A client handed us a marketing dashboard last quarter with 38 widgets on it. Impressions, reach, sessions, bounce rate, follower growth, a dozen channel breakdowns. It was genuinely impressive to look at.
We asked one question: which of these changed a decision you made in the last 90 days?
The honest answer covered about four of them. The other 34 were there because the tools produced them, not because anyone used them.
That is the test we run on every dashboard we inherit. A metric earns its place only if a realistic move in it — up or down — would change what you do next week. If nothing changes, the number is decoration. Worse, it's decoration that costs attention, because every vanity metric on a screen is a real metric someone isn't looking at.
Three numbers get deleted on day one, almost every time. Here's each one, why it survives despite being useless, and what we put in its place.
1. Impressions and reach
Impressions measure how many times something could have been seen. Not seen. Could have been. A number that goes up when you spend more and down when you spend less, with almost no relationship to whether the right people encountered you.
The failure mode is subtle. A campaign racks up two million impressions and the report reads like a win. But if those two million are the wrong audience — cheap inventory, broad targeting, a viral post that reached everyone except buyers — the number rewards exactly the behavior you want to stop.
What we track instead: qualified reach. Impressions filtered to your actual target segment, or share of search on the terms your buyers use when they're in-market. One tells you that a billboard exists. The other tells you whether the people who can pay you are seeing it.
2. Total sessions
Traffic is the most seductive metric on the list because it always feels like progress. More sessions, more people, more top-of-funnel. And it climbs reliably when you publish more or bid harder.
The problem is that total sessions average your best and worst visitors into a single line. A post that draws 5,000 students who will never buy sits in the same number as 200 sessions from procurement leads at your target accounts. The line goes up. Pipeline doesn't move. Everyone congratulates the content team.
What we track instead: sessions from target segments, and conversion-qualified traffic. We segment by the audiences that can actually become revenue, then watch whether those sessions grow and what share of them take a meaningful action. A flat total with a rising qualified segment is a good quarter. A rising total with a flat qualified segment is a treadmill.
3. MQL count
The marketing-qualified lead was invented to give marketing a number it controlled. That's precisely the problem: a metric you can hit by loosening your own definition is a metric that measures your generosity, not your results.
We've watched teams celebrate a record MQL month while sales quietly ignored 90% of them, because "qualified" had drifted to mean "filled out a form." The count went up. The definition went down. The two moved together and nobody said so out loud.
What we track instead: pipeline created and sales-accepted opportunities. The moment a number requires sales to accept it, marketing can no longer inflate it alone. That single dependency does more for alignment than any amount of shared-goals workshopping. When marketing's headline number is pipeline that sales agreed to work, the two teams are finally looking at the same scoreboard.
The dashboard we actually keep
After the deletions, what's left is short — usually five or six numbers:
- Qualified reach or share of in-market search
- Conversion-qualified sessions from target segments
- Conversion rate at the one or two steps that matter most
- Pipeline created, attributed honestly to source
- Sales-accepted opportunities
- Cost to produce a unit of pipeline, by channel
Every one of these passes the test: a realistic swing changes next week's plan. Reach drops, we fix targeting. Qualified conversion falls, we fix the page. Pipeline stalls at one source, we move budget.
The point was never to have fewer metrics for its own sake. It's that a dashboard is a decision tool, and most of them have quietly become a comfort object. Delete the three that comfort. Keep the ones that make you do something.