Somewhere in your portfolio there is probably a company that underwrote 150 to 250 basis points of margin expansion, reports green on every KPI, and is behind plan anyway. Nobody can say exactly why. The board pack gets discussed, the tracker gets reviewed, and the gap persists.
Imagine this hypothetical company. A specialty industrial distributor, roughly $85M of revenue, seven branches, one add-on acquired 18 months ago that still runs its own ERP. Its board pack reads:
- Revenue at 103 percent of plan.
- Gross margin at 34.2 against a 33.0 target.
- Rebate income at 101 percent of plan.
- Top ten customer concentration comfortably under the watch line.
- Every branch is green on margin.
- Six packs delivered on time, six months running.
Every number is honestly produced but that’s what makes it dangerous. A dashboard can only show what the data underneath can produce, and at most mid-market companies the data cannot produce anything except green.
Five mechanisms do the hiding:
Accuracy. The number is right in the system and wrong
in the world.
That 34.2 percent margin is computed from the ERP standard
cost file, which is list cost. Freight and duty are allocated at period level,
never to the SKU. A third of active SKUs carry negotiated exception prices
maintained in branch spreadsheets, untouched for an average of 19 months
through a period of known input cost inflation. The real drag on affected
revenue is 80 to 140 basis points, and the instrument measuring the margin
thesis cannot see the leak it was bought to find.
Consistency. The systems disagree about the facts, and
the pack picks one.
Concentration sits at 28 percent because there is no
canonical customer master. The same national account holds different records
across two ERPs and the CRM, so it counts as several customers. Deduplicate the
top accounts and concentration crosses 35. Same question, different answer
depending on which system you ask. Nobody chose the flattering answer. The
report was simply built on the system that gives it.
Granularity. Green at the level reported, unknowable at
the level the plan requires.
Every branch is on plan. But the value creation plan was
priced at customer and category level, and contribution margin at that level
cannot be produced, because cost never attaches below the period. The business
reports where it can see, the plan lives where it cannot, and the gap between
those two altitudes is where margin expansion theses go to die.
Lineage. Nobody can trace the number to the source.
Rebate income is on plan, and both the number and the plan
come from the same three manual workbooks, with no tie-out between negotiated
entitlement, purchase receipts, and claims filed. When we reconstructed one
quarter for the two largest supplier agreements, claims ran below entitlement.
Extrapolated, that is $150K to $400K a year left uncollected, sitting behind a
green tile. The metric agrees with the plan because they share the same blind
spot.
Repeatability. The green is a person, not a process.
Six packs on time reads like operational discipline.
Underneath, the close is a manual mapping between two charts of accounts, held
together by two and a half full-time equivalents of reconciliation work, and
the only person who knows the add-on’s close retires this year. The metric
measures heroics. Heroics do not survive diligence, and they do not survive
retirement or leaving parties either.
None of this stays private forever. EY’s 2026 exit
readiness study makes the point that buyers now separate AI activity from AI
strategy, and warns that without clean, accessible, well-governed data, claims
made in the equity story become difficult to evidence in diligence.
In a market where DPI has displaced IRR as the metric that
raises funds, a value creation story you cannot evidence is a discount you have
already accepted. You just have not been told the number yet.
The good news is that finding this takes questions, not a
platform. Five of them, one per mechanism.
1.
Which of our KPIs would change if we recomputed
them from source rather than the reporting layer?
2.
Do our systems agree on who our customers are,
and which answer does the board pack use?
3.
What level of granularity is the value creation
plan priced at, and can the business report at that level?
4.
Can anyone show where each number on the pack
came from without opening a spreadsheet?
5.
Which green numbers are produced by a person
rather than a process, and what happens when that person leaves?
Ask them at the next board meeting and watch which ones
produce a pause.
The distributor in the hypothetical company was not run by
fools, and its dashboard was not a lie. The dashboard was green because green
was the only color it could produce.
Graeme Crawford is the founder of Crawford McMillan, a data
strategy consultancy that helps PE firms and their portfolio companies turn
management-grade data into investor-grade data. He writes Inside The Data Room, a
weekly memo for operators who know the data is the deal, and hosts the PE Data Guy podcast.
