Fewer metrics won’t fix a scorecard nobody believes

An operator opens the monthly report from the brand and goes to one number first. Not to read it. To check it against the number sitting in their own system.

If the two match, the rest of the report gets read. If they don’t, the report gets closed, and every report after that one gets closed faster.

There is an argument running through franchise operations right now, and it comes from good people. The network is drowning in numbers. Dashboards, scorecards, point-of-sale reports, marketing analytics, a separate portal for every vendor. Brands spend real money standing all of it up and then find out almost nobody opens it. The fix on offer is simplification: find the handful of numbers that actually move unit performance, put those in front of the owner, coach against them, let the rest go.

I agree with the observation. Brands do buy reporting the network never opens. I think the diagnosis underneath it is wrong, and wrong in a way that makes the cure expensive.

Operators don’t stop logging in because there are too many numbers. They stop because the numbers disagree.

The moment it actually breaks

Somewhere in the last two years, an owner in your network pulled a lead count off the marketing dashboard, then pulled what looked like the same count out of the booking system, and got two different figures for the same week. Not off by a rounding error. Off by enough that one of them had to be wrong.

So they asked somebody at headquarters, and the honest answer was some version of “those are counted differently.”

That answer is usually true. It also ends the relationship between that owner and that report.

What the owner learned in that exchange was not which number was right. It was that nobody had checked. That is the durable damage. Not the gap, the discovery that it had been sitting there in front of the whole network and that it took an operator to find it.

Why simplifying on top of that makes it worse

Two things happen.

Six numbers that disagree with the unit’s own systems is still six numbers that disagree. Nothing about the truth of any of them changed. The count went down, and the count was never the complaint.

The second one is less obvious. Those extra numbers were doing work. A careful operator triangulates: if the lead count looks high, they check it against appointments and against revenue, and the three together tell them whether to believe any of it. Cut the scorecard to six and the cross-checks go with everything else. The owner can no longer catch the contradiction, which is not trust restored. It is the same contradiction with nobody positioned to see it, until it surfaces some other way and lands worse, because by then it looks buried on purpose.

Simplification on top of unreconciled sources is progress you can demo and can’t defend.

What the Framework asks instead

The Functional Technology Framework, the lens I run this work through, judges technology by the function it performs for the business rather than by the category it sits in. So the question about a franchisee scorecard was never how many metrics belong on it.

The function of a scorecard is to change what an owner does on Monday.

A number the owner doesn’t believe can’t do that job. Not at forty metrics and not at six. So counting metrics was never the path. What matters is whether the number matches what the operator already sees in their own system, and whether somebody can explain the gap in one sentence when it doesn’t.

Reconciliation, which is the unglamorous part

Reconciliation here means proving a number lands the same value in every system that shows it, and where it doesn’t, deciding on purpose which version is right. Most of the work is not technical.

Pick your handful of numbers. Everybody already agrees on that step.

Then, for each one, list every system in the network that displays it or displays something with a similar name. That list runs longer than the person making it expects, and it usually includes a spreadsheet.

Pull the same period out of each and write down the deltas, the gaps between what the systems say.

Then run each gap to ground. There are really only three causes. Definition: a lead counted at form submission in one place and at qualification in another. Timing: a week that starts Sunday in the point-of-sale and Monday in the reporting layer. Population: one system still counting a location that closed in March.

Then the step that gets skipped. Decide which version is right, write down why, and make every other system either match it or stop showing the number.

That last clause is the entire job. Brands will run the audit, produce a genuinely good document explaining every discrepancy, and leave the systems exactly where they were. An explained contradiction is still a contradiction.

Why nobody funds it

Because it has no output. You spend six weeks and end up with the scorecard you started with, except now it is right. No launch, no screenshot, nothing to present at the convention. Meanwhile the simplification project has a redesigned dashboard and it demos in four minutes.

I understand why the dashboard wins. I have watched it win. It is still the wrong order.

Where I get this from

Both chairs, which is usually the sign it is real.

As a multi-unit owner, me and about 25 other owners in our region were frustrated enough with the reporting coming out of the platform we all had to run that I went and built my own. The group’s reaction is what told me there was a business in it.

Then I was on the other side of it, running that platform for franchise brands. The part I did not expect: the connector was never the hard part. Pulling the data was a week. Getting two systems to report the same number for the same week, and explaining in one sentence the ones that still didn’t, took months. That was the part that decided whether a network believed the output.

What I’d do

Thought starter, not a project plan.

If you are about to cut the scorecard from forty numbers to six, do it. The instinct is right. But pick the six, and then, before the network ever sees them, pull each one out of every system that displays it and write down what you get.

If they match, ship it. You spent a week confirming you were in better shape than you thought.

If they don’t, you found the actual project, and it is cheaper than the dashboard redesign you were about to fund. It is also the one that gets the login back.

If you want to know what that reconciliation pass looks like in practice, I’m happy to walk through it.