The exit data book: how KPI preparation also makes a company licensing-ready

What an exit data book is

An exit data book is the sell-side collection of KPI data, reconciled to the financial statements, that sits behind the confidential information memorandum. Where the CIM tells the story, the data book lets a buyer rebuild it: revenue by customer and month, cohorts, margins by service line, pipeline, pricing and headcount, usually in spreadsheet form with definitions attached.

It is distinct from the quality of earnings report, which normalizes earnings, and from vendor due diligence, which is a written third-party report. The comparison of vendor due diligence and a data licensing review sets out where each starts and stops.

How buyers test KPI quality

Buyers and their advisers test three things: does each KPI reconcile to the ledger, has its definition stayed the same over time, and can the company show where each number came from. A data book that fails any of these invites a retrade.

A sound build tends to follow five passes:

1. Agree KPI definitions with the CFO and bankers, then freeze them.
2. Map each KPI to its source system, report or table, data owner and date range.
3. Extract and reconcile to the general ledger month by month, logging every manual adjustment.
4. Run the analyses buyers will run, such as customer concentration, cohort decay and price-volume splits, and fix what breaks.
5. Lock a version for the data room and keep the extraction logic and logs for follow-up questions.

Pass two is where licensing readiness starts.

The shared source map: one artifact, two uses

The source map built in pass two lists systems, date ranges and owners. A data licensing review needs the same three columns, plus two more.

The two extra columns are rights (who created the records, and what contracts, privacy notices and employee policies allow) and sensitivity (personal data, health data, client-confidential material). Adding them while the map is open is far cheaper than rebuilding it later.

What a licensing review needs that a data book does not

A data book is built from aggregates drawn mostly from finance systems. AI developers license something different: the underlying records of work.

• Operational systems: email, Teams or Slack, shared drives, ticketing, engineering and project tools rarely feed a data book but hold most licensable records.
• Record-level history: a ticket thread with its resolution, a quote with its revisions and outcome, a project file with its change orders.
• Rights review: whether the company created the material and may license it.
• Exportability: whether whole records, not just summaries, can be exported from each system, including retired ones.

Strong companies often run 10-15+ systems, and a data book typically draws on only a few of them. That gap is why an operating partner should widen the source map rather than assume the data book already covers it.

Why it matters to a portfolio

For a sponsor, one round of preparation can support two outcomes: a cleaner sale and, for companies that qualify, a one-time license payment. Timing is a choice. A license completed before marketing gives buyers a documented asset and a closed agreement to diligence; a license signed after closing is the new owner's decision. Under longer holds the window for a pre-exit license widens; see value creation during longer hold periods.

Sectors with deep operating records, such as HVAC and plumbing roll-ups, IT services and professional services, are where widening the source map pays off most.

How a data license should appear in the data book

Present license proceeds as non-recurring. A one-time payment for an agreed dataset should not sit in run-rate revenue or adjusted EBITDA, and a quality of earnings provider will strip it out anyway. Disclose the license's exclusivity and term in the data room.

Revenue recognition depends on the terms. Under ASC 606, an entity assesses whether a license gives the customer a right to access its intellectual property over the license period, recognized over time, or a right to use it as it exists when granted, recognized at a point in time (Deloitte Revenue Recognition Roadmap 12.4). The company's auditors decide how a specific data license is recorded, so ask them before the data book is locked.

This is general information, not legal, tax or financial advice. Confirm with your auditors and counsel before acting.

What it means for a referral partner

Operating partners watch data books being built across the portfolio, and the moment the source map exists is the cheapest time to run a licensing screen. To qualify, the company needs to be US-based, with 50+ full-time employees at peak (contractors excluded), years of operations captured in its systems, clear rights to license its records and an owner or executive able to sponsor the deal. A first pass through the company fit checker is non-binding and needs no contact details, and the first-month referral plan for operating partners shows how to fit introductions into an existing portfolio calendar.

Partners earn 25% of the eligible platform fees SourceX actually collects from the referred company's licensing deals, capped at $100,000 per referred company. Rewards are paid only after the buyer pays and SourceX receives its fee, and no reward is guaranteed.

Limits and open questions

• A finance-led data book covers only a fraction of a company's licensable records.
• Aggregated KPIs and the data book itself are sale materials, not licensable datasets.
• Rights and privacy questions sit outside the scope of most data books.
• Revenue recognition and covenant treatment of a license depend on its terms and the company's agreements.
• Running a license and a sale process at the same time needs coordination with the deal team and counsel.

The definition of exit readiness puts the data book in the wider context of sale preparation.

Next step

When the next portfolio company starts its data book, add rights and sensitivity columns to the source map and run the fit screen. Then register as a partner to introduce companies that pass, and see referral opportunities for operating partners for the wider playbook.