Insights
Does AI outreach still work for M&A deal sourcing?
Mitchel Campbell, Co-founder and CEO
· 6 min read
The short answer
AI-written outreach on its own has stopped working. Owners now get several near-identical approaches a week and ignore them, and what still earns a reply is proof you know their sector, a specific reason to talk, and a relationship behind the first message.
That does not mean AI has no place in sourcing. It means the value has moved. Eighteen months ago, an AI-written email to a business owner was an edge. Today it is the default, and owners can tell. The firms still getting results use AI to free up time for the work that does not scale, and to make sure they can be found when an owner starts looking.
Key takeaways
- Generic buyer language now reads as spam – lines like "I have a qualified buyer" land in the same pile as every other template
- Outreach is table stakes – volume keeps the funnel full, but it does not win the deal
- What else still earns a reply – a real buyer overview, a relevant closed deal and value up front
- First-draft research is now the starting line – since the September 2026 launches, every firm can produce a market summary or target list in minutes
- Your own data is the edge – relationship history, intermediary coverage and a clean CRM are what a model cannot see, and cleaning that data is the real first project
- Owners are starting to ask ChatGPT who to call – specific content, case studies, sector pages and consistent directory listings appear to help a firm get cited
Why AI outreach stopped working
When everyone can send it, volume stops setting anyone apart.
Funds, searchers, independent sponsors and advisors all moved into AI outreach at roughly the same time. The cost of sending a personalised-looking email fell to almost nothing, so the number of emails reaching each owner went up and the reply rate went down.
A PE partner put it well on a call with me recently: everyone dove into AI for deal sourcing first, and it is getting harder and harder to find value there. Advisors describe the same thing from the other side. Generic buyer language is now recognised as spam.
"I have a qualified buyer for your business" is the clearest example. Owners have read that line so often it lands in the same pile as the extended car warranty call. Several advisory founders I have spoken with now receive cold emails claiming a PE client wants to buy their own advisory firm.
Outreach still matters. It has become table stakes. Volume keeps the funnel full. It does not win the deal.
What still earns a reply from an owner
Effort a template cannot fake.
Based on what advisors and funds are telling me, four things still get owners to respond:
- Proof you know their sector – a specific observation about their market that a template could not have written
- A real buyer overview – the type of buyer, what they are looking for and why this business fits
- A relevant closed deal – something similar you have done, described in plain terms
- Value up front – a valuation range, a readiness checklist, or a view on what buyers are paying for right now
Each of these takes work before you hit send. That is the point. Owners are filtering for advisors and buyers who did the work first.
Where AI does help sourcing now
The edge has moved to what a model cannot see: your relationships, your history and the time you free up.
September 2026 made this clearer. On Sep 10, OpenAI launched ChatGPT for Financial Services, shaped with Morgan Stanley and Evercore as design partners, with PitchBook and LSEG data built in (PYMNTS, Sep 10, 2026). The same day, Intapp released a plug-in that gives ChatGPT a firm's own DealCloud client, deal and relationship history (Intapp, Sep 10, 2026). On Sep 29, Datasite set out a strategy to connect AI across the whole deal, from market intelligence to the data room (Datasite, Sep 29, 2026).
When every firm can produce a first-draft market summary or target list in minutes, that work becomes the starting line. What the model cannot know is your firm. Which owners you have spoken to and when. Which buyers have closed with you before. Which relationships sit with which partner.
That is where AI earns its keep in sourcing today:
- Your relationship history – the owners, buyers and intermediaries you already know, kept current and searchable
- Intermediary coverage – bankers, brokers, CPAs and lawyers who call you first because you stayed in touch
- Clean data – a CRM accurate enough that an agent working from it does not embarrass you
- Time – the hours AI saves only matter if they go into calls, site visits and relationships
That only helps if your own data is clean enough to plug in. For most firms, that is the real first project.
The other direction: owners now ask ChatGPT
Sourcing also runs the other way. More owners start their search with an AI tool instead of Google.
An owner thinking about selling used to search Google for an advisor. Increasingly, they ask ChatGPT. An AI tool does not show ten blue links. It gives one answer and cites a handful of sources, so whether your firm is one of them is starting to matter.
What seems to get a firm cited, based on what I am seeing so far:
- Specific content – a real view on valuations or buyer appetite in your sectors, not generic "how to sell" articles
- Case studies – closed deals described in plain terms
- Sector pages – one page for each industry you actually cover
- Consistent descriptions – the same name, focus and location on your website and in every directory you are listed in
This is early, and nobody has a reliable playbook yet. Firms that already publish specific, useful content are starting from a better place.
Frequently asked questions
Does AI-written outreach still work for M&A advisors?
Not on its own. Owners now receive many near-identical AI-written approaches and treat generic buyer language as spam. Outreach still keeps the funnel full, but replies come from specific proof of sector knowledge, a relevant deal or something useful up front.
What should a small PE fund use AI for in sourcing?
Use it to keep your relationship history and intermediary coverage current, and to free partner time for calls and visits. Proprietary sourcing now depends on what you know that a scraped list does not.
Should we stop sending outreach altogether?
No. Treat it as table stakes rather than the strategy. Fewer, better-researched messages tend to outperform volume.
How do owners find advisors through AI search?
They ask a tool like ChatGPT who to call, and it cites a small number of sources. Specific content, case studies, sector pages and consistent directory listings appear to help a firm get cited.
Where to start
If you want to see where your own firm's hours go before deciding what to change, take the AI readiness check. It takes about four minutes and shows where AI would help most.
For the full picture by audience, read the implementation guide for M&A advisory firms or private equity.
At Analyst3, we start with an audit of where a firm's time actually goes, then build the workflow that returns the most hours first, with a person approving anything that reaches an owner.