Buying a company for its records: when the data is the thesis

When is data a real acquisition thesis?#

Data is a real acquisition thesis when the target's records would change what the acquirer can do after close, and when the acquirer may legally use them for that purpose. Examples include a roll-up buying a firm whose job history covers a new region, or a software group buying a product whose support and usage history would strengthen its own.

It is a weak thesis when the records are described in general terms, such as decades of proprietary data, without evidence of what they contain, who owns them and what contracts say about their use. Deal teams should treat data claims like any other synergy: specific, tested in diligence and priced with caution.

The four tests: rights transfer, depth, linkage and refresh#

The four tests are rights transfer, depth, linkage and refresh. Each needs evidence from the target's systems and contracts, not a management presentation, and a failed test should change the price, the structure or the decision to proceed.

Run the tests in that order. Rights failures are the hardest to fix after close, while weak depth or linkage usually just narrows the thesis to the record families that pass.

Rights transfer: promises travel with the data#

Rights transfer is the test most often failed, because the promises a company made when it collected data usually travel with the records. Customer contracts may restrict use to providing the service, privacy policies may limit sharing, and vendor terms may restrict bulk export from the systems holding the history.

Regulators have treated personal data in transactions this way for years. When RadioShack's customer data was proposed for sale in bankruptcy, the director of the FTC's Bureau of Consumer Protection wrote to the consumer privacy ombudsman in May 2015. The letter recommended against selling the data on its own and urged limiting any transfer to a buyer in a substantially similar business that would honor the existing privacy policy and seek consumers' affirmative consent before material changes.

Business records raise different questions from consumer lists, but the lesson holds: the acquirer inherits the commitments, not just the files. Which laws apply is assessed deal by deal with counsel.

Measuring depth, linkage and refresh#

Depth, linkage and refresh are measured from the systems themselves. Ask for a metadata inventory listing each system, its date coverage, approximate volumes and key fields, and review a small sample under NDA where necessary.

These measures line up with drivers in the SourceX Enterprise Data Value Framework, such as scale, recency, data cleanliness and human-generated signal, and with the factors that reduce net value: preparation cost and privacy burden.

• 1Depth: count accessible history by system, not years in business; an old company may hold only recent history after a migration.
• 2Linkage: check that tickets connect to engineering issues, estimates to jobs and invoices, or nonconformances to corrective actions.
• 3Refresh: confirm the business and its systems keep running, so the record set keeps growing and stays recent.
• 4Cleanliness: note duplicates, thin free text and missing fields that raise preparation cost.
• 5Privacy burden: gauge how much personal or confidential detail must be removed before any use.

Paying for data in the price versus licensing it#

Paying for data in the purchase price means the acquirer funds the records up front and must then earn the return through use. ASC 805 has acquirers book an acquired intangible apart from goodwill when it rests on contractual or legal rights or could be separated from the business, and databases appear in the Codification's examples of technology-based intangibles. Data that drives the price can therefore show up in purchase accounting, along with the expectations attached to it; the treatment is settled with accounting advisors.

If the plan is to use the records to run a combined business, buying may be the only route. If the acquirer only needs the records for a defined purpose, such as improving its own product, licensing usually fits better: the current owner keeps ownership, the records are licensed for defined uses, and no acquisition premium is paid for something the buyer does not need to control.

A third case is a thesis that the target's records could themselves be licensed to AI developers after close. There is no price list for operational records and value is known only once a buyer engages, so that income should sit outside the purchase price as upside to be tested, not as a synergy the seller is paid for in advance.

Signals that the data story is weaker than it looks#

Signals that a data thesis is weaker than it looks usually appear early in the data room, well before any rights memo is written. Treat each one as a question for management rather than a verdict.

None of these signals has to end a deal. They move value out of the headline price and into structure, such as an earnout tied to records verified after close, or out of the thesis entirely so the business is priced on its operations.

• 1Management cannot name the systems holding the records or say how far back exports reach.
• 2The data room shows dashboards and screenshots but no export samples or field lists.
• 3Customer contracts were never reviewed for data-use or confidentiality language.
• 4Key history sits in a vendor account owned by a founder or set to lapse at close.
• 5The thesis depends on combining records held by different entities under different promises.

Illustrative: a roll-up weighs a target's dispatch history#

Illustrative: a fictional PE-backed home services platform is considering a regional plumbing company largely for its long dispatch and service history in a new metro area. The deal team runs the four tests before assigning any value to the records.

Rights pass for internal use, but older jobs sit in a retired system with partial exports, so depth is real only for recent years. Linkage is strong: estimates, jobs, invoices and warranty callbacks connect in one system. Refresh depends on keeping the local brand and its dispatchers.

The platform prices the business on its operations alone, treats the records as upside for scheduling and pricing, and leaves any outside use of them to a later, separate rights review.

How SourceX fits a data-driven deal#

SourceX does not buy companies or records; it manages licensing transactions for operating companies that keep ownership. For a sponsor weighing a data thesis, a metadata-only fit check on a portfolio company or a closed acquisition shows whether its records are licensing candidates, without moving files.

If a package proceeds, the SourceX five-step transaction, Supply, Rights, Preparation, Approval and Delivery, applies. Its SourceX Evidence Packet captures provenance, licensing rights, permitted use, the privacy record and release authorization, which is exactly the documentation a future acquirer will ask to see.