14 reports, 4 decisions

The client was a multi-crew home services company — call it a mid-size electrical and HVAC contractor, though the specifics have been changed. Three years of growth had taken them from one truck to five crews, and their reporting had grown the same way most reporting grows under pressure: one report at a time, added whenever someone asked a question that the current stack could not answer.

By the time Orivian was brought in through a QuickBooks dashboards engagement, the owner was pulling fourteen separate reports every month: a full P&L, a balance sheet, four different QuickBooks class reports (one per crew), an A/R aging summary, an A/P aging summary, three spreadsheets a bookkeeper rebuilt by hand from CSV exports, a payroll summary, a sales-by- service-line report, and a job costing report that hadn’t been opened in four months because nobody remembered exactly how to read it.

None of that was incompetence. Every one of those fourteen reports had been added for a real reason, at a real moment, to answer a real question. The problem was that nobody had gone back and asked whether the reason still held.

The method was simple to describe and slower to actually do: for each of the fourteen reports, write down the specific decision it was supposed to inform. Not “stay informed” or “keep an eye on things” — an actual decision, the kind with a yes/no or a number attached. “Do we have the cash to make payroll in two weeks” is a decision. “See how the business is doing” is not.

Four reports passed that test cleanly. The P&L informed pricing and hiring decisions. The A/R aging summary informed who to call about an unpaid invoice. The four per-crew class reports, taken together, informed which crew was actually profitable versus which one just looked busy. And the job costing report — the one nobody had opened in months — turned out to be the only place that answered “should we take another job like the one we just finished,” which was a question the owner asked out loud in the first meeting, unprompted, before ever seeing the report that already answered it.

The other ten reports fell into two buckets. Five were duplicates in disguise — the same numbers as one of the four core reports, recut a different way, kept alive because someone once needed that specific cut and nobody ever asked if the need was still there. The other five were being generated out of habit: pulled every month, filed, and never actually opened to make a decision. They were maintenance work pretending to be reporting.

Once the mapping was done, the rebuild was straightforward. The fourteen reports collapsed into four dashboard views, each one owned by a single decision instead of a single report:

• Owner overview — replaced the standalone P&L for the “are we okay” question, with revenue, margin, and cash trend in one view instead of a report to read line by line.
• Cash flow — replaced the balance sheet plus a manually rebuilt spreadsheet, answering whether cash in will cover payroll and vendor bills before the gap shows up in the bank balance.
• Job profitability by crew — replaced the four separate class reports and the job costing report nobody opened, ranking every crew and job by margin instead of requiring four side-by-side report windows.
• AR & collections — replaced the aging summary and the follow-up spreadsheet, ranking overdue invoices by urgency instead of by whoever remembered to check.

These are, deliberately, the same categories covered in Orivian’s QuickBooks dashboard examples guide — not because every business needs exactly these four, but because these four decisions come up often enough that they are close to a default starting point before a business’s specific reporting quirks get layered on top.

The bookkeeper got roughly four hours a month back that had been going into rebuilding spreadsheets from CSV exports. The owner started actually reviewing job profitability weekly instead of quarterly, because it took thirty seconds to open instead of twenty minutes to reconstruct. And the crew that “looked busy” on the old per-crew reports turned out to be running the thinnest margin of the five — invisible in fourteen separate reports, obvious in one ranked view.

None of that came from more data. The business already had every number it needed, spread across fourteen places. The change was making each view answer to exactly one decision, and being willing to delete a report once its decision no longer needed asking.

You do not need Orivian in the room to start this. List every report you currently pull on a recurring basis. Next to each one, write the specific decision it is supposed to inform — if you cannot write one down in a single sentence, that is the answer. Then group what is left by decision, not by report source, and merge anything answering the same question. What remains is usually somewhere between three and six views, not fourteen.

The hard part is rarely finding the four decisions that matter. It is admitting that the other ten reports, however carefully built, were never actually one of them.