Due Diligence: The Questions That Actually Matter

Part 8 of The Founder's Guide to AI-Enabled Roll-Ups

You've answered 847 questions in the data room. You've produced three years of financials, client lists sorted by revenue, employee contracts, and a narrative explaining every revenue fluctuation since 2019. The buyer knows your largest client's payment terms, your smallest employee's start date, and exactly which months you took owner distributions.

What do you know about them?

If you're like most founders at this stage, the answer is: what they've told you. You've seen the pitch deck. You've met the deal team. You've heard the vision for how AI will transform your business. You might have Googled the partners and skimmed their LinkedIn profiles.

That's not diligence. That's hope dressed up as research.

The Founder Who Asked Questions

Michael ran a 22-person accounting firm in the Midwest. Solid practice, good clients, $3.2 million in revenue. When a PE-backed roll-up approached him in early 2024, everything looked right. The multiple was fair. The integration plan seemed reasonable. The technology pitch was compelling.

But Michael had watched a friend sell to a different platform two years earlier. That friend's experience—missed earn-outs, delayed technology, key staff departing—made Michael cautious. So he did something most founders skip: he investigated the buyer as thoroughly as he was investigating them.

He requested five references. The buyer provided them readily. He found three more independently through LinkedIn—two current portfolio company founders and one who had left after his earn-out period. He asked all eight the same questions. He documented responses in a spreadsheet.

The pattern that emerged wasn't damning. The buyer was legitimate, reasonably competent, and generally delivered on commitments. But three consistent themes appeared across every conversation:

First, technology deployment took eighteen months, not the twelve months promised in pitch meetings. Second, the first six months post-closing involved far more meetings and reporting requirements than anyone expected. Third, earnouts tied to year-one performance were difficult to achieve—not because targets were unreasonable, but because integration disruption temporarily reduced productivity.

Michael didn't walk away. He renegotiated.

His earn-out was restructured around eighteen-month metrics rather than twelve. He negotiated a specific cap on required meeting hours during the first quarter. He got written confirmation that his earn-out targets would be adjusted if the buyer delayed technology deployment beyond agreed milestones.

"The deal economics were basically the same," he told me. "But my expectations were completely different. When month eight rolled around, and we were still in pilot phase with the AI tools, I wasn't frustrated. I wasn't calling my lawyer. I just kept working, because that's exactly what the references told me would happen."

The founder who doesn't ask questions signs the same deal with twelve-month earn-out targets, expects rapid transformation, and spends year one confused and resentful when reality diverges from the pitch.

Why Founders Skip This Step

The asymmetry is partly psychological. By the time you're deep in negotiations, you've already mentally committed to selling. The buyer has been courting you for months. Their team seems professional. The offer looks attractive. Asking hard questions feels like it might derail the process—or worse, signal that you're not committed.

There's something else, too. A discomfort that's hard to articulate.

The buyer has positioned themselves as the evaluator. You've been the one answering questions, producing documents, and justifying decisions. The power dynamic has been established: they judge, you perform. Flipping that dynamic—becoming the investigator rather than the investigated—feels presumptuous. Who are you to question them?

This is exactly backwards. You're not asking for a job. You're transferring ownership of something you built over years or decades. You're entrusting your clients to their stewardship and your employees to their management. The buyer is making a financial investment. You're making a life decision. Your need for information is at least as legitimate as theirs.

Good buyers understand this. As one private equity firm notes, they actively encourage sellers to conduct reference calls and get to know their buyer through conversations with founders who've been through the process before. A buyer who discourages your diligence is telling you something important about how they'll treat you after closing.

Reference Call Framework

When you hire a senior executive, you call their former employers. When you're about to transfer ownership of your life's work, you should do at least as much.

Request references from three to five founders or executives of companies the buyer has previously acquired. This request is entirely reasonable—any serious acquirer will comply. If they hesitate or provide only very recent acquisitions (within the past 6 months), that's information worth noting.

But don't rely solely on the references they provide. Those will be curated—founders who had positive experiences or have ongoing relationships that constrain honest feedback. Find two or three additional references independently. LinkedIn makes this straightforward. Search for companies in their portfolio, identify former owners or executives, and reach out directly.

The most valuable conversations are often the hardest to arrange: founders who sold to the platform but have since left. They have no ongoing relationship to protect. Their earn-outs have paid out or not. They can speak freely in ways that current portfolio company leaders cannot.

Questions that reveal reality:

"What happened when things weren't going well? How did they respond to the first major problem post-closing?"

Good answer: "We had a significant client departure in month four. They were on a call with me within 24 hours, helped develop a retention strategy for at-risk accounts, and adjusted my earn-out timeline to account for the revenue impact. It wasn't fun, but they were responsive."

Concerning answer: "I didn't really have any problems." (Either they're not being candid, or they haven't been in the portfolio long enough to encounter friction.)

"Did the deal terms change between LOI and closing? If so, how did they handle those conversations?"

Good answer: "They found an issue in diligence that affected valuation. They explained exactly what they found, showed me their math, and we negotiated a fair adjustment. I didn't love it, but the process was transparent."

Concerning answer: "They came back with a lower number right before closing and basically said take it or leave it." (This pattern, called "re-trading," predicts future behaviour.)

"How has the buyer contributed to your business post-closing? Give me a specific example."

Good answer: "They introduced us to a fractional CFO who restructured our pricing model. Revenue per client is up 18% since implementation."

Concerning answer: "They've been supportive." (Too vague. Press for specifics.)

"If you could go back, would you do the deal again? What would you negotiate differently?"

This question generates the most useful information. Listen for what they'd change—those are the terms you should negotiate now.

Questions that waste time:

"Were they professional?" (Everyone says yes.)

"Did they follow through on commitments?" (Too vague to generate useful answers.)

"Would you recommend them?" (Social pressure makes honest answers rare.)

The goal is to understand behaviour under stress, not behaviour during courtship. Every buyer looks good when the deal is closing. What matters is how they act when earn-out targets are missed, when key employees resign, or when technology integration runs behind schedule.

Diligence by Acquirer Type

Not all buyers are the same. The questions that matter depend on who's across the table.

PE-backed roll-ups (Crete, Alpine, similar models):

These buyers run standardised playbooks across dozens of acquisitions. Your diligence should focus on:

• Playbook consistency: "Is the integration process the same for every acquisition, or customised to each firm?"
• Fund timeline pressure: "When does your current fund's investment period end? Are you under pressure to deploy capital?"
• Portfolio company interaction: "Will I have access to other founders in the portfolio? How often do portfolio companies collaborate?"
• Exit timeline: "What's the typical holding period? When and how do you expect to exit?"

PE-backed buyers optimise for repeatability. That's a strength (they've done this before) and a limitation (your unique situation may not fit their standard approach).

VC-backed technology builders (General Catalyst model, similar):

These buyers are building proprietary technology and acquiring distribution channels. Your diligence should focus on:

• Technology maturity: "How many firms are currently using your AI tools in production? What's the longest deployment?"
• Your role as beta tester: "Will my firm be testing new features before they're proven? What happens if they don't work?"
• Product roadmap input: "How do portfolio companies influence technology development priorities?"
• Long-term vision: "Is the goal to build a standalone technology company, or is there an expected sale to a larger acquirer?"

VC-backed buyers may offer more technology upside but also more technology risk. You're joining earlier in their journey.

Technology companies acquiring distribution (Crescendo, similar):

These buyers already have working technology and want your client base. Your diligence should focus on:

• Integration speed: "What's your expected timeline from close to full technology deployment?"
• Cultural integration: "How have previous acquisitions adapted to your company culture? What's been hardest?"
• Role clarity: "Am I primarily responsible for client retention, or will I have broader operational responsibilities?"
• Technology non-negotiables: "Which systems and processes must change immediately versus over time?"

Tech acquirers often move faster but expect faster adaptation. Understand whether their pace matches your capacity.

Technology Verification: Is the AI Real?

AI-enabled acquirers are selling a transformation thesis. They claim their technology will automate tasks, increase capacity, and improve margins. Before you bet your business on that thesis, verify it.

This is harder than it sounds. Technical due diligence on AI systems requires evaluating whether a product is a viable solution or merely a polished demo. The difference matters enormously for your post-acquisition experience.

Request a live demonstration with your actual data. Not a prepared presentation with perfect examples. Not a recorded video showing ideal scenarios. A real-time demonstration using documents from your practice—the messy ones, the edge cases, the exceptions that define your daily work.

Provide them with:

• Scanned documents at odd angles
• Handwritten notes in margins
• Non-standard formats from different clients
• Documents with poor image quality
• Exceptions to standard workflows

Watch what happens. If the demonstration only works with carefully prepared inputs, that tells you something important about the realities of deployment. The pitch might promise 40% time savings. The reality might deliver 15% while creating new categories of exceptions requiring human review.

Ask specific implementation questions:

"What's the average time from close to first AI deployment in portfolio companies?"

Good answer: "Typically four to six months for initial pilot on select workflows, twelve to eighteen months for scaled deployment across the practice. Our fastest was three months at [specific firm], slowest was fourteen months at [specific firm] due to legacy system complexity."

Concerning answer: "It varies by firm." (True but evasive. Press for specifics.)

"What percentage of promised automation has actually been achieved in your earliest acquisitions?"

Good answer: "Our first three acquisitions are seeing 25-35% time reduction on targeted workflows. That's below our initial projections of 40%, and we've adjusted our models accordingly."

Concerning answer: "We're still measuring." (After multiple acquisitions, they should have data.)

"Can you share metrics from a firm similar to mine—same size, same service mix—showing before-and-after productivity?"

Good answer: Specific numbers, specific firm characteristics, willingness to connect you with that founder.

Concerning answer: "Our metrics are confidential." (They're asking you to share everything about your business. Reciprocity is reasonable.)

The American Bar Association notes that buyers conducting due diligence on AI-enabled targets should examine the type, function, provenance, and use of applicable AI tools, and verify that necessary rights exist to use any AI-generated outputs. The same rigour applies in reverse. You're acquiring a future that depends on AI performance. If the buyer is asking for proprietary information about your business, asking for performance data about their technology is entirely appropriate.

Financial Capacity and Closing Risk

A deal isn't done until it closes. Between LOI and closing, financing can fall through, investment committees can baulk, and market conditions can shift. If the buyer's financial capacity is questionable, you bear the risk of a failed transaction—wasted time, disclosed confidential information, and potentially damaged client relationships if word got out.

Questions to assess financial certainty:

"How is this acquisition being funded? Cash from existing fund, debt facility, or co-investment?"

Good answer: "This will be funded from Fund III, which closed at $450 million in 2023. We have approximately $180 million in remaining dry powder. No debt financing required for acquisitions at this size."

Concerning answer: "We're still finalising the capital structure." (At the LOI stage, this should be clear.)

"If debt is involved, has the lender provided a commitment letter? Can I review it?"

Good answer: "Yes, here's the commitment letter from [lender]. The terms are [specific terms]."

Concerning answer: "We have strong lender relationships and don't anticipate any issues." (Relationships aren't commitments.)

"Have you ever failed to close a signed deal? What happened?"

Good answer: "Once, in 2021. We discovered material misrepresentation in the seller's financials during confirmatory diligence. Here's what we found and why we walked away." (Honest, specific, reasonable explanation.)

Concerning answer: "No, never." (Either they haven't done many deals, or they're not being forthcoming.)

"What conditions could cause you to renegotiate the purchase price between now and closing?"

Good answer: "Material adverse changes in revenue trajectory, discovery of undisclosed liabilities, or key employee departures before closing. Here's our standard list of diligence items that could affect valuation."

Concerning answer: "We stand behind our LOI terms." (Sounds good, but doesn't answer the question. Get specifics.)

According to PwC, both buyers and sellers are becoming increasingly sophisticated in seeking to exploit value through negotiation of transaction documents. Understand the buyer's history. Ask your references directly: "Did the purchase price change between LOI and closing? How was that conversation handled?"

Pay attention to exclusivity requests. Buyers typically want 60-90 days of exclusivity during confirmatory diligence. If they're asking for significantly longer, ask why. Extended exclusivity periods can signal financing uncertainty, a pattern of protracted negotiations, or simply a slow-moving organisation—all of which predict post-closing behaviour.

Post-Closing Plans: Where Generic Meets Specific

The buyer has told you their vision for your business. Now make them get specific.

On employees:

"What is your typical employee retention rate in the first twelve months post-acquisition?"

Good answer: "Across our portfolio, we see 15-20% voluntary turnover in year one, concentrated in months three through six. We've found that middle managers are the highest risk, and we've developed specific retention approaches for that tier."

Concerning answer: "We have excellent retention." (No data, no specifics.)

"How do you handle retention bonuses? Who qualifies? What's the typical structure?"

Good answer: "Key employees identified during diligence receive retention bonuses structured as [specific terms]. Here's our standard retention agreement template."

Concerning answer: "We evaluate each situation individually." (Means they either don't have a playbook or won't share it.)

Research on post-acquisition employee retention shows that more than a third of acquired employees leave within the first year. Buyers who acknowledge this reality and have specific mitigation strategies are more trustworthy than those claiming perfect retention rates.

On clients:

"Do you have a standard client communication playbook? Can I review it?"

Good answer: "Yes, here's our communication timeline and template language. We typically do [specific approach] for top-tier clients and [different approach] for the broader base."

Concerning answer: "We work with each founder to develop an appropriate communication plan." (Probably means they don't have one.)

"Have you experienced significant client attrition in any previous acquisition? What happened?"

Good answer: "In one acquisition, we lost approximately 12% of clients in the first year. Here's what we learned and what we changed."

Concerning answer: "No significant attrition." (Either they haven't done enough deals or they're defining "significant" very generously.)

On your role:

"What will my typical week look like at month three? Month twelve?"

Good answer: A specific description of meetings, reporting requirements, client responsibilities, and operational duties—with acknowledgement that it differs from your current week.

Concerning answer: "That's really up to you." (Sounds empowering, but probably means they haven't thought it through.)

"What would cause you to remove a founder from an operating role before their earn-out period ends?"

This matters more than it might seem. Earn-out provisions often include "termination for cause" clauses that can void your contingent payments. Understand exactly what behaviours would trigger those provisions—and get specific examples, not just legal language.

Red Flags and Green Flags

After multiple conversations and document reviews, you'll have accumulated signals. Some suggest trustworthy partners. Others suggest problems ahead.

Red flags:

• Reluctance to provide references, or only providing references from very recent (within six months) acquisitions
• Significant gaps between what the buyer claims and what the references report
• Technology demonstrations that only work under controlled conditions
• Vague answers about funding sources or financing structure
• History of purchase price reductions between LOI and closing
• High turnover in their own deal team or integration team
• Unwillingness to put post-closing commitments in writing
• Aggressive timelines that leave insufficient time for your diligence
• References who hesitate noticeably before answering questions
• Defensiveness when you ask hard questions

Green flags:

• Transparent sharing of both successes and challenges from previous acquisitions
• References who speak candidly, including about difficulties, and still recommend the buyer
• Technology that performs on messy, real-world examples—not just prepared demos
• Clear documentation of fund status, financing, and approval processes
• Willingness to make earn-out terms contingent on factors you can actually control
• Stability in their integration team and consistent playbooks across portfolio companies
• Specific, detailed answers rather than generic assurances
• Patience with your diligence process and encouragement to be thorough
• Ability to name specific mistakes from previous integrations and what they learned

What's Counterintuitive About Reverse Diligence

Aggressive diligence doesn't scare away good buyers—it scares away bad ones.

Founders worry that asking hard questions will make them seem difficult or uncommitted. The opposite is true. Sophisticated buyers expect thorough diligence. It signals that you're a serious operator who thinks carefully about major decisions—exactly the kind of founder they want in their portfolio.

A buyer who gets defensive when you investigate them is telling you something important. If they can't handle scrutiny during courtship, they won't handle disagreement well during integration.

The references the buyer provides can still be valuable—if you ask the right questions.

Yes, these references are curated. But curated doesn't mean useless. Even founders with positive experiences will share their concerns if you ask the right questions. "What would you negotiate differently?" generates useful information even from satisfied sellers. "What surprised you most?" surfaces issues the buyer didn't anticipate and therefore didn't warn you about.

Perfect track records are a red flag, not a green flag.

Every integration encounters problems. Every technology deployment hits obstacles. Every culture clash produces friction. A buyer who claims uniformly positive outcomes is either lying, hasn't done enough deals to encounter real problems, or lacks the self-awareness to recognise difficulties.

The buyer who says "Our third acquisition was really hard—here's what went wrong and what we learned" is more trustworthy than the one who says "All our founders are thriving." Honesty about past difficulties predicts honesty about future ones.

What We Don't Know (And Can't Learn Through Diligence)

Reverse diligence improves your information. It doesn't make you omniscient.

Survivorship bias shapes every reference call. The founders willing to talk are the ones still engaged with the platform or who left on reasonable terms. The ones who left angry, who felt cheated, who would tell you to run—they're harder to find and less likely to respond to cold outreach. Your sample is biased toward acceptable outcomes.

Technology performance elsewhere doesn't guarantee performance at your firm. The AI that worked beautifully at a 30-person tax practice might struggle with your 15-person audit-heavy firm. Different client bases, different document types, different workflows. Past performance, as they say, does not guarantee future results.

Buyer behaviour can change. The deal team courting you today may not be the integration team managing you tomorrow. The fund's strategy can shift. Key personnel can depart. A buyer with a strong track record can stumble on their fifteenth acquisition in ways they didn't on their fifth.

Personal chemistry is hard to predict. References can tell you how the buyer treated them. They can't tell you how the buyer will treat you. Personality, communication styles, and conflict-resolution approaches vary from person to person. The partner who was responsive and supportive with one founder might be dismissive with another.

None of this means diligence is pointless. It means diligence improves your odds rather than guaranteeing your outcome. You're making a probabilistic decision, not a certain one. The goal is to be well-informed, not to be certain.

Building Your Diligence Timeline

Reverse diligence takes time. Build it into your deal timeline explicitly, running parallel to the buyer's investigation of you.

Week 1-2 post-LOI: Request reference list from buyer. Begin independent research to identify additional contacts. Request a technology demonstration with your data.

Week 2-4: Conduct reference calls. Aim for at least five conversations—three provided by the buyer, two found independently. If you can find a founder who's left the platform, prioritise that conversation. Document responses systematically using consistent questions.

Week 3-5: Technology verification. Observe a live demonstration with your actual documents. Request portfolio company metrics showing AI deployment timelines and results. Ask for a connection to a comparable firm already using the technology.

Week 4-6: Financial diligence. Review fund documentation, financing commitments, and approval requirements. Understand closing conditions and timeline. Investigate any history of price renegotiation.

Week 5-7: Post-closing planning. Review integration playbooks, employee communication templates, and client retention strategies. Get specifics on your role, reporting requirements, and decision authority. Negotiate modifications based on diligence findings.

Throughout: Update your M&A advisor and legal counsel with findings. Adjust deal terms as needed to reflect risks discovered during the investigation.

This timeline assumes 60-90 day exclusivity. If the buyer wants longer, ask why. If they want shorter, push back—thorough diligence protects both parties.

When Diligence Changes the Deal

Sometimes what you learn changes your view of the transaction.

If references reveal consistent problems—price adjustments after LOI, aggressive behaviour during integration, failure to deliver on technology promises—you have options. You can walk away entirely. You can negotiate deal terms that protect against the specific risks you've identified. Or you can proceed with adjusted expectations.

If technology verification shows the AI is less mature than advertised, consider whether earn-out terms should reflect that reality. If the buyer promised 40% productivity gains but portfolio companies are seeing 15%, your earnout targets should be calibrated to achievable levels, not pitch-deck projections.

If financial diligence reveals uncertainty, a fund near the end of its investment period, debt financing that isn't yet committed, or approval processes that could delay closing, build protections into your timeline. Set a drop-dead date after which the deal terminates. Negotiate a reverse break-up fee if they fail to close through no fault of yours.

Michael, the Midwest founder from earlier, used his diligence findings to restructure three specific deal terms. He didn't get more money. He got more realistic terms—which, twelve months later, proved more valuable than a higher headline number on a structure he couldn't achieve.

Asymmetry You Can Correct

The buyer will always know more about your business than you know about theirs. That asymmetry is structural—they're in the business of acquiring companies, and you're selling one for probably the first time.

But you can narrow the gap. You can ask questions that reveal how they actually operate, not just how they present themselves. You can talk to founders who've walked this path and learn from their experience. You can verify technology claims against real-world results rather than polished demonstrations.

The goal isn't to catch the buyer in a lie. Most buyers are legitimate operators running real businesses. The goal is to ensure that the partnership you're entering matches your expectations—and to adjust those expectations, or the deal terms, when reality differs from the pitch.

You built something valuable. Before you hand it over, make sure you know who's receiving it.

← Back to Chapter 7: The Integration Playbook

Continue to Chapter 9: When to Walk Away →

Or return to: The Founder's Guide to AI-Enabled Roll-Ups (Hub)

Capital Founders OS is an educational platform for founders with $5M–$100M in assets. We focus on frameworks for thinking about wealth—so you can make better decisions.

If it's your first time here, don't forget to Subscribe.