Can't Change What You Can't Measure: The PE Data Access Problem
Diligence validated the financials. Revenue: $35M. EBITDA margin: 14%. The QofE held. The market study confirmed the thesis. The 100-day plan got board approval the week before close.
Day 1 post-close, you ask the controller for unit economics by product line. He gives you a one-page P&L. You ask for cost per unit by SKU. He says they don’t track that. You ask for customer-level margin. He says the system can’t pull it. You ask for the source data so you can build it yourself. He says the seller isn’t comfortable sharing that level of detail “for privacy reasons.”
Welcome to the gap nobody underwrites. The deal model says you can take EBITDA from 14% to 19% in twenty-four months. The operating data required to actually do that work was never part of what got handed over.
The Question Every PE Operator Eventually Asks
A PE-adjacent operator posted the question in an industry forum a few months ago: Do you need access to a P&L to do value creation? The setup was specific — a quasi-operating role inside a family-owned business that gives him “a very basic, non-detailed P&L” and keeps the rest private. He’s been asked to create value. He has no operational data to work with.
The top reply was five words: “Can’t change what you can’t measure.”
That line resonated because every PE operator has lived some version of it. Sometimes the company is a family business that won’t open the books. Sometimes the company is well-meaning but truly does not have the data — the systems were built to file taxes, not to manage operations. Sometimes the data exists in a single founder’s head and nobody else can see it. The cause changes. The result is the same. You’re standing in the middle of a value creation plan with no instrumentation underneath it.
We’ve covered how board-level KPIs cascade down to role-level actions — how the math has to flow at every level for any of it to work. That whole chain assumes the metrics exist in the first place. This post is the prerequisite. Before you can cascade anything, somebody has to be able to see the numbers.
In a meaningful share of lower middle market acquisitions, that’s not where you start. You start before that. You start with visibility problems.
Why Diligence Validates the Wrong Thing
The deal model is built on financial statements. Trailing twelve, last three years, normalized for adjustments. The QofE asks: are these earnings real? Commercial diligence asks: is the market growing? Legal asks: is the cap table clean?
Nobody asks: can we measure how this business actually works at the unit level? Not because the question doesn’t matter — because by the time it surfaces, the deal is already closing.
Financial accounting tells you what happened. It doesn’t tell you why. Revenue went up — was it volume or price? Gross margin compressed — was it mix shift, raw material costs, or pricing pressure on a specific customer segment? Financials aggregate. They smooth. They report at the level the IRS needs to see, not the level you need to act on.
What you need is management/cost accounting information — the layer underneath the P&L. Direct cost per unit. Contribution margin by SKU. Capacity utilization by line. Customer profitability by segment. Data that explains the numbers rather than summarizing them. Most sub-$50M companies don’t have that layer at all. They have a tax accountant, a QuickBooks file, and a one-page P&L. The system was built to satisfy compliance, not to drive operations. As one operator put it, the realities of how a business actually runs don’t fit into a nice cell on a spreadsheet.
The Paradox of Buying What You Can’t See
PE buys companies to improve them. Improving them requires measuring them. But the measurement infrastructure usually doesn’t exist at the granularity improvement requires — and the seller has no incentive to build it for you. In family-owned businesses especially, the reflex is privacy. The operator who posted in the forum described it precisely: the owner doesn’t give you access for privacy reasons. Detailed data feels like a vulnerability, not a tool. They’ve run the company for two decades on instinct. Why suddenly let a new operator pull every transaction?
So the thesis assumes data. The 100-day plan assumes data. The board deck assumes data. The data either doesn’t exist in the right form, or it exists and isn’t being shared. While you wait, the value creation runway is burning.
The bad operators get stuck here. They write memos. They escalate to the operating partner. They wait for the IT migration that’s six months out. They blame the data gap for missing month-three milestones. The hold period is finite. They’ve used three months of it on a permission slip.
The good operators do something different.
Build the Visibility Layer Yourself
We’ve made the broader case that this is the norm, not the edge case — six PE leaders on the Investors & Operators podcast arrived at the same diagnosis independently. This post starts from there and goes tactical: once you accept the data won’t come, what do you actually do?
You can measure more than you think without the formal data. The operators who execute in the first 100 days don’t wait for clean financial access. They build a parallel operational visibility layer — one they own, in parallel to whatever the seller is willing to share.
It looks like this.
Estimate from the outside in. You don’t need a perfect data set to identify the big problems. Rough sales volume plus verbal history from the controller plus a walk-through of the operation plus the basic P&L gets you 80% of the way to a working hypothesis. As one operator put it bluntly: can develop value creation plan with rough sales volume estimate and verbal history. Not because that’s the ideal — because it’s enough to start. The operators who insist on perfect data before they act don’t act.
Sample, don’t demand. You don’t need every invoice. You need fifty. Pull a sample of fifty transactions across product lines, customers, and time periods, and you can reverse-engineer unit economics within a week. Sampling is how auditors do it. It’s how every operations consultant does it. It’s how you should do it. Get it through sampling. Fifty transactions tells you more than a summary P&L because the summary hides the variance and the sample exposes it.
Build operational KPIs that don’t require financial access. A surprising amount of useful information lives outside the P&L. Customer count and concentration. Order volume by week. Cycle time from order to delivery. On-time delivery rate. Defect or rework rate. Headcount turnover by role. None of these require touching the books. All of them are observable through interviews, walk-throughs, and the systems the company already runs. Build five of them in the first thirty days and you have a parallel dashboard the seller didn’t have to give you.
Then make the data case. When you walk into the next sit-down with the seller and you can say, “Your shift two has 2x the rework rate of shift one — here’s what it’s costing you,” something shifts. You found a problem they couldn’t see from their own P&L. The privacy reflex starts to soften. They open up because you’ve earned the right to ask. Documenting impact requires financial access — but the insight doesn’t. The insight comes first. The access follows.
This is hand-to-hand combat. It’s not the kind of work that fits on a slide. It’s calling the operations manager at 7am to ask how she actually counts a “completed job.” It’s spending two hours in a warehouse with a clipboard. It’s pulling fifty random invoices and rebuilding contribution margin on a notepad because the system can’t do it. The framework above is genuinely simple. The execution is the work.
The data ask is also a relationship ask. One operator called it a test of coachability and change capacity — can you get management to share, discuss, and act on operational data openly? How the seller and the management team respond to “show me how a unit gets made” tells you almost as much as the data itself. The team that walks you through their process, points out the gaps, and asks for help is the team you want to keep. The team that stonewalls is telling you something else — and that signal is part of your diligence.
The “go fix it guy” doesn’t get the keys to the data warehouse on day one. He gets a basic P&L and a politely guarded back office. What separates him from the operators who stall is whether he treats that as a stop sign or a starting point.
Where to Start When You Walk Through the Door
Pick one outcome. Not the whole 100-day plan — one outcome. The single number in the value creation thesis the deal doesn’t work without. EBITDA margin expansion. Gross margin recovery. Customer retention.
Now ask: what operational driver, if I could see it weekly, would tell me whether we’re moving the right direction? Unit cost. Cycle time. Win rate. Pick one, not five. Build the simplest version you can stand up in two weeks — a sampled spreadsheet, an interview-based estimate, a clipboard-and-stopwatch operation if that’s what it takes. Get a working signal. Then refine it.
The phrase — can’t change what you can’t measure — is true. The corollary matters more: you can measure more than you think before the data is officially handed over. This is where deal assumptions and operational reality diverge. The data gap isn’t an excuse for inaction. It’s the first thing the value creation plan actually asks of you. If you’re waiting for perfect visibility before you act, you’ve already lost your first 90 days.
Alex Escoriaza helps PE-backed operators build operational visibility before the data is clean — a parallel measurement layer that doesn’t require waiting on IT, the seller, or the next ERP migration. If your 100-day plan is stuck behind a data access problem, reach out.