Six systems, six definitions
One company puts hosting in cost of revenue, another in opex. Both are defensible and the portfolio gross-margin comparison is meaningless until somebody normalises it by hand every quarter.
By structure
The recurring problem in a mid-market portfolio is not that any single company reports badly. It is that eight companies report differently — different charts, different definitions of gross margin, different close calendars — so the consolidated view is assembled by an analyst rather than computed, and every board pack is a reconciliation exercise.
184 consecutive days tied · variance $0.00
The situation
One company puts hosting in cost of revenue, another in opex. Both are defensible and the portfolio gross-margin comparison is meaningless until somebody normalises it by hand every quarter.
One company closes on day five, another on day nineteen. The portfolio view waits for the slowest, and nobody can see what is blocking it.
Mapping happens at consolidation time, differently each period, usually in a workbook maintained by one analyst who is now indispensable.
Every exit or add-on acquisition rebuilds the same normalisation work, because none of it was ever captured as a durable mapping.
A $40M company with a strong controller and a $12M one with a bookkeeper cannot be held to the same reporting standard without doing something about the second.
Requiring every company onto one ERP is the standard answer and it produces two years of implementations, distracted management teams, and at least one failure.
The instinct in most portfolios is a platform mandate: everyone moves to one ERP, then the numbers are comparable. It is coherent and it is expensive — two years of implementations across companies whose management teams have other priorities, with the predictable outcome that at least one goes badly and becomes the reason the programme stalls.
The alternative is to standardise a layer above the ledgers. Each company keeps whatever it runs — QuickBooks, NetSuite, Sage Intacct, Acumatica, Dynamics — and a shared model reads all of them into one set of definitions. The mapping is captured once per company and maintained rather than rebuilt quarterly.
Diligence, at both ends. An add-on acquisition arrives with its own chart and its own definitions, and normalising it into the portfolio view is a mapping exercise rather than a project. At exit, consolidated statements that trace to source transactions turn a multi-week quality-of-earnings exercise into a data-room link.
The second-order benefit is that underperformance surfaces earlier. When every company reports on the same definitions on the same day, a margin drift at company four is visible in month two rather than in the annual review.
Some will. A portfolio company still on QuickBooks Desktop with four entities and a fifteen-day close is a genuine constraint, and the reporting layer will make that visible rather than hide it. But that becomes a targeted decision about one company with a clear business case, rather than a portfolio-wide mandate applied to companies that did not need it.
Where to start
Agree the reporting chart and the metric definitions with the operating partner and two or three controllers. This is the part that determines everything and it is a decision exercise, not a technical one.
Read-only, mapped, reporting live. Start with the best-run companies so the pattern is established before the harder ones.
Each additional company is a mapping exercise of one to two weeks. The weaker finance functions take longer and surface real issues, which is the point.
A newly acquired company is mapped into the portfolio view during the first close rather than after the first annual cycle.
Questions
Tell us the portfolio shape and what you report today, and we will scope what standardising actually takes.