebitda-3-10m, ISO 13485 current.
Facts (illustrative): adjusted EBITDA $4.1 million after a replacement GM. Top customer 19%. No real estate in the deal. Tooling mostly customer-owned. QoE not yet started.
Cell: medical (ISO 13485), EBITDA $3–10M = 6.5x–9.0x.
Illustrative enterprise value: $26.7–$36.9 million before peg and excess cash.
What the fight will be: add-backs, tooling ownership, resin and WIP peg — not whether “manufacturing should be 3x.” GF Data’s 7.2x average can rhyme with the middle of this cell because the size is institutional. It still does not let a commodity printer borrow the row.
Buyer set: searchers at Stanford-like math may compete; add-ons and strategics should be in the process. IBBA Pulse Q1 2026: 83% of deals over $5 million drew three or more offers (Source: IBBA). Run that process. Do not take the first letter.
Example: commodity fab near the asset floor
Label: metal fabrication / commodity stamp, sde-under-1m leaning tired.
Facts (illustrative): trailing SDE $290,000 after refusing heroic add-backs. Top customer 37% on verbal releases. 1990s brakes and a tired laser. Owner wants “plus inventory” at book ($310,000), of which $90,000 is obsolete plate.
Cell: metal fab SDE under $1M = 2.0x–3.5x.
Illustrative earnings value: $580,000–$1,015,000. After writing down obsolete plate, the peg shrinks. Concentration pushes you toward the low half and an earn-out — see customer concentration.
Floor test: suppose written OLV is $1.15 million and clean NWC is $180,000. Asset stack $1.33 million before environmental and rebuild chips. Earnings lose. The honest conversation is an asset deal plus any premium for the remaining crew. Do not split $1.33 million and $700,000 into $1.00 million because the number is round.
Fed G.17 utilization 76.0% (July 2026; Source: Fed G.17) is climate. This shop’s overtime story does not restore a going-concern cell.
| Example | Band × sector | Cell used | Earnings value (illustrative) | Floor note |
|---|---|---|---|---|
| CNC job shop | SDE < $1M × general CNC | 2.0x–3.6x SDE | $0.76–$1.37M | OLV may win |
| Medical molder | EBITDA $3–10M × ISO 13485 | 6.5x–9.0x | $26.7–$36.9M | Earnings win; fight is QoE |
| Commodity fab | SDE < $1M × fab | 2.0x–3.5x SDE | $0.58–$1.02M | OLV plus NWC likely wins |
What these examples refuse to do
They refuse to average SDE and EBITDA. They refuse to give the fab shop the medical row. They refuse to apply GF Data’s 8.8x $100–$250 million observation to any of the three. They refuse to treat CTA’s wider observed SDE band for some cleaner small industrials as a contradiction of the 2.75x median — the median is the middle of a Main Street set; certified small shops can sit higher, still on SDE.
A fourth sketch, packaging, ebitda-3-10m: $6.8 million adjusted EBITDA, no customer above 14%, clean GFSI. Specialty packaging cell 5.5x–7.5x. Illustrative EV $37.4–$51.0 million. A commodity printer with the same EBITDA and a 44% customer does not inherit that cell.
Use the calculator to pick a cell. Use methods for the sequence. Sellers: sell my manufacturing business. Educational information only. Not an appraisal. No value is guaranteed.
A fourth sketch: electronics EMS at $1–3M EBITDA
Label: electronics / EMS, ebitda-1-3m, no medical or aero cert.
Facts (illustrative): $1.7 million adjusted EBITDA after GM. Top customer 27% on cancelable releases. Inventory includes $240,000 of last-time-buy components the customer has not taken. SMT mix is one late line and two older ones.
Cell: electronics / EMS, $1–3M = 4.0x–6.0x.
Illustrative EV: $6.8–$10.2 million before peg. Concentration and E&O push you toward the low half and a named-account earn-out. The inventory write-down happens before the multiple. Do not “plus” the last-time buy.
This sketch exists so electronics files do not borrow the medical row because someone said “high tech.”
A fifth sketch: food co-pack at $3–10M with GFSI current
Label: food / co-pack, ebitda-3-10m, SQF current.
Facts (illustrative): $5.2 million adjusted EBITDA. No customer above 16%. Working capital heavy. Building carved out at market rent. Last GFSI audit clean.
Cell: food / co-pack $3–10M = 5.5x–7.0x.
Illustrative EV: $28.6–$36.4 million before a season-aware peg. The fight is lots, allergens, and the peg — not whether manufacturing “should be 3x.” A lapsed GFSI would leave this row.
What readers should do instead of scaling these dollars
Pick your cell on the calculator. Move inside it with your facts. Commission OLV if the floor might win. Book an intro call if you want the sequence on your binder. Do not email these sketches to a lender as a report. They are teaching tools. Appraisal versus broker opinion is the boundary.
How not to scale these sketches onto a live credit memo
Pick your cell on the calculator page. Move inside it with your facts. Commission OLV if the floor might win. Book an intro call if you want the sequence on your binder. Do not email the CNC, medical, fab, EMS, or food sketches to a lender as a report. They are teaching tools. They refuse blended bases. They refuse borrowed rows. They refuse 8.8x on a job shop. An appraisal versus a broker opinion is still a different document. Copy the discipline, not the dollars.
A reader contract that keeps sketches off credit memos
These files are teaching tools. They name size band, sub-sector, and base before a cell. They test OLV. They refuse blended bases and borrowed rows. They refuse 8.8x on a job shop. Copy the discipline. Do not copy the dollars onto a live memo. Pick your cell on the calculator page. Move inside it with your facts. Commission OLV if the floor might win. Book an intro call if you want the sequence on your binder. An appraisal remains a different document. A broker opinion remains a marketed range. Neither is a sketch on this page.
A closer read on reader contract
These sketches name size band, sub-sector, and base before a cell. They test OLV. They refuse blended bases, borrowed rows, and 8.8x on a job shop. Copy the discipline. Do not copy the dollars onto a live credit memo. Pick your cell on the calculator page. Move inside it with your facts. Commission OLV if the floor might win. Book an intro call if you want the sequence on your binder. An appraisal remains a different document. A broker opinion remains a marketed range. Neither is a sketch on this page. The EMS and food addenda exist so those families do not steal medical or packaging rows.
How to steal the method without stealing the dollars
Each sketch on this page names a size band, a process family, and an earnings base before it names a cell. That order is the only thing a live file should copy. The dollars are teaching furniture. They are not comps. They are not a credit-memo appendix. They are not a partner exhibit. If your plant is a $2.1 million EBITDA precision plastics shop, you do not scale the medical-molder sketch by a ratio. You open the publishing table, sit in the $1–3 million EBITDA precision-plastics cell of 5.0–6.5 times, and move inside that cell with your own concentration, certificates, and iron.
The commodity-fab sketch exists because asset floors win more often than slogans admit. Test orderly liquidation plus clean working capital before you argue about whether 5.0 or 5.5 times is “fair.” If the floor wins, the multiple debate was entertainment. The EMS and food addenda exist so those families stop borrowing medical or packaging rows. Electronics assembly is not ISO 13485. A GFSI food plant is not a converting line.
GF Data’s 8.8 times average at $100–250 million enterprise value is in the research file so readers can see a larger-EV series. It is not a job-shop quote. Stanford’s 6.2 times search-fund median is a different population again. If a sketch on this page ever seems to rhyme with those series, that is coincidence of teaching, not a license to import them.
Commission an orderly-liquidation report if the floor might win. Book an intro call if you want the sequence run on your binder. An appraisal remains a different document. A broker opinion remains a marketed range. Neither is a paragraph you highlight on this page and forward to a lender.