Nobody Opens the Dashboard. And That’s the Dashboard’s Fault.

Go into any marketing team’s BI tool right now and check the view counts. Not the number of dashboards. The number of times anyone opened them in the last 30 days. You already know what you’ll find. A wall of beautifully built dashboards, and maybe three that anyone actually looks at. The rest are a graveyard.

Somebody built each one. Somebody sat in a meeting where a VP said “we need visibility into this,” and an analyst spent two days wiring it up, and everyone nodded when it shipped, and then nobody ever opened it again. Multiply that by two years of quarterly planning and you get forty dashboards, thirty-seven headstones.

The usual explanation is that people are lazy or the data is dirty. Both are convenient because both blame someone else. The real reason is more annoying: the dashboard was built to answer a question nobody was actually asking. It was built to answer a question someone said out loud in a meeting, which is a very different thing.

The gap between what people say and what they need

When a stakeholder asks for a dashboard, they are almost never telling you the real question. They’re telling you a proxy for it. “I want to see channel performance” really means “I have a nagging feeling we’re wasting money somewhere and I want to be able to find it fast when I get nervous at 4pm.” Those are not the same spec. The first one gets you a grid of channels and spend. The second one gets you a way to ask, in the moment, “where did we waste money last month.”

Dashboards are terrible at the second kind of question because they are frozen. You build them ahead of time, against the questions you can predict. But the questions that matter are the ones you didn’t predict. They show up when the number looks weird, when the board asks something specific, when a campaign tanks and you need to know why right now. A pre-built view can’t help you there, because nobody built the view for a question you didn’t know you’d have.

So what happens? You open the dashboard, it doesn’t quite answer the thing, you Slack the analyst, the analyst is in three other meetings, and by Thursday you’ve got a one-off CSV that also becomes a headstone. The dashboard didn’t fail because it was ugly. It failed because it was an answer to yesterday’s question.

Why more dashboards make it worse

The instinct, when a dashboard doesn’t get used, is to build a better one. More filters. More drill-downs. A tab for every team. This is how you end up with the graveyard. Every unmet question spawns a new artifact, and every new artifact needs maintenance, and now your one analyst spends their week keeping forty dead things technically alive instead of answering the live question in front of them.

There’s a tax nobody puts on the invoice: the more views you have, the less anyone trusts any of them. Two dashboards show slightly different revenue because they pull from different sources or use a different date logic, and now every meeting starts with ten minutes of “wait, which number is right.” A wall of dashboards doesn’t create clarity. It creates a committee of disagreeing numbers, and people cope by trusting none of them and going back to gut.

A dashboard nobody opens isn’t a reporting problem. It’s a sign the tool made you predict your questions in advance, and you guessed wrong.

The thing dashboards were a workaround for

Step back and notice what a dashboard actually is. It’s a workaround. It exists because, historically, you could not just ask your data a question in plain language and get an answer. Querying required SQL, or a BI analyst, or a ticket. So we pre-computed the answers we thought we’d need and pinned them to a screen. The dashboard was never the goal. It was the compromise we made because asking directly was too expensive.

That compromise is the part that’s expiring. The reason you built forty static views is the same reason you don’t build them at home for your bank account: you just want to ask “how much did I spend on this thing last month” and get a number. Marketing data has been stuck one generation behind that because the data lived in nine tools that don’t talk, and asking a question across all nine meant a human had to go stitch it together first.

What actually breaks the cycle

The fix is not a prettier dashboard. It’s making the cost of asking a question approach zero, so you stop pre-building answers to questions you’re guessing at. When anyone on the team can type “which channels lost us money in Q2 and what changed” and get a straight answer pulled across every source, three things happen. You stop maintaining a graveyard. You stop arguing about which number is real, because there’s one source. And the analyst goes back to doing analysis instead of playing dashboard janitor.

The tell that you’re ready for this is simple. Count how many of your dashboards got opened this month. If the honest answer embarrasses you, the problem was never that you needed more views. It’s that you were forced to guess your questions in advance, and real questions don’t work that way.

You don’t need to build the fortieth dashboard. You need to be able to ask the thing you’re actually wondering about and get an answer before the meeting starts. That’s the whole idea behind THE DASHBOARD: one place where your fragmented stack becomes one source of truth you can just talk to. Fewer headstones. More answers. Go check your view counts first, though. That number will tell you everything.