How to talk to business owners about AI without hype: scripts for M&A advisors
Talk to business owners about AI in plain words tied to their own company: what it means for their value, what buyers will ask in diligence and whether their records are an asset, since AI developers license real operational records. Promise no price or deal, disclose any referral reward, and ask permission before introducing the company to SourceX.
The short answer: plain words, their company, their permission
Talk to business owners about AI the way you would discuss any market force that touches their value: in plain words, tied to their own company, with no promises. Owners tend to want three things answered: what AI means for what the business is worth, what buyers will ask about it, and whether anything they already have is an asset. The third is where data licensing comes in, because AI developers license real operational records from established companies. End each conversation with a question the owner can decline, and introduce no one without permission.
Why M&A advisors get the question
Owners raise AI with the people they trust on value. A sell-side advisor sees the financials, learns the systems while preparing for diligence and hears directly from buyers, many of whom now include AI questions in their diligence requests. That makes the advisor the natural person to separate noise from what actually moves a valuation.
These conversations will multiply. McKinsey estimates that by 2035 about six million US small and medium-size businesses will face ownership transitions as baby boomers retire, and that more than one million are viable candidates for sale (McKinsey, February 2026). Owners heading into that market will meet buyers who ask about AI.
The 3P rule: plain, particular, permission
Check every AI script against three tests before you use it.
- Plain: could the owner repeat it to their CFO or spouse without a glossary?
- Particular: does it mention their systems, their records or their likely buyers rather than AI in general?
- Permission: does it end with a question the owner can say no to?
A script that fails any test goes back for a rewrite. Owners discount advisors who sound like vendor marketing, and they remember the ones who were straight with them.
Words to swap out
| Jargon | Plain alternative |
|---|---|
| Large language model | AI that reads and writes text |
| Agentic AI | AI that carries out multi-step tasks, such as working a support ticket through to resolution |
| Training data | Examples an AI system learns from |
| Evaluation data | Real cases used to check whether an AI system did the job correctly |
| Monetize your data | License copies of your records for a one-time payment while you keep ownership |
| Data moat | Records nobody else has |
What AI developers license, in one paragraph
AI is shifting from tools that answer questions to agents that perform tasks, and teaching agents to do real work takes records of real work: tickets and how they were resolved, quotes and whether they were won, engineering reviews, approvals and exceptions. That material is thin on the public web. Researchers at Epoch AI have projected that, if current trends continue, language models will fully use the stock of public human-generated text sometime between 2026 and 2032. It is a forecast with wide uncertainty, but it explains why licensed, permissioned business records have become a scarce input. AI labs and data buyers license copies under contract, and the company keeps ownership.
Which owners to raise it with, and when
| Client situation | What you will hear | First question to ask |
|---|---|---|
| Two to five years from an exit | I want to know what buyers will look at | Which of your systems goes back the furthest? |
| Just lost a buyer or paused a sale | Frustration about price | Is there value in the business that the bids ignored? |
| Software, IT services or professional services founder | Pride in how the team works | How many years of tickets, projects or reviews do you still have? |
| Replacing an ERP, CRM or help desk | Cost and disruption worries | What happens to the old system's history after cutover? |
| Rolling out AI tools for their own staff | Worry about data leaking | Have you ever looked at your records as an asset rather than only a risk? |
The best moments sit inside meetings you already hold: the annual check-in with a future seller, the valuation conversation before a mandate, the kickoff where you collect the system list for the CIM, and the week after a bidder sends AI diligence questions. Those questions resurface later in the purchase agreement; see AI representations and warranties in M&A purchase agreements.
Scripts for the five questions owners ask
Will AI make my company worth less?
What will buyers ask about AI?
Is our data worth something?
Is it safe?
Can you introduce us?
What never to promise
- A price, a range or a comparison with a publicly reported deal.
- That a buyer is waiting or that a deal will close.
- That a license will not affect a later sale; a signed license has to be disclosed to acquirers.
- Anything about how a buyer will use the data beyond what the signed agreement says.
- Timing you cannot control. Buyer responses typically take about two weeks, and that clock starts only after qualification and the inventory make the company deal-ready.
How the introduction works after a yes
- You register as a partner and either send the owner your referral link or submit the company through the referral form.
- SourceX talks with the owner or another authorized sponsor about peak headcount, years of operation, the systems in use and who holds the rights.
- The company lists its systems, years of history and what can be exported, without sending any records.
- SourceX and the company agree price and terms; the company sees one all-in price with no separate charges.
- AI labs and data buyers look at the opportunity; the owner can still walk away at any point before signing.
- After signature and the owner's authorization, the records go out with the agreed redactions and the company is paid once.
Your part ends at step one. You never export, upload or describe confidential records.
How your reward works, and what to disclose
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, and rewards become payable only after the buyer pays and SourceX receives its fee. Because it comes out of SourceX's fee, the owner's proceeds are untouched, and like any success-based reward it is not guaranteed. Tell the owner about it in writing before the introduction. Advisors who hold securities registrations or professional licenses should check their own rules first; the guide on whether investment bankers and M&A advisors can accept referral fees covers the main questions.
When not to raise it
- The business is too small: the baseline is 50+ full-time employees at peak (contractors excluded).
- Its records mostly belong to clients, or are mainly consumer personal data or medical records.
- The owner has already ruled out an exclusive license for an agreed term.
- A sale is days from signing and the deal team has not been consulted.
- Nobody at the company can run an export.
Data licensing sits comfortably beside the other non-sale work an advisor can offer; see M&A advisor value-added services beyond the sale and business development ideas between mandates.
Next step
Pick one owner you will speak with this month and try their company in the company fit checker or against the who qualifies baseline. If the screen looks promising, ask the owner's permission, then register as a partner and make the introduction. The M&A advisor referral overview shows where these conversations can lead.
- 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 an advisor bring up AI before the owner does?
Yes, when you can tie it to the owner's company. Many buyers now raise AI in diligence, so an owner who hears those questions first from their own advisor is better prepared. Start with what buyers will ask and what the company already documents, and raise data licensing only if the company has the size, history and records to make it relevant.
What should an advisor say if the owner asks what their data is worth?
Say plainly that nobody can quote a price before the company is qualified, its records are inventoried and buyers have reviewed the opportunity. Price depends on the records, the length of history, the rights position and buyer demand at the time. Nothing is binding until the owner agrees price and terms and signs, so checking fit commits the owner to nothing.
How technical does an advisor need to be to discuss AI with owners?
Not very. Owners want judgment, not architecture diagrams. Know the difference between tools that answer questions and agents that carry out tasks, why real business records matter for training and testing those agents, and what buyers ask in diligence. For anything deeper, point the owner to their CTO, their counsel or a specialist adviser.
Does raising data licensing put the client's confidential information at risk?
Not if the advisor stays in the introducer's role. With permission, you share the owner's contact details and basic fit facts, such as headcount and the kinds of systems used, and nothing from the records themselves. What gets removed or masked is settled with the company before work starts, and delivery waits for an executed agreement and the company's sign-off.
Should the advisor tell the owner about the referral reward?
Yes, in writing and before the introduction. Explain that the reward is a share of SourceX's fee, paid only after the buyer pays and SourceX receives its fee, and that it never reduces what the company receives. Openness protects the relationship, and some professional rules require disclosure of referral compensation in any case, so check yours.
Related pages
- AI representations and warranties in M&A: what buyers ask and how sellers answer
- Can investment bankers and M&A advisors accept referral fees?
- M&A advisor value-added services: adding a records and data-rights review
- M&A advisor business development ideas for the months between mandates
- Check Company Fit for Data Licensing
- Which US businesses are a fit for a SourceX data licensing introduction
Free resources
- Enterprise value calculator — Enterprise value from equity value, debt and cash.
- Earnout scenario calculator — Probability-weighted earnout value and its present value.
- Profit margin calculator — Profit and margin across three scenarios.
- 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