By Jose Antonio Díaz Infante, Founder of Matrits
AI & Automation Lead for Growth Companies
Matrits helps Search Funds, PE teams, M&A advisors, and growth companies build AI-powered analysis infrastructure that compresses timelines and eliminates friction from deal workflows.
To make it easier for you to read this article, we suggest downloading it in PDF format.
You just closed on a B2B services company. EBITDA looked solid. Customer retention at 85%. Margins climbing. Your diligence was solid.
Six months later: 60% of your operations depend on one person who quit. The infrastructure you thought could scale runs on systems from 2012. Your data lives in six spreadsheets. There’s zero automation—every invoice, every report, every customer touchpoint is manual.
Your IRR dropped. Not because the numbers were wrong. Because nobody asked what was actually running the business before you signed.
This is what M&A and Search Fund teams miss.
- Financial due diligence isn’t technical due diligence
I talk to Searchers, PE operators, and M&A advisors who’ve walked into this. The pattern never changes: we crushed the EBITDA analysis. Validated customers. Stress-tested margins. What we never did was ask what keeps the lights on.
A company with solid cash flow can still be built on broken systems. You won’t see it until it’s yours.
What deal teams overlook:
Infrastructure is fragile. Legacy systems. Vendor lock-in. Single points of failure that work today but can’t scale. Lose a key engineer and the whole thing seizes. You inherit that.
Data is scattered. Customer information in one system. Operations in another. Financials nowhere near either. Nobody has a complete picture of the actual business. You can’t make smart decisions. You can’t automate. You can’t compete on data.
There’s no automation. Most companies still do everything manually. Manual invoicing, manual reporting, manual data entry. That’s not simplicity—that’s salary trapped in repetitive work. It’s a hard ceiling on what your margins can be.
Then there’s technical debt. Code nobody understands. Systems that fail randomly. Zero documentation. Not deal-breakers. Landmines that explode later.
Reality: 18-36 months spent fixing things you didn’t budget for. A 15-20% efficiency gap you can’t close without a complete rebuild. An IRR that bears no resemblance to your model.
- Pre and post-close assessment
The fix sounds simple. Execution is harder: ask different questions at different stages.
Pre-close:
Add three technology pillars to your decision tree: Infrastructure, Data, AI/Automation. Don’t think like a technologist. Think like an investor: how fragile is this operation really?
- Infrastructure. What actually runs this business? Where are the breaking points? Will it survive 3x the current volume? What happens if their lead engineer leaves? How much custom code are they stuck with? How locked in are they to one vendor?
- Data. Where does critical data actually live? Centralized or scattered everywhere? Can you actually access it? Is it clean? When you ask management about data governance, their silence is more informative than any answer.
- AI/Automation. Walk through their main processes. Invoicing, onboarding, customer support, reporting. How much is still manual? What could be automated? How many hours of labor are buried in repetitive work?
Look at the code too. Not a full audit—just a scan. Old dependencies? Systems nobody documented? Any tests? This doesn’t kill a deal. It tells you what fixing it actually costs.
You end up with a risk matrix. Infrastructure fragility: 7/10. Data accessibility: 4/10. Automation potential: 9/10. Now you know what you’re actually buying.
Post-close:
Day one priorities are two things: stabilize critical systems and find wins you can execute in 90 days.
Stabilizing means documenting how things actually work. Map what depends on what. Cross-train people so the operation doesn’t break if someone leaves. Write down the basic runbooks. You’re not rebuilding—you’re making it transparent.
Quick wins are different. Find cheap friction to fix. Maybe it’s a data integration that kills three manual reports. Maybe it’s automating invoicing and freeing up 40 hours a month. Maybe it’s consolidating customer data so you can actually run the business from data instead of spreadsheets.
Pattern: small technical fixes that release cash and margin immediately. Then plan the bigger changes.
- How AI actually changes this
Pre-close, Claude reads your technical documentation—architecture, codebase, systems specs—and gives you a real risk assessment in minutes. You ask “What’s the EBITDA impact if we can’t fix the infrastructure debt in year one?” and get an answer grounded in their actual systems.
Post-close, AI finds quick wins fast. Tell it your current processes, team size, budget, your timeline to next funding. It ranks technical improvements by ROI per hour. No guessing. Real sequencing.
Then the tactical work: automating data pipelines that took your team weeks to manually map. Pulling documentation from code. Building connections between your data silos.
Doing this manually takes months. Doing it with AI infrastructure takes weeks.
What actually changes:
Deal teams that added technical assessment describe the same three shifts:
You disqualify targets faster. Not because the financials don’t work, but because you understand the complexity early. You price it in or you walk.
Your integration moves faster. No surprises at month one. You know the technical landscape before close. Your fix roadmap already exists.
You find automation opportunities the seller never thought of. Improvements that weren’t in your model. Margin expansion happens faster because you’re working from reality.
One Searcher I know did infrastructure assessment pre-close, discovered their invoicing was 30% manual, built that automation upside into the thesis. Year one: that one automation project freed up $180K in labor. Funded everything else.
She didn’t discover anything revolutionary. She just asked better questions first.
What happens when you skip this:
Deal teams moving fast in 2026 understand: EBITDA quality depends partly on technical foundation. A company making steady margins on fragile systems isn’t the same as one built to last.
Financial diligence is table stakes. Technical assessment is now the differentiator.
Here’s the real cost of skipping it:
You close a deal that looks good. Six months in you’re staring at $400K-$800K in infrastructure debt you didn’t see coming. Your 90-day automation plan becomes an 18-month rebuild. People leave because systems are broken. Margin targets slip. The deal you thought would hit 25% IRR is tracking 12%.
Worst part? All of it was visible before you signed. You just didn’t ask.
The teams getting this right:
The Searchers and M&A advisors avoiding this ask technical questions before close. They price in complexity early. They walk from deals with hidden problems. They close companies that actually scale.
- What you do next
Two paths:
- Path one: Close your next deal without this assessment. Hope nothing breaks. Fix problems after close.
- Path two: Spend some time with my team before you close. We run the technical assessment. You know exactly what you’re buying and exactly what it costs to fix. You either price it in or you don’t close.
One path protects your IRR. The other doesn’t.
I work with Search Funds, PE teams, and M&A advisors on exactly this: technical assessment pre-close so you close with your eyes open, and automation priorities post-close so you hit your margin targets in 90 days instead of 18 months.
→ Schedule a consultation Let’s talk about what technical assessment actually changes on your next deal.
No magic. Just better questions.
