Why I'm skeptical of EHS dashboards.

The first time I sat with an EHS director and walked through someone else's safety dashboard, she had it pulled up on her phone in a Carhartt jacket, standing in a windowed conference room above a yard. The numbers were green. The trend lines were going the right way. TRIR was below the rolling 12-month baseline. She looked at me and asked, "What does this actually want me to do today?"

I didn't have a good answer. The dashboard didn't either.

The part I never got past.

EHS dashboards are very good at showing you what already happened in a shape safe enough to print on a slide. They are not very good at telling you what is about to go wrong, or what to do in the next hour to keep it from going wrong. The two questions an EHS leader actually carries to work — "where is the next incident going to come from?" and "what should I be doing about it right now?" — neither one is well served by aggregate numbers in a frame.

I have watched programs go all-in on dashboards. The investment is real. Six months later, the EHS director still opens the inspection backlog and the open-CAPA list first thing in the morning. The dashboard is a screen in the conference room that someone wipes the dust off before a quarterly review.

So I am skeptical. Not of measurement — measurement is the job. Skeptical of the dashboard as the surface where the work actually gets done.

What we built instead.

Kinetiq Nexus does have charts. We are not anti-numbers. But the home screen of the platform is not a dashboard. It is a queue — what needs attention this shift. Inspections that are due. Findings without an owner. CAPAs that have rolled past their target close date. Broadcasts that haven't been acknowledged on a Pulse kiosk. Anything that came in flagged by the analytics engine as outside its category baseline.

Numbers are downstream of the queue, not the headline. If you want to see your TRIR trend, your category-level safety indicators against the federal benchmark, your CAPA aging — they're there, one click in. They earn their place by being useful when you have a specific question. They do not get to live on the home screen and dominate the attention of the people who run the program.

The threshold alerts are tied to real baselines — federal datasets filtered to your mine type, scored at the category level — not to a green-yellow-red gradient someone picked because it looked good in a deck. When the analytics engine surfaces an alert, it is because a category metric crossed a baseline, and the surface text says what the baseline is, how far you are from it, and what to look at next.

What we got wrong.

I almost shipped a dashboard-as-home-screen. The early prototype had a 4x3 grid of KPI tiles — green where you wanted green, a few sparklines, a leaderboard for sites with the lowest TRIR. It looked good. It demoed beautifully.

The same EHS director from the conference room saw it, sat with it for ten minutes, and told me she would never open it past week two. Not because the design was bad. Because she still had to go somewhere else to figure out what to do today.

I rebuilt the home screen around the queue model the next month. The KPI tiles still exist; they live one level deeper. The first thing you see when you log in is a list of things waiting on you, in priority order, with deep links into the workflow that closes them. The dashboard became a destination, not the headline. It made the product worse in screenshots and better in daily use. I would make the same trade again.

What we deliberately don't do.

A few things we said no to, even after they were on the wishlist:

• No TRIR ticker on the home screen. The number is real; the surface signal is misleading. A ticker invites the wrong response — staring at the metric instead of working the queue that moves it.
• No site-vs-site leaderboard by default. A leaderboard rewards underreporting at the bottom and rewards looking good at the top. We will show site-level comparison when the comparison is the question; we won't make it the framing.
• No "executive view" toggle. The CSO and the site lead should see the same primitives — different scope, same shape. If the executive view is more honest than the operator view, the operator view is the problem; fix that instead of building a second screen.
• No vibes-based color gradients. If a metric is yellow, the surface tells you what threshold made it yellow and what the actual category baseline is. Color without a number is just a feeling.

What I am watching next.

The thing I want to be honest about: the queue model has held up so far on programs that own a handful of sites. An EHS director managing eight sites with five hundred contractors could see something different, and the model has to either hold up at that scale or get rebuilt before it breaks. Programs that size are starting to come on, and I am watching their usage closely. If you are running one and you want to compare notes, please write.

And if you have shipped an EHS program with a dashboard at the center and seen how the team actually uses it after the rollout glow — I would like to hear from you. Some of what we are skeptical of, we are skeptical of for our own reasons. A lot of it we are skeptical of because the people in the field told us to be.

Thanks for reading.

— Jeffrey