Primary Research for Search Fund Principals

Search fund diligence math is different.

You are one person, maybe two. The work that a PE fund hands to a deal team of eight — customer interviews, former-employee calls, industry scoping, competitive intelligence — lands entirely on you. You have no analysts to delegate to. You cannot run the expert-network motion that large funds run, because $500–1,000 per expert hour burns through a 1% search budget in a week.

And yet primary research is still the single highest-leverage activity you do during a live deal. Customer reference calls alone tell you more about whether a target is worth buying than any management deck or quality-of-earnings report. The problem isn't whether you need them. It's how you run enough of them without burning out, missing the window, or cutting corners that show up later as a broken thesis.

This post is for search fund principals, ETA acquirers, and independent sponsors running diligence solo or with one partner. It covers the three interview types that actually move a buy/kill decision, what to ask in each, and how AI-moderated interviews make the arithmetic possible for a 1-person team.

The three interview types that matter

You don't have bandwidth for comprehensive diligence interviews. You need to concentrate your primary research on the interviews that change outcomes.

1. Customer reference calls — the highest-signal interviews

These are the single most important interviews you run during a live deal. Why? Because customers are the only stakeholders who can tell you what's real about the target's revenue story. Management will tell you retention is "strong." Customers will tell you the retention number, and the reason behind it, and whether they'd leave for a competitor if the right alternative showed up.

What to ask:

• How long have you been a customer?
• Describe the purchase process — what did you evaluate, and why did you pick this vendor?
• What's your actual renewal probability this year? What would change it?
• If this vendor disappeared tomorrow, who would you switch to?
• On a 1–10 scale, how likely are you to recommend this vendor?
• Has your spend with this vendor grown, stayed flat, or shrunk over the past 2 years?
• Describe the quality of the service or product you actually receive.

How many to run: 5–10 for a typical small-business acquisition. More if the business is customer-concentrated. The counts that matter: 3 should be tenured customers, 2 should be newer customers, 1 should be a recently-churned customer if you can find one.

The signal that kills a deal: Customers who describe the target as "the one we happened to pick" instead of "the one we picked because X, Y, Z." Commodity positioning in a small business is a value-creation trap — the revenue is real today but the defensibility is fake, and you'll find out the hard way during the hold.

2. Former employee calls — especially former sales and ops leaders

Former employees tell you the internal story. How the sales team actually performs. What the culture of the business is. Why the best people left. What the founder won't admit about competitive threats. Former sales leaders are the highest-signal former employees because they know the deal pipeline, the win/loss reasons, and the customer-level economics.

What to ask:

• Why did you leave?
• What did you think about the sales pipeline when you left? Was it healthy?
• Who was the real competition, and what did you lose deals to?
• What was the founder or leader actually like to work with?
• If you were running the business today, what would you change immediately?
• Is there anyone I should talk to who isn't on management's reference list?

How many to run: 2–4. You want at least one former leader (sales, ops, or finance) and at least one former individual contributor.

The signal that kills a deal: Multiple former employees independently naming the same competitive threat, the same customer-concentration risk, or the same founder dysfunction. Consistent stories across independent sources are almost always true.

3. Vertical scoping interviews — for thesis expansion, not deal diligence

Between deals, search funds face a different research problem: is this vertical actually worth the next 12 months of your search time? That's where vertical scoping interviews come in. Industry experts — former executives at comparable businesses, trade association staff, vertical consultants — can tell you whether the category has the economics and dynamics your search thesis assumes.

What to ask:

• What's the typical margin structure for a business like this?
• Who are the 5 biggest competitors, and how are they differentiated?
• Is the category consolidating, fragmenting, or steady?
• What's the typical customer acquisition cost and lifetime value?
• Where are the leverage points for a new owner — pricing, operations, sales team, technology?
• If you had $3M and 10 years, would you buy a business in this category? Why or why not?

How many to run: 3–6 when you're seriously scoping a new vertical. Fewer for incremental thesis refinement.

The execution problem (and why this used to be so hard)

Here's the math that blocks solo searchers and independent sponsors from running the interview counts above:

• A typical customer reference call takes 30–45 minutes, plus scheduling coordination (15–30 minutes), plus post-call notes (20–30 minutes). Total: ~90 minutes per interview end-to-end.
• Running 8 customer calls for a single LOI takes ~12 hours of searcher time, concentrated in the LOI-to-SPA window when you're also working through the financials and negotiating terms.
• Quality degrades under load. After five calls in a day you stop asking good follow-up questions, and the calls produce worse signal than the ones you ran fresh in the morning.
• You miss the window. By the time you've run 8 calls, the seller is frustrated with the timeline or another buyer has moved faster.

The traditional solo-searcher workarounds are all bad:

• Cut the call count. You run 3 customer calls instead of 8, and you find out in year 2 of ownership that the concentrated customer you didn't interview was the one who was planning to leave.
• Hire a junior analyst. Most searchers can't justify the cost, and the training time exceeds the deal window.
• Use an expert network. Per-hour pricing eats your search budget, and the experts are usually mismatched to small-business diligence.

How AI-moderated interviews change the math

AI-moderated expert interviews change the calculus for solo operators in three specific ways:

1. No team required. You don't need an analyst. You don't need a research team. The AI runs the interviews on your question list, on your schedule.
2. Parallel execution during LOI week. You can have 4 customer calls running in parallel on a Tuesday afternoon while you're working through the QoE. Your personal calendar stops being the bottleneck.
3. Structured output for your LOI memo. Every interview comes back as a transcript with key quotes, data points, and a summary. You paste directly into your LOI memo or investor update. No retyping your own scribbled notes.

The cost structure is also different from expert networks: you pay per call, not per expert hour, which means the economics work for small deals.

A rough LOI-to-close interview playbook

Here's what this looks like in practice across a typical 4-week search-fund diligence window.

LOI week:

• 3–5 customer reference calls
• 1–2 former employee calls (prioritize a former sales or ops lead)
• Run these in parallel while you work through financials and draft the LOI memo

Diligence weeks 1–2:

• 3–5 more customer calls, covering tenured/new/churned buckets
• 1–2 more former employee calls
• 1 vertical expert call if you want an outside read on the category

Pre-close week:

• 1–2 final reference checks on specific risks flagged during earlier calls
• Any gap-fill interviews the diligence surfaced

Total interview count: 10–16 across a typical small-business acquisition. For a solo searcher, that used to mean either skipping interviews or blowing the timeline. With AI running the calls in parallel, it's a day of setup and a week of reading transcripts.

What to actually do this week

If you have a live LOI or a target you're seriously scoping, here's the practical path to try this without betting the deal on it:

1. Pick your next target. Not one you're already halfway through — give yourself setup time.
2. Write a 10-question customer reference script. Start with the questions above. Adapt to the business.
3. Run 3 customer calls via AI-moderated interviews during LOI week. Run the other 2 yourself for comparison.
4. Compare the transcripts. Did the AI-run calls surface the same quality of signal? Did you catch things in the AI transcripts you'd have missed running them live?
5. Scale or revert. If the AI interviews landed, run more of your diligence this way. If they didn't, you've lost a week — not a deal.

The goal isn't to automate yourself out of primary research. The goal is to let you run the interview count your diligence actually needs, at the timeline your LOI actually has, without breaking yourself in the process.

---

InsightAgent runs AI-moderated expert interviews for search fund principals, independent sponsors, and entrepreneurial acquirers. See how the solo-searcher workflow fits together.