Most agency owners discover the valuation conversation has multiple frameworks the moment a buyer's number doesn't match what they expected. The buyer cites an EBITDA multiple. The owner cites a revenue multiple. The owner's accountant cites a third number off the tax return. Three frameworks, three answers, one deal — and the seller who didn't know there were three frameworks usually ends up anchored to whichever one the buyer surfaces first.
The fix is to walk into the conversation already familiar with all four valuation methods, the two profit metrics that ground the most-used one, and the normalization process that produces a defensible number. None of these are hard. They are just not what an agency owner has typically been taught.
Four ways to value the same book.
Every institutional buyer's valuation model fits one of four categories. They aren't mutually exclusive — most sophisticated buyers triangulate across two or three — but each method makes a different assumption about what's being bought, and the seller who understands which method a buyer is leaning on can predict the offer before it lands.
- Income approach (DCF). Project free cash flow for 5–10 years, discount to present value at a cost-of-capital rate. Theoretically rigorous, but the inputs (growth, margin, retention) are negotiable, which makes the output negotiable. Used by buyers with strong financial-modeling shops and by sellers running highly transparent processes.
- Market-based approach. Apply a multiple from comparable transactions. The multiple comes from observed deal data — agency size, line mix, region. Used by virtually every aggregator and PE platform; this is the framework the seller is most likely to encounter.
- Asset-based approach. Value the book as the sum of its identifiable assets (client lists, carrier appointments, equipment) minus liabilities. Rarely used for going-concern agency sales; common in distressed dispositions or when a buyer plans to fully absorb the operation.
- EBITDA multiple (the dominant framework). Multiply a normalized profit metric by a market-supported multiple. This is the canonical model for institutional agency M&A above ~$1M EBITDA, and it is the framework against which the other three are usually reconciled.
The market-based and EBITDA-multiple approaches converge in practice: the "comparable transactions" used in the market approach almost always reduce to an EBITDA multiple. The income approach (DCF) and asset-based approach surface only in specific contexts — DCF for very large or very strategic deals, asset-based for distressed or unusual structures.
Two profit metrics, two answers.
Inside the dominant EBITDA-multiple framework, the profit metric a buyer applies the multiple to is itself a choice. The decision tree is small but consequential.
For owner-operated books under $1M.
- Adds back the full owner's compensation — buyer assumes they're stepping into the operator role.
- Reflects the actual cash a single owner-operator extracts from the business.
- Used by individual buyers, SBA-financed transactions, and books small enough to be a job-and-a-book.
- Multiples typically 2.0× to 3.5× SDE.
For institutional buyers, everything above.
- Adds back only the difference between owner comp and a market-rate replacement.
- Reflects the operating cash flow a hired manager would generate.
- Used by aggregators, PE platforms, and any buyer who isn't planning to operate the book personally.
- Multiples in the 4–6× distressed / 8–10× market / 10–12× competitive / 12–19× kill-zone bands.
The size break is the dividing line. SDE is structurally too kind to small-book sellers (because the full owner-comp add-back inflates the metric); Normalized EBITDA is structurally too harsh on owner-operated micro-books (because the market-rate replacement is large relative to the book). Buyers know which side of the line your book sits on; the seller's job is to know it first.
From tax return to defensible price.
Tax-return EBITDA is the wrong starting number for almost every agency sale. Owners run their books for tax minimization — generous owner compensation, family payroll, personal expenses on the corporate card, lease payments to owner-controlled real estate, one-time legal and consulting fees absorbed as operating expense. Each of those line items legitimately depresses tax-return profit; each of them is also a defensible add-back in normalization.
Buyers categorize add-backs into three families, and a seller's documentation should follow the same structure:
| Add-back family | Examples | Documentation buyer wants |
|---|---|---|
| Owner-specific | Excess owner comp; family payroll without role; personal expenses on corp card; club / travel / car | Payroll register; market-rate benchmark (BLS, Reagan, Big I); expense detail |
| Non-operating | One-time legal on unrelated matters; investment income; discontinued line of business; carrier bonus tied to prior period | Invoices; bank statements; matter-specific detail |
| Non-recurring | CAT-year comp surge; AMS migration consulting; office relocation; one-time hire / severance | Engagement letters; three-year trend showing the line as outlier |
The test buyers apply isn't "can the seller explain this line?" — it's "can the seller prove it with a document a diligence team can audit?" Every add-back that survives that test flows to enterprise value at the full multiple. Every one that doesn't survive becomes a negotiation chip the buyer uses to take the multiple down.
An add-back worth fighting for is an add-back backed by a document a buyer can audit. If your only evidence is your word, you're not normalizing — you're negotiating.
What this means in the real conversation.
The buyer who walks into your first conversation has a method, a metric, and a multiple already chosen. They're going to apply that framework to your numbers and produce an offer. If the offer feels low, it's usually not because the buyer is being aggressive — it's because the seller's numbers aren't yet in the format the buyer's framework reads.
Two practical moves close the gap:
- Decide which framework fits your book before you talk to anyone. Sub-$1M owner-operated book? SDE × 2–3.5×. Above that? Normalized EBITDA × market band. Knowing which conversation you're in lets you set your own anchor instead of receiving the buyer's.
- Build the normalization workbook before LOI, not during diligence. Every add-back you cite at LOI without documentation becomes a retrade target post-LOI. Documented add-backs survive; undocumented add-backs disappear at the closing table — and they take a turn or two of multiple with them.
The full Pillar on this topic — Agency Valuation Methods for Sellers — walks the framework decision and the multiple bands in detail. For the normalization workbook itself, the deepest deep-dive is in the dedicated valuation Tacticals; start with the definitive guide and the Normalized EBITDA deep-dive when you're ready to do the math.