Deal economics
Go/no-go criteria for a data licensing opportunity
By SourceX Editorial · Reviewed by Noah Loul ·
Short answer
Make a data licensing go/no-go call by scoring a specific opportunity on five criteria: rights to license the records, effort to export and prepare them, brand and customer risk, evidence of buyer interest, and the control you keep over use. A blocked score on rights or control is a no-go for the affected records, however strong the other scores look.
Key takeaways
- Score the opportunity in front of you, a named record set and a named use, not your company's data in general.
- Rights and control are gating criteria: a blocked score on either is a no-go regardless of the rest.
- Buyer interest counts only when it is specific: a record type, an intended use and a willingness to review terms.
- Re-score at each step, because rights findings and term sheets change the answer.
- Every no-go should name what would change it, so the opportunity can be revisited later.
What should a go/no-go decision cover?#
A go/no-go decision on a data licensing opportunity covers one specific package: a defined set of records, a buyer's intended use and the terms on the table. Scoring a whole company's data in the abstract produces a vague yes; scoring a package produces a decision someone can act on.
The decision also covers the next commitment, not the final signature. The first go approves a rights review and an inventory; a later go approves preparation work; the last approves release. Each step spends more staff time and advisory cost, so each deserves its own check.
The five-criteria scorecard#
The five-criteria scorecard rates each criterion as clear, workable or blocked, using evidence rather than instinct. Clear means no open issue; workable means a known fix with an owner; blocked means no practical fix exists.
Score each criterion separately before discussing the whole. Teams that start with an overall feeling tend to explain away a weak rights or control score because the buyer seems keen.
| Criterion | Question to answer | Evidence to collect | Blocked when |
|---|---|---|---|
| Rights | Can the company license these records for this use? | Customer and vendor contracts, notices, ownership of the records | Contracts prohibit the use and consent is impractical |
| Effort | Can the team export, review and prepare the package? | Export routes, history available, redaction scope, named staff | No one can do the work without harming operations |
| Brand and customer risk | Would customers or staff be surprised or harmed? | Record content, customer expectations, the buyer's intended use | The use would damage key relationships |
| Buyer interest | Is there a specific, credible request? | Record type requested, intended use, readiness to review terms | Interest is general, with no defined use or terms |
| Control | Do the terms keep the company in charge of its records? | Exclusivity, field of use, deletion, audit, ownership of improvements | Terms transfer ownership or bar internal use |
Which findings are hard stops?#
Hard stops are findings that no amount of effort or price can fix, and they should end the opportunity or remove the affected records from the package. Write them down before the first buyer conversation so enthusiasm does not soften them later.
A hard stop on part of the archive is not a hard stop on the opportunity. Removing the affected record family and re-scoring the rest is often the right move.
- Customer or client contracts forbid the use, and consent cannot realistically be obtained.
- The value of the records depends on personal, health or financial details that preparation would have to remove.
- The records are a client's deliverables, customer-owned designs or export-controlled technical data.
- No one with authority will sign, or a required consent from a lender, board or investor is refused.
- The buyer requires ownership of the records, perpetual exclusivity over core records or uncapped liability.
- The buyer's intended use would compete directly with the company's own products or services.
How to read a mixed scorecard#
A mixed scorecard is read with a fixed rule set, so the decision does not depend on who argues hardest in the room. Rights and control act as gates, effort and brand risk can usually be fixed by scoping, and buyer interest decides how much to spend on the next step.
Write the rule set into the memo template before scoring. A rule agreed in advance is much harder to bend once a buyer is waiting for an answer.
| Score pattern | Decision | Next step |
|---|---|---|
| All five clear | Go | Approve the next step and name its owner |
| Rights or control blocked | No-go for the affected records | Remove them and re-score, or stop |
| Effort or brand risk blocked, rights and control clear | Go only after rescoping | Narrow the record set or the use, then re-score |
| Buyer interest workable, everything else clear | Go to low-cost steps only | Run the metadata fit check and hold preparation spending |
| Several workable scores, none blocked | Go with conditions | List each fix, its owner and the step by which it must close |
How to score effort and brand risk honestly#
Effort is scored honestly by naming the people and systems involved, not by estimating hours in the abstract. List who will run exports from each system, who will check records for confidential material and who will answer the buyer's questions, then ask their managers what will slip while they do it.
Brand risk is scored by testing a plain description of the deal on the people who matter most. If you could not explain the license in a few plain sentences to your largest client or your senior staff without discomfort, the score is not clear, whatever the contract says.
Both scores improve with scoping. Narrowing a package to internal review comments and scheduling records, for example, can turn a blocked effort score into a workable one and remove most of the brand question.
Illustrative: an engineering firm scores a request for project records#
Illustrative: a fictional civil and structural engineering firm keeps project financials and staffing in Deltek, RFIs and submittals in Procore and review markups in Bluebeam. A model developer asks about records that show how engineers answer RFIs and review submittals.
Rights score blocked for drawings and calculations, because many are client deliverables, but workable for internal review comments and RFI responses once client names and project identifiers are removed. Effort scores workable, with the project controls lead running exports. Brand risk scores clear after the principals agree the package excludes anything a client would recognize.
Buyer interest scores clear because the request names a record type and an evaluation use. Control scores workable: the draft asks for exclusivity across all project records, and the firm counters with non-exclusive terms limited to the reviewed set. The firm proceeds on the narrowed package and records the drawings as a standing no-go.
Recording the decision and revisiting a no-go#
A go/no-go decision should be recorded in a short memo that names the package, the five scores, the hard stops checked and the person who approved. Boards and lenders may ask for this trail later, and it keeps a decision from being reopened informally.
A no-go memo should also name what would change the answer: a contract renewal with different terms, a system retirement that frees an archive, or a buyer request for a narrower use. Opportunities that fail usually fail on one criterion, and that criterion can move.
How the scorecard maps to the SourceX five-step transaction#
The scorecard maps to the SourceX five-step transaction, so each criterion is re-checked at the step where new facts appear. The Supply check works from metadata alone, such as system names and years of history, and no files change hands during that initial assessment; from there, no step proceeds without the supplier's approval.
Each go decision leaves a written trail in the SourceX Evidence Packet, which documents provenance, the licensing rights relied on, permitted use, the privacy record and the release authorization.
| Step | Go/no-go question | Criteria tested |
|---|---|---|
| Supply | Is there a specific package worth scoping? | Buyer interest, effort |
| Rights | Can the company license this package for this use? | Rights, brand and customer risk |
| Preparation | Can personal and confidential details be removed without destroying the value? | Effort, rights |
| Approval | Do the final terms keep the company in control? | Control, brand and customer risk |
| Delivery | Is the release authorized and the package exactly as approved? | Control |
Frequently asked questions
Who should make the go/no-go call?
The chief executive or owner, with general counsel or outside counsel on rights and the COO or CTO on effort. Where a sponsor owns the company, its consent may be needed too, and the governing documents may reserve the decision for the board. The person who will sign the license should be part of the decision from the first step.
Do we need a price before deciding to go?
Not for the early steps. A price exists only once a buyer has seen a described package, and there is no published price list for these records. Early go decisions approve scoping and rights work; the final go weighs the actual offer, net of costs, against the effort and risk scored earlier.
What if only part of the records pass?
Proceed with the part that passes if it still forms a coherent package. Partial packages are common: drawings out, review comments in; customer pricing out, exception notes in. Record the excluded part as a standing no-go with its reason so no one adds it back without review.
How much information do we need for the first score?
Metadata is enough: which systems hold the records, roughly how many years can be exported, what the records contain and which contracts might restrict them. No files or samples are needed for the first score, and nothing has to leave the company to complete it.
Should we tell the buyer why we said no?
A short, factual reason usually helps, such as noting that the requested records include client deliverables the company cannot license. It can prompt the buyer to propose a narrower package that does pass. Keep contract details, customer names and internal scores out of the explanation, since confidentiality obligations still apply.
Related resources
See if your company qualifies
A short company assessment. No data uploads are needed.