The seller's revenue the book won't support

Every agency acquisition comes with a revenue figure from the seller and a multiple the seller would like applied to it. The multiple gets the attention in the room. The revenue figure is where the money actually goes: apply 2.9× to revenue the book does not support and the gap is paid for nearly three times over.

This post follows an illustrative agency — Ironwood Risk Partners, a construction and energy services agency in Fort Worth, invented for the demo book — from the book export to the investment-committee memo. The revenue engine finds that the book supports 8.2% less revenue than the seller states, and that 17.1% of what it does support is contingent or supplemental income. Every figure here is generated demo data, shown on the running product.

Use case card

Actors	Deal lead (owns the acquisition), analyst, practice leader (signs off what is outside policy), the investment committee
Trigger	The seller's book export imported; Analyze run at analysis and again at IC memo
Agency Illustrative	Ironwood Risk Partners — 3,232 book rows, seller's revenue $3.44M, asking 2.9× revenue
Agents involved	Revenue quality analyst, retention analyst, supervisor agent, valuation memo writer
Controls that fire	`REVENUE_SHORTFALL` (book more than 5% below the seller's — refer), `VOLATILE_REVENUE` (contingent and supplemental over 15% — warn)
Decisions and authority	Approving the findings for the memo (deal lead), the practice leader's sign-off on the referral, the price (the committee)
Outcome	A memo that shows the asking multiple on the seller's revenue, on the book's revenue and on recurring revenue, side by side
What is measured	Book revenue vs seller's revenue; recurring share; the price the multiple implies on each base

Why the seller's number and the book's number differ

Neither side is usually lying. The seller's figure comes from the agency's accounts: commission received in a financial year, with contingent commission, supplemental income and fees in it, and sometimes a business that has since left. The book export is the policies themselves. InsightRollup computes revenue from the book — commission and fees written in the twelve months to the export date, never premium and never a half-finished calendar year — and sets it next to the seller's figure.

Source of difference	What it looks like in the book	Who sees it
Contingent and supplemental income	Rows with commission type `contingent` or `supplemental`: paid by carriers on profit and volume, so it moves year to year	Revenue quality analyst
Lost clients still in the accounts	Clients who were due to renew in the last twelve months and did not — revenue that is not coming back	Retention analyst
Different periods	A financial year against twelve rolling months to the export date	The engines (every citation states its period)
Rows that cannot be counted	Commission missing or above the premium; rows with no client	Book intake agent

flowchart LR S["Seller's revenue
$3.44M"] --> G{"Book vs seller"} B["Book revenue
$3.15M, 12 months
to the export date"] --> G G -->|"-8.2%"| R["REVENUE_SHORTFALL
refer to the practice leader"] B --> Q["Recurring 82.9%
contingent + supplemental 17.1%"] Q --> V["VOLATILE_REVENUE
warn: value it separately"] R --> M["Memo shows the multiple
on each base"] V --> M style R fill:#fdecec,stroke:#e57373 style V fill:#fff4e2,stroke:#f4a03a

What the screens show

The Acquisition Workspace for an illustrative agency at IC memo — The Acquisition Workspace for the illustrative Ironwood Risk Partners at IC memo: the stage strip, the headline figures from the book, and the first decision the agents prepared, with Accept, Reject and Ask.

The workspace hero puts the headline figures in one row: book revenue $3.2M, client retention 90.7%, organic growth 3.6%, top producer 26.1% of revenue, top carrier 29.3% of premium, 21.2% of revenue renewing in 90 days. The rail shows the key facts — seller's revenue $3.4M, asking 2.9×, book against seller −8.2% — and the four controls that fired.

The key facts and acquisition controls rail, on illustrative data — The rail: key facts (seller's revenue, asking multiple, book revenue, book against seller) and the acquisition controls that fired, with who each refers to.

What the revenue is made of, and data quality, on illustrative data — What the revenue is made of — renewal, new, supplemental, contingent — beside the rows the seller should correct.

The revenue mix card is the second half of the story. Of the $3.15M the book supports, renewal commission is 73.8% and new business 9.1%; supplemental income is 8.9% and contingent commission 8.2%. That 17.1% depends on the carriers' profit-sharing and volume bonuses, and the VOLATILE_REVENUE control says to value it separately from recurring commission.

Step by step

From the book export to the memo Ironwood Risk Partners, illustrative. Days are from the import.

Book export importedImportDay 0

ironwood_book_export.csv: 3,232 rows mapped onto one schema, each keeping its row number.

Revenue computed and comparedrevenue_quality_agentDay 0

The engine: $3.15M of commission and fees written in the twelve months to the export date, 8.2% below the seller's $3.44M; 82.9% recurring. The agent's finding cites 1,079 rows.

Controls firesupervisor_agentDay 0

REVENUE_SHORTFALL refers to the practice leader; VOLATILE_REVENUE warns. The supervisor agent prepares the practice leader's sign-off.

Findings approved for the memoDeal leadDay 2

The deal lead reads each finding against its rows and approves it for the memo. Only approved findings reach the memo.

Seller asked to reconcileDeal leadDay 3

The gap goes back to the seller as a question: which accounts carry the difference, and how much of it is contingent income from which carriers.

Memo draftedmemo_agentDay 5

The memo opens with the valuation arithmetic: 2.9× on the seller's revenue is $9.97M; the same multiple on the book's recurring revenue is $7.59M.

Practice leader signs offPractice leaderDay 6

The sign-off on the referral is answered on the decision record, with a note. Signing cannot be recorded while it waits.

The arithmetic, as the memo shows it

Base	Revenue	At 2.9×
The seller's revenue	$3.44M	$9.97M
The book's revenue (12 months to the export date)	$3.15M	— (shown in the memo header)
The book's recurring revenue	$2.6M	$7.59M

The memo does not propose a price. It shows what each base implies, cites the rows, and leaves the number to the committee. All figures above are the illustrative book's.

The valuation-support memo, on illustrative data — The memo: the valuation arithmetic on top, then each section built from approved findings, every figure with the book rows behind it.

Who talks to whom

sequenceDiagram participant Seller participant Lead as Deal lead participant RU as InsightRollup participant Ag as Agents participant PL as Practice leader Seller->>Lead: Book export, seller's revenue, asking multiple Lead->>RU: Import and Analyze RU->>Ag: Engine facts with citations Ag-->>Lead: Findings and prepared decisions RU-->>PL: REVENUE_SHORTFALL referral (prepared sign-off) Lead->>Seller: Reconcile the difference (sent by a person) Lead->>RU: Approve findings for the memo Ag-->>Lead: Memo drafted from approved findings PL->>RU: Sign-off, with a note (decision record)

What the agents prepared, and what each rests on

Agent	What it prepared	What it cites	Who decides
Revenue quality analyst	The book supports $3.15M of revenue, 8.2% below the seller's $3.44M; 82.9% recurring, 17.1% contingent or supplemental	1,079 book rows written in the 12 months to the export date; rows by commission type	Deal lead approves for the memo
Retention analyst	Client retention 90.7%, revenue retention 89.0%: 460 of 507 clients due to renew did; $269.94K left with the 47 that didn't	2,098 rows across two rolling periods	Deal lead
Supervisor agent	Practice leader sign-off: two controls fired that refer	The acquisition controls and the engine figures they read	Practice leader
Valuation memo writer	The valuation-support memo from the approved findings	Approved findings only; the engines' figures quoted	Deal lead finalizes

On the decision record

Section	Hit	Outcome
Revenue	REVENUE_SHORTFALL: book revenue against the seller's −8.2%	refer
Revenue mix	VOLATILE_REVENUE: contingent and supplemental 17.1%	warn
Key people	KEY_PRODUCER: top producer 26.1% of revenue	refer
Carriers	UNAPPOINTED_CARRIERS: 58.7% of premium not fully appointed	warn

For the deal team

You stop rebuilding revenue in a spreadsheet. The gap is on the screen with its rows, and the question for the seller writes itself.

For IT

Revenue is computed in pandas from the stored rows; each figure carries a citation (file, row count, filter). The model never produces it.

For corp-dev leadership

The committee sees the multiple applied to three bases, and who signed off the referral, before it agrees a price.

Questions we get asked

Is the seller's figure wrong?

Not necessarily. It is usually a different period, with contingent income and fees in it. The point is to see the difference and ask about it before the price is agreed.