Investment committee questions to answer before backing a new revenue line

Why answer the IC's questions before they are asked

An investment committee sees a new revenue line inside an existing portfolio company as a risk to the thesis it already approved: it might dilute management focus, create liabilities or blur the exit story. Expect it to test six areas: durability, rights, competitive exposure, management time, exit treatment and reputational or legal risk. A one-page answer sheet that covers each with evidence turns an open-ended debate into a short decision.

This checklist uses licensing operational records to AI developers through SourceX as the worked case, because it touches every area at once: one-time income, data rights, AI-related reputational questions and exit treatment. The same structure works for any new line, from a services add-on to a marketplace listing.

The question bank

1. Durability: will the revenue last, and does it need to?

• Is the revenue recurring or one-time? For a data license: a one-time payment for an agreed dataset, so model it as one-time and keep any further licensing out of the base case.
• What drives demand, and how exposed is it to a shift? For a data license: AI developers are moving from models that answer questions to agents that carry out multi-step work, and training those agents needs records of real work that are thin on the public web. Epoch AI's analysis of human-generated training data projects that, if trends continue, models will fully use the stock of public human-generated text between 2026 and 2032, a forecast with wide uncertainty.
• What ongoing obligations does the line create? For a data license: delivery under agreed redaction rules and compliance with the license terms, which are typically exclusive for AI training for an agreed term. There is no ongoing service team to staff.

2. Rights: does the company own what it would sell?

• Who owns the underlying material? For a data license: work that employees create within the scope of their jobs is generally owned by the employer as a work made for hire, as the Copyright Office's Circular 30 explains. Contractor-created material may need a written assignment.
• Do customer contracts, privacy notices or employee policies restrict use? For a data license: a January 2024 FTC staff post on privacy and confidentiality commitments states that promises not to use customer data for undisclosed purposes, such as training models, are enforceable, so check what the privacy policy and terms actually promised.
• Is regulated data involved? For a data license: health information generally must be de-identified under HIPAA's standard or otherwise authorized; HHS guidance on de-identification describes the two accepted methods. Records that are mainly consumer personal data or protected health information are usually out of scope.
• Who decides what is redacted? For a data license: de-identification and redaction requirements are agreed with the company before any work begins, and nothing is delivered without an executed agreement and the company's authorization.

3. Competitive exposure: could this help a rival or undercut our own plans?

• Could licensed records reveal pricing, customers or methods? For a data license: nothing is binding until the company agrees price and terms, so list sensitive categories, such as customer identities or price files, for the scope and redaction discussion that happens before any work begins.
• Does it conflict with the company's own AI roadmap? For a data license: an exclusive AI-training license for an agreed term limits licensing the same data elsewhere, so clarify how internal use is treated before signing. The guide to the chief AI officer at a PE-backed company covers the internal side.
• For software companies, does it touch product positioning or pricing? The guide on seat compression and AI covers how PE-backed SaaS companies are repricing.

4. Management time: who carries it?

• Who is the internal owner? For a data license: an authorized sponsor such as the owner, CEO or CFO, with an IT or operations lead for the inventory and later exports.
• What does the sponsor team do? For a data license: the operating partner makes one introduction. SourceX runs qualification, inventory, pricing, buyer review, contracting and delivery with the company.
• What does it displace? Show where the initiative sits against the existing value creation plan, and which meeting or project gives way if it goes ahead.

5. Exit treatment: how will a buyer see it?

• How will it appear in a quality of earnings review? Present one-time license income separately from run-rate EBITDA; a buyer will strip it out of the multiple anyway.
• Does an exclusivity term bind a future owner? For a data license: disclose the license in the data room and expect buyers to read its term, scope and assignment clauses. The continuation vehicle diligence questions show how buyers probe management on items like this.
• When is the revenue recognized? Under ASC 606, a license is recognized at a point in time or over time depending on whether the customer receives a right to use or a right to access the intellectual property; Deloitte's roadmap chapter on licenses explains the test. Ask the auditors early.

6. Reputation and legal risk: what could go wrong in public?

• How would employees and customers react if they heard about it? Prepare a short internal note: the company keeps ownership, the data is licensed rather than sold, and redaction rules are agreed before delivery.
• Do lenders need to know? Some credit agreements treat an exclusive license as a disposition or restrict licensing pledged intellectual property; check with lender counsel before signing.
• Does anyone on the sponsor side benefit personally? If an operating partner or the firm would receive a referral reward, disclose it to the IC and compliance, and check the LPA and firm policy first.

This is general information, not legal, tax or financial advice. Confirm with your own counsel, tax adviser or professional body before acting.

How to use the results

Red flags that should end the discussion

• The records mainly belong to the company's clients, as at many agencies and outsourcers, and those clients have not consented.
• The data is mainly consumer personal information with no licensing basis, or protected health information without authorization or de-identification.
• Archives were deleted, or nobody can export the data.
• The same data is already licensed for AI training.
• The company does not reach 50+ full-time employees at peak (contractors excluded).
• The owner will not consider an exclusive license.
• Records were generated with AI in order to sell them.

The full baseline is on who qualifies.

What the IC memo paragraph can say

Next step

If the answers hold up, register as a partner and introduce the company, or have the CFO apply directly at sourcex.si/apply. The page for private equity operating partners shows how sponsors run this across a portfolio.