Justin PenningtonJustin Pennington
All posts

Nobody Opens the Dashboard

Most reporting dashboards in an operating business get a look in the first month and then sit unopened. The fix is not another chart. It is who gets told when something is off, and what they do next.


A dashboard goes live and people look at it for a week. Someone walks through the tiles. A number gets quoted in a meeting. Then the ordinary week returns. The tab stays pinned. The day goes back to texts, a spreadsheet one person trusts, and "did you see that?"

I see that more often than a broken chart. People stopped coming back because the next hour of work did not require the visit.

I run Infraxio from Ponte Vedra, Florida. Growth systems, the business hub, bid workflows, dock and yard work, and the custom systems we build all get the same request: can we have a dashboard. The useful question is quieter. When something is off, who gets told, and what do they do next.

The First Month Is the Life of the Dashboard

The launch looks like adoption. The tool is new, so opening it feels like using it. A few people click through, the numbers look familiar, and they go back to the job.

A month later the split is plain. The person who bought the dashboard still thinks of it as how the company sees itself. The people who could change an outcome have stopped opening it. The screen asks them to leave the work, read a picture, and decide whether any of it is theirs.

CRM and ERP rollouts already taught this. People return to the tools they already had open, and the system of record becomes a place leadership visits. I wrote about that in nobody lives in the CRM. A reporting layer has the same shape. If the only way to learn that something is wrong is to go look, the system is built for the people who have time to look.

Business intelligence made the picture cheap. A Monday email full of screenshots still left the picture outside the job. After the first month, checking starts to feel optional, because the tiles are calm often enough. On the day a tile is not, the person who needed it has moved on. The day runs on memory and a question in the hallway.

A true chart can still sit unread. I have written about why reporting lies when the numbers do not agree. Clean data does not rescue a screen nobody opens.

Who Gets Told, and What They Do

Before anyone draws a tile, I want three answers.

What counts as off. A lead with no first touch. A bid section with no owner as the date gets close. A door that slipped. An invoice that did not go out. A task sitting past the point where someone should have moved it. If everything deserves a chart, nothing is an exception. The line has to be specific enough that an operator can name the miss.

Who hears it. One named person, and a backup for the day that person is out, on a site, or already in the work. A channel called leadership is not an owner. A shared inbox is not an owner. Several people can see a red tile and each one can assume another has it. That is the same hole that loses a new inquiry.

What they do next. A call, a decision, a task with a date, a hold, a no-bid. The exception should arrive as work, in the place that person already has open, with enough of the record attached that they do not have to hunt for why it showed up.

A wall of charts waits for someone to wander past. A business should still know the one number that matters. A number that a real decision hangs on can live where a person already works. A wall of tiles will not carry the day.

A Morning Summary Is Still a Report

The temptation is to have AI explain the dashboard every morning. A paragraph in a chat. A digest before the meeting. For a few days it feels like the dashboard has a reader. A summary of everything still asks a human to find the one line that required a decision. You have the chart again, in prose.

Name the thresholds instead. Let the system watch them. When one trips, tell the owner, attach the record, and put a next action where that owner will treat it as work. Leave the calm parts of the day alone. Calm does not need a narrator.

That is the AI use worth building into operations. Flag the exception and route it, so the owner does not have to go looking. The agent does not replace the owner. I have written about putting that work on the task board, where a person reviews it the same way they review anyone else's work. If the message is a task with a name on it, someone can act the same day. A summary of the whole business can wait for a meeting that asked for one.

Build It Where the Work Already Is

This is the rule across the products and the custom systems we build.

IFX Hub is the business hub. Deals, tasks, client delivery, and the record live in one place, and a person or an agent can hold a task. A follow-up that stalled, or a delivery that stopped moving, should land as owned work on that board.

Growth7 is the growth and lead system. An inquiry matters when a stall, or a reply that needs a person, becomes a next action for someone named, with the context still attached. A funnel picture after the fact does not call anyone back.

IFX Bid is how a pursuit gets owned. The submission date is the part everyone can already see. The misses are internal: a section with no owner, a price that is still a question, a pursuit that should have been declined and is still being written. An owner and a next action are how a response goes out.

DockOps is the dock and yard. The supervisor already knows what the day was supposed to be. What they need is the break: a door that slipped, a yard that is blocked, an appointment that no longer matches the truck that showed up. The exception belongs in front of the person who can move the door.

Custom systems follow the same test. When the daily job is a price, a dispatch, a pick, or a client update, the screen where that job happens is where the exception belongs. A separate reporting product is a second place to go, and second places lose after the first month.

Start From the Miss

If the dashboard is already furniture, take the recent misses you can still name. The lead that went cold. The invoice that waited. The bid that went out thin. The dock that backed up. For each one, write who should have heard, when, and what they should have done next. If those three are fuzzy, another chart was not going to catch it.

Put that path where the work already is. One owner. A backup. A line that means something is actually off. A next action with a date. Let a model watch the line and route the exception.

From Ponte Vedra, with operators across St. Johns County and Jacksonville, the purchase is rarely the gap. The shops that stay current decided, before the screen was pretty, who gets told when the day goes sideways and what that person does next.

If the only way to know something is wrong is to open a dashboard, you will find out late.