Twelve AI questions operating partners should ask portfolio company CEOs
Ask a portfolio company CEO twelve questions in three groups: adoption (where AI is used and what it has changed), risk (vendors, customer promises, disruption and governance) and assets (which records the company holds, how far back they go and who owns them). The asset questions show whether the company could license its records without anyone sharing a file.
Why a fixed set of AI questions helps
CEO answers on AI drift toward demos and optimism unless the questions pin them down. A fixed list, asked the same way at every company, lets an operating partner compare answers across the portfolio, see where pilots are stalling and find assets nobody has priced.
The list has twelve questions in three groups. The first two test adoption and risk. The third, assets, asks about operational records in a way that surfaces data licensing candidates without requesting any confidential file. Use it at a 100-day plan review, a quarterly operating review or a board pre-meeting.
The twelve questions
Adoption: is AI changing how work gets done?
- Q1. Which three workflows use AI today, and who owns each one by name?
- Q2. What has changed in a measurable way, such as cycle time, cost per ticket or win rate, and how was it measured?
- Q3. Which pilots stopped, and why: data quality, integration, adoption or cost?
- Q4. What would you do with one more AI hire or one more quarter of budget?
Risk: what could go wrong, and who is watching?
- Q5. Which AI vendors touch company or customer data, and do their terms let them train on it?
- Q6. What have we promised customers, in contracts or the privacy policy, about using their data for AI training?
- Q7. Which of our services could an AI agent deliver to clients directly, and what is the plan if clients ask for a lower price?
- Q8. Who approves new AI tools, and is there a written acceptable-use policy?
Assets: what records does the company hold?
- Q9. Where does the company's operational history live (email, Slack or Teams, CRM, finance, support, engineering, operations), and how many years does each system cover?
- Q10. Are retired or archived systems still exportable, and who could run an export today?
- Q11. Did we create these records ourselves, or does a meaningful share belong to clients?
- Q12. Would you consider licensing a defined, de-identified set of these records to AI developers for a one-time payment, if we kept ownership and approved the terms?
None of the asset questions asks the CEO to send, upload or describe the content of any record. Answers about systems, years and ownership are enough.
Q6 deserves board attention. FTC staff wrote in January 2024 that companies' promises not to use customer data for undisclosed purposes, such as training or updating models, are enforceable, whether those promises sit in privacy policies, terms of service, promotional materials or marketplaces. It is staff guidance rather than a rule. This is general information, not legal, tax or financial advice. Have the company's counsel review what it has actually promised.
How to read the answers
| Pattern in the answers | What it likely means | Next action |
|---|---|---|
| Named owners and measured results on Q1-Q2 | Adoption is real and managed | Add the next two use cases to the value creation plan |
| Pilots stalled on data quality (Q3) | Systems or data ownership are weak | Pair with the AI readiness assessment |
| No one knows vendor training terms (Q5) | Contract exposure the board has not seen | Have counsel review the main AI vendor contracts |
| Unclear customer commitments (Q6) | Privacy and contract risk | Map existing promises before any new use of data |
| Agent exposure acknowledged but no plan (Q7) | Pricing pressure ahead | Put repricing and service redesign on the next board agenda |
| Ten or more systems, five-plus years of history, exportable and company-created (Q9-Q11) | Possible licensing candidate | Run a fit screen and discuss an introduction |
| Openness on Q12 | A sponsor for the conversation exists | Introduce the company or share your referral link |
The company fit checker gives a preliminary, non-binding read on the asset answers, and the portfolio company screening workbook keeps results for the whole portfolio in one place.
Red flags in the asset answers
- Archives were deleted after a migration, or a retired system was cancelled without an export.
- Most of the records are client conversations or files the company holds on clients' behalf.
- The data is mainly consumer personal information or patient records.
- The same records were already licensed for AI training.
- Peak headcount never reached 50 full-time employees, contractors excluded.
- The CEO would not consider an exclusive license for an agreed term.
Any one of these makes licensing unlikely for now. None of them makes the rest of the AI conversation less useful.
When to ask which questions
| Setting | Which questions | Tip |
|---|---|---|
| 100-day plan review | All twelve | Record the answers as the baseline |
| Quarterly operating review | Q1-Q4, plus any risk that changed | Track the same owners and metrics each quarter |
| Board pre-meeting with the CEO | Q5-Q8 | Agree which risks go on the board agenda |
| Annual budget | Q4 and Q9-Q12 | Decide whether a records inventory belongs in next year's plan |
| Before a system migration | Q9-Q10 | Preserve full exports before the old system is shut down |
If you are new to the seat, the explainer on what a PE operating partner does shows where these reviews fall in the hold.
Turning a yes on Q12 into an introduction
- Confirm that the CEO or CFO is the person who would sponsor and sign.
- Register as a partner, then send the CEO your referral link or submit the company through the referral form.
- SourceX checks headcount, history, data breadth and rights directly with the company; you stay out of the records.
- The company completes its inventory and decides on price and terms, and nothing binds it until it signs.
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. The reward is paid only after the buyer pays and SourceX receives its fee, is not guaranteed, and never reduces the company's proceeds. Operating partners new to the program can start with the operating partner referral overview, and the CEO's likely first question is answered in whether a portfolio company can license its data.
Next step
Take the twelve questions into your next operating review. If the asset answers look strong, register as a partner and make the introduction.
- Step 1Share your linkSend your personal link to a company you know.
- Step 2Company appliesThe company applies itself at /apply.
- Step 3Buyer selects and paysThe buyer selects and pays for the data and SourceX receives its fee.
- Step 4You get your rewardYour share of SourceX fees becomes payable.
Common questions
Should these questions go to the CEO, the CIO or the CFO?
Start with the CEO, because whether the CEO can answer is itself informative. Route the detail afterwards: the CIO or IT lead for system history and exports, the CFO or general counsel for vendor terms and customer commitments. Q12 should stay with the CEO or owner, since only an authorized executive can decide to explore a license.
How often should an operating partner ask the full set?
Ask all twelve once at the start of the hold or at the 100-day review to set a baseline, then annually at budget time. Between those points, the adoption questions fit naturally into quarterly operating reviews, and the records questions should come back whenever a system migration, consolidation or retirement is planned.
What if the CEO says the company has no AI strategy yet?
Start with the asset questions, Q9 to Q11, which do not depend on any AI adoption. A company that has done little with AI can still hold years of well-organized operational records. The adoption and risk answers then give you the starting point for a plan rather than a measure of progress.
Is it appropriate for an operating partner to raise data licensing with a CEO?
Yes, framed as a question about assets rather than a pitch. The CEO decides whether to explore it, and nothing binds the company until it signs. If you intend to introduce the company as a referral partner, tell the CEO you may receive a reward from SourceX's fee and check your firm's policy on such fees first.
Can the questions be sent as a written survey instead of asked live?
Questions one to eleven work well as a pre-read or short survey, which also makes portfolio-wide comparison easier. Q12 is better asked in conversation, because the CEO's reaction, including hesitation about exclusivity or confidentiality, tells you more than a yes or no on a form.
Related pages
- AI readiness assessment for portfolio companies: what to check and what it tells you
- Check Company Fit for Data Licensing
- Portfolio Company Screening Workbook for Partners
- What does an operating partner do in private equity?
- Referral opportunities for private equity operating partners
- Can a PE portfolio company license its data to AI companies?
Free resources
- Client data licensing eligibility checker — A transparent preliminary screen for one company.
- Enterprise value calculator — Enterprise value from equity value, debt and cash.
- Earnout scenario calculator — Probability-weighted earnout value and its present value.
- All free tools · MCP resource center
By SourceX Partnerships Team · Published 2026-10-09 · Updated 2026-10-09
Know a US company with valuable proprietary data?
Become a referral partner from anywhere we support, get your link and introduce an owner or authorized decision-maker.
Refer a company →I own a business
Explore licensing your company's data to AI developers worldwide. Start a short assessment; no uploads needed.
Start an assessment