How does this calculator work?
There is no form and no engine. The “calculator” is the publishing table from /research/valuation-benchmarks.md, shown in full below, plus the rules that keep people from abusing it.
Step 1. Choose the earnings base. Owner-operated discretionary earnings under about $1 million → SDE column. Institutional files after a replacement manager → an EBITDA column. See SDE versus EBITDA.
Step 2. Choose the row that matches the process, not the NAICS code you like. Multiples by sub-sector explains why aerospace and medical sit high and why commodity print sits low.
Step 3. Read the cell as a range. Midpoint is not a quote. Outside the cell needs an auditable reason — scarce certificate, proprietary product, or an OLV floor.
Step 4. Move inside the cell with concentration, backlog quality, certificates, owner dependence, equipment age, union withdrawal liability, and environmental RECs. Those are broker rules of thumb, labeled as such in the research file.
Step 5. Compare earnings value to OLV plus normalized NWC minus corrections. If the floor wins, stop arguing a turn.
This is educational copy. It is not a certified appraisal. Jason Taken is a business broker. Appraisal versus broker opinion is the boundary page.
What size-band × sub-sector ranges are sourced?
Label for every cell below: advisor-survey / reported-deal ranges, not an appraisal.
Synthesized from BizBuySell manufacturing 2021–2025 (median 2.75x SDE; IQR 2.04x–3.59x), IBBA / M&A Source Market Pulse Q1 2026 (advisor-survey LMM plots; $5–$50M sparkline near 4.5x–5.1x EBITDA), GF Data Q1 2026 manufacturing recap (7.2x average; 5.9x under $25M EV; 8.8x at $100–$250M EV), CTA’s wider observed SDE band for some cleaner small industrials, and Pepperdine 2025 as a secondary PE-style cite. If a cell cannot be tied to a source, the research file says to label it broker rule-of-thumb. The grid below is the site publishing table.
SDE under $1 million (owner-operated)
| Owner-operated process family | SDE range this worksheet uses | Why that cell exists |
|---|---|---|
| General CNC / job shop without a specialty cert | 2.0x–3.6x SDE | BizBuySell manufacturing quartiles, 2021–2025 |
| Commodity metal fabrication | 2.0x–3.5x SDE | Same Main Street set, discounted versus certified shops |
| Specialty machining, medical, or aero parts still under $1M SDE | 3.0x–5.0x SDE | Wider observed CTA/IBBA synthesis; still SDE, still small |
Adjusted EBITDA, going-concern, cash-free / debt-free
| Earnings column | Commodity fab / print family | Precision / plastics family | Aero AS9100 / medical ISO 13485 family |
|---|---|---|---|
| $1–3 million EBITDA | 4.0x–5.5x | 5.0x–6.5x | 6.0x–8.0x |
| $3–10 million EBITDA | 5.0x–6.5x | 5.5x–7.5x | 6.5x–9.0x |
| $10 million-plus EBITDA | 6.0x–8.0x | 6.5x–9.0x | 7.0x–11.0x |
Calculator worksheet by size band (same sourced cells, different layout)
Each cell below is an advisor-survey / reported-deal range, not an appraisal. Pick one band. Do not blend columns.
Owner-operated SDE under $1 million
| Process family on this worksheet | SDE times to use |
|---|---|
| General CNC and job-shop work | 2.0x to 3.6x SDE |
| Commodity metal fab and stamp | 2.0x to 3.5x SDE |
| Injection molding and extrusion | 2.2x to 3.8x SDE |
| Electronics EMS assembly | 2.0x to 3.5x SDE |
| AS9100 aerospace and defense | 3.0x to 5.0x SDE |
| ISO 13485 medical device | 3.0x to 5.0x SDE |
| IATF automotive programs | 2.0x to 3.4x SDE |
| Food co-pack | 2.2x to 3.6x SDE |
| Specialty packaging | 2.2x to 3.8x SDE |
| Commodity print | 2.0x to 3.2x SDE |
| Wood and millwork | 2.0x to 3.6x SDE |
| Chemicals and coatings | 2.2x to 3.8x SDE |
| Tooling, dies, and molds | 2.4x to 4.0x SDE |
| Building products | 2.0x to 3.6x SDE |
| Textiles and sewn goods | 2.0x to 3.2x SDE |
| Machinery OEMs | 2.2x to 3.8x SDE |
Adjusted EBITDA $1–3 million
| Process family on this worksheet | EBITDA times to use |
|---|---|
| General CNC and job-shop work | 4.0x to 6.0x EBITDA |
| Commodity metal fab and stamp | 4.0x to 5.5x EBITDA |
| Injection molding and extrusion | 4.5x to 6.5x EBITDA |
| Electronics EMS assembly | 4.0x to 6.0x EBITDA |
| AS9100 aerospace and defense | 6.0x to 8.0x EBITDA |
| ISO 13485 medical device | 6.0x to 8.0x EBITDA |
| IATF automotive programs | 4.0x to 5.5x EBITDA |
| Food co-pack | 4.5x to 6.0x EBITDA |
| Specialty packaging | 4.5x to 6.5x EBITDA |
| Commodity print | 4.0x to 5.0x EBITDA |
| Wood and millwork | 4.0x to 6.0x EBITDA |
| Chemicals and coatings | 4.5x to 6.5x EBITDA |
| Tooling, dies, and molds | 5.0x to 7.0x EBITDA |
| Building products | 4.5x to 6.0x EBITDA |
| Textiles and sewn goods | 4.0x to 5.5x EBITDA |
| Machinery OEMs | 4.5x to 6.5x EBITDA |
Adjusted EBITDA $3–10 million
| Process family on this worksheet | EBITDA times to use |
|---|---|
| General CNC and job-shop work | 5.0x to 6.5x EBITDA |
| Commodity metal fab and stamp | 5.0x to 6.5x EBITDA |
| Injection molding and extrusion | 5.5x to 7.5x EBITDA |
| Electronics EMS assembly | 5.0x to 7.0x EBITDA |
| AS9100 aerospace and defense | 6.5x to 9.0x EBITDA |
| ISO 13485 medical device | 6.5x to 9.0x EBITDA |
| IATF automotive programs | 5.0x to 6.5x EBITDA |
| Food co-pack | 5.5x to 7.0x EBITDA |
| Specialty packaging | 5.5x to 7.5x EBITDA |
| Commodity print | 4.5x to 6.0x EBITDA |
| Wood and millwork | 5.0x to 6.5x EBITDA |
| Chemicals and coatings | 5.5x to 7.5x EBITDA |
| Tooling, dies, and molds | 5.5x to 7.5x EBITDA |
| Building products | 5.5x to 7.0x EBITDA |
| Textiles and sewn goods | 4.5x to 6.0x EBITDA |
| Machinery OEMs | 5.5x to 7.5x EBITDA |
Adjusted EBITDA $10 million and above
| Process family on this worksheet | EBITDA times to use |
|---|---|
| General CNC and job-shop work | 6.0x to 8.0x EBITDA |
| Commodity metal fab and stamp | 6.0x to 8.0x EBITDA |
| Injection molding and extrusion | 6.5x to 9.0x EBITDA |
| Electronics EMS assembly | 6.0x to 8.5x EBITDA |
| AS9100 aerospace and defense | 7.0x to 11.0x EBITDA |
| ISO 13485 medical device | 7.0x to 11.0x EBITDA |
| IATF automotive programs | 6.0x to 8.0x EBITDA |
| Food co-pack | 6.5x to 8.5x EBITDA |
| Specialty packaging | 6.5x to 9.0x EBITDA |
| Commodity print | 5.5x to 7.0x EBITDA |
| Wood and millwork | 6.0x to 8.0x EBITDA |
| Chemicals and coatings | 6.5x to 9.0x EBITDA |
| Tooling, dies, and molds | 6.5x to 9.0x EBITDA |
| Building products | 6.5x to 8.5x EBITDA |
| Textiles and sewn goods | 5.5x to 7.5x EBITDA |
| Machinery OEMs | 6.5x to 9.0x EBITDA |

What the table will not do
It will not output a single “manufacturing multiple.” It will not accept a form. It will not add book inventory on top of enterprise value. It will not put the building inside the cell. It will not treat GF Data’s 8.8x $100–$250 million observation as a job-shop quote. It will not treat BizBuySell’s 223-day 2025 manufacturing median close as a price (Source: BizBuySell year-in-review).
IBBA Pulse Q1 2026: 67% of advisors saw no material AI effect on value; 83% of deals over $5 million drew three or more offers (Source: IBBA). Those facts are process climate. They are not a cell.
How should you move inside a cell?
Start at the midpoint only as a teaching habit, then argue with evidence. A 16% top customer, current AS9100, and a 2018+ CNC mix sit high in the aerospace row. A 41% customer, lapsed ISO, and 1990s manuals sit low — or they fail the going-concern test and fall to OLV. Worked examples walk three files. Methods is the sequence. EBITDA by size is the column essay.
If your process is a mix — 20% medical, 80% commodity fab — do not take the medical column. Weight the conversation. If your certificate lapsed, leave the certified row.
Disclaimer and next step
Disclaimer. The ranges on this page are advisor-survey / reported-deal ranges, not an appraisal. They are synthesized from the sources listed above and from /research/valuation-benchmarks.md. They are educational. They are not a quote on your plant. They are not a lender value, a court value, an ESOP value, or a divorce value. HedgeStone Business Advisors and Jason Taken provide broker opinions of value for marketed sales. Jason Taken is a business broker, not an appraiser, not an attorney, and not a CPA. No value is guaranteed. Confirm with M&A counsel, a CPA, and qualified appraisers.
The conversion path is an on-site intro call at /book via the button on this page — not a third-party calendar URL in the body. Sellers who want a marketed range continue at sell my manufacturing business. The valuation pillar remains the long essay.
Spec: inputs a live tool would take if one ever ships
If a later engineering pass adds a widget, the inputs should be: sub-sector row, earnings base (SDE or adjusted EBITDA), earnings dollars, and four toggles — top-customer band, certificate status, owner-dependence flag, and whether a written OLV exceeds a mid-cell earnings value. The output should be a range, a label that says advisor-survey / reported-deal, not an appraisal, and a sentence that says Jason Taken is a business broker. The output should refuse to run if the user tries to apply an EBITDA column to a sub-$1M SDE file without subtracting a replacement.
Until that widget exists, this page is the spec and the table. Do not paste a spreadsheet on a website and call it a calculator. Do not hide a form. The conversion path is the intro call.
Spec: what a live tool must never do
Never average rows. Never apply GF Data’s 8.8x bracket to a Main Street input. Never add book inventory on top of EV. Never put an off-site calendar URL in the body. Never call the output an appraisal. Never store a plant’s financials in a public form. Never name PE firms. Never promise a close date.
Those “never” lines are why this page is copy. They are easier to enforce in prose than in a half-built script.
How to use the table in a partner conversation without turning it into a court exhibit
Show the cell. Show the chips. Agree that a buy-sell that names an appraiser still needs the appraiser. Use the table as shared vocabulary, not as the number in the agreement. Buyouts, estates, and divorce is the assignment page. Printing this table does not satisfy it.
How partners should use this table without turning it into an exhibit
Show the cell. Show the chips. Agree that a buy-sell that names an appraiser still needs the appraiser. Shared vocabulary is the useful output. A printed table is not a court number, not a lender number, and not an estate number. If a later widget ever ships, it must refuse to run an EBITDA column on a sub-1 million SDE file without a replacement, and it must label every output as advisor-survey and reported-deal, not an appraisal. Until then, this page is the spec. There is no form. The next step is the intro call.
Educational information only. Not tax, legal, or engineering advice.
A partner-use rule that keeps this table from becoming a decree
Show the cell. Show the chips. Agree that a buy-sell naming an appraiser still needs the appraiser. Shared vocabulary is the useful output. A printed table is not a court number, not a lender number, and not an estate number. If a later widget ever ships, it must refuse to run an EBITDA column on a sub-1 million SDE file without a replacement, and it must label every output as advisor-survey and reported-deal, not an appraisal. Until then this page is copy plus a spec. There is no form. The next step is the on-site intro call.
A closer read on partner use
Show the cell. Show the chips. Agree that a buy-sell naming an appraiser still needs the appraiser. Shared vocabulary is the useful output. A printed table is not a court number, not a lender number, and not an estate number. If a later widget ever ships, it must refuse to run an EBITDA column on a sub-1 million SDE file without a replacement, and it must label every output as advisor-survey and reported-deal, not an appraisal. Until then this page is copy plus a spec. There is no form. The next step is the on-site intro call. Do not hide a spreadsheet and call it a calculator.
What to tell a partner who wants a single number from this table
Tell them the table is an advisor-survey and reported-deal range, not an appraisal. Tell them to pick the size-band row and the process-family column that match the plant they actually own. Tell them the midpoint of a cell is not a quote. If they still want a single number for a decree or a return, they need a credentialed appraiser and a named standard of value. This page will not become that document by being printed.