Your KPIs Were Picked Because They Were Easy to Get, Not Because They Matter

Here is a question worth sitting with for ten seconds: who actually decided the metrics on your marketing scorecard?

Not who presents them. Who chose them.

The story everyone tells is that the metrics came from strategy. Leadership sat in a room, argued about what mattered, and landed on a set of numbers that reflect the business. That happens sometimes. Mostly it doesn’t.

What actually happened is that someone was asked to build a report on a Thursday, opened the four tools they had credentials for, and picked whatever exported cleanly. That set of numbers got a name, then a template, then a recurring calendar invite, then a quarterly target attached to it. Five years later it is presented to the board as if it were the result of deliberation.

It was the result of an export button.

Availability bias, but for reporting

Psychologists have a name for judging importance by ease of recall. The reporting version is simpler and dumber: a number becomes important because it was already sitting there.

You can spot it fast. Look at your monthly deck and ask, for each metric, what tool it comes from. Almost always the answer is the same four or five tools, and almost always they are the tools with the friendliest CSV export. Sessions, because the web analytics tool hands them over. Open rate, because the email platform puts it on the front page. MQLs, because someone built that scoring rule in 2019 and the field is right there on the contact record. Pipeline created, because the CRM report builder can do it in three clicks.

Now ask what is missing. Usually: anything that requires two systems talking to each other. Which customers came back a second time. What the ad spend actually did to revenue rather than to clicks. Whether the accounts your sales team calls “good fits” behave any differently after they buy. How long the gap is between first touch and closed business, by segment, in a way you’d bet money on.

Those questions are not harder to think of. Anyone can think of them. They are harder to assemble, because the pieces live in different vendors’ databases with different customer IDs and different definitions of a week. So they quietly fall off the scorecard, and the scorecard becomes a map of your integration difficulty rather than a map of your business.

The metrics you can get are the metrics you optimize

This would be a mildly interesting observation if it stopped at reporting. It doesn’t, because measurement is not passive. Whatever is on the board deck becomes a target, and whatever becomes a target starts steering real decisions about real money.

So the chain goes: the number was easy to export, therefore it went on the scorecard, therefore it got a target, therefore someone’s comp depends on it, therefore the team optimizes for it, therefore the strategy is now downstream of a CSV export.

Watch how that plays out. If traffic is on the deck and revenue per segment is not, you will get more traffic. Some of it will be worthless, and nothing in the reporting will tell you which part. If open rate is on the deck, you will get subject lines engineered for opens by people who know perfectly well that opens stopped meaning much once mail clients started prefetching images. If MQL count is on the deck, you will get MQLs, and the definition of MQL will drift downward every quarter the target gets missed, because the definition is the only variable anyone can actually move.

Nobody in this story is stupid or lazy. Every individual step is reasonable. The person building the first report was being efficient. The person who set a target on it was doing their job. The team hitting the target is doing what they were paid to do. The failure is structural: the easiest data won, and nobody ever revisited whether it was the right data, because revisiting it means an integration project and integration projects mean a quarter of nobody having any numbers at all.

Two questions that expose it

You don’t need an audit for this. You need two questions, asked in a room with the people who build the reports.

One: if we could get any number instantly, which numbers would we put on this deck? Not “which would be nice.” Which would actually change what we do next quarter. Write that list down without looking at the current deck.

Two: what is the overlap between that list and what we report today?

In most mid-market marketing teams the overlap is uncomfortable. Half the current deck is on the wish list, which is fine. The other half is there because it has always been there and removing it would raise questions. And the wish list has three or four items that have never appeared in a report, not because anyone decided they weren’t important, but because getting them meant a ticket to a data team that has a six-week queue and bigger problems.

That gap is the actual state of your analytics. Not the tool count, not the dashboard count. The gap between the numbers you’d choose and the numbers you can reach.

Why this gets worse as you add tools

The intuitive fix is more tooling, and the intuitive fix makes it worse. Every new platform ships with its own reporting layer, its own definitions, and its own convenient export. So each addition increases the supply of easy single-system numbers while doing nothing for the cross-system questions that were the actual gap.

You end up with more metrics available and fewer answers, which is the specific flavor of frustration that makes people say things like “we have plenty of data, we just can’t get anything out of it.” They’re right. The data volume went up. The reachable question set did not.

And the cost is invisible on any invoice. There is no line item called “decisions we did not make because assembling the number was a two week project.” It shows up as budget defended by whoever had a chart, campaigns that ran a quarter too long, and a persistent low-grade feeling that the team is busy but nobody can say what’s working.

The point

Your scorecard is a historical artifact. It records which systems were easy to query at the moment somebody had to build a report, plus whatever got bolted on since. Treating it as a considered statement of what the business cares about is a small, expensive fiction that most companies are running right now.

Fixing it is not a reporting exercise. Nobody’s slide template is the problem. The problem is that the set of questions you can answer is capped by how much of your stack can be looked at together, and until that cap moves, better reporting just means prettier versions of the same convenient numbers.

Move the cap and the whole thing changes. When the CRM, the site, the ad platforms, the billing system, and the email tool sit in one place with one definition of a customer, the metric you’d choose and the metric you can get finally converge. Then the scorecard is a decision instead of an accident.

That’s the entire reason THE DASHBOARD exists: one place where your stack lives together, so you can ask the question you actually have instead of the one your export button allows. Worth a look if the gap between those two lists is bothering you.