Run The Region

The Regional KPI Dashboard People Actually Read

Fewer metrics. Unit-ranked. Phone-first. One click. Updated daily. The design constraints that separate a dashboard general managers open every morning from a report only the regional director reads.

Kiyong (Eric) Lee·Operations Executive·August 2026·9 min read

The short version

The test of a dashboard is not what is on it. The test is who opens it. If the general managers are not opening it every morning, it is a report, not a dashboard.
Five to seven metrics. Not fifty. More metrics means fewer opens.
Unit-ranked worst to best on every line. Peer visibility does more work than manager pressure.
Phone-first, one click, no VPN, no login gymnastics. If it takes three taps, it will not become a habit.
Every metric on the daily dashboard has to be one the general manager can move in the next 24 hours. Everything else belongs on a monthly view.

The first regional dashboard I inherited had 47 metrics. It was updated every Wednesday for the prior week. It required three clicks and a VPN to open from a phone. And exactly one person in the region opened it: the regional director I was replacing. It looked comprehensive when it was projected on the boardroom monitor. It moved nothing at the unit level, because no general manager had ever really used it. When I asked one of them what their labor variance was last week, they told me they did not know, and they were embarrassed about it. They should not have been. Nobody had built them a tool that would let them know.

The dashboard is a small thing that decides a large amount. If the general managers open it every morning, the region has a shared reality. If they do not, everyone is running on their own private numbers, and the weekly meeting becomes a debate about what is even true. Getting the dashboard right is one of the highest-yield operator moves you can make, and getting it wrong is one of the most common ways operators waste the first year of a new job.

The test is who opens it

Before you argue about which metrics belong on a dashboard, argue about who is going to open it. That is the real test. A dashboard with the perfect set of metrics that only the regional director opens is a report. A dashboard with a merely good set of metrics that every general manager opens every morning is an operating tool. The second is worth ten of the first.

The instrument you want to design against is not "does this cover the P&L." It is "will the general manager at unit 14 open this on their phone at 7:15 a.m. tomorrow while they are drinking coffee." If the honest answer to that is no, the design is wrong, no matter how comprehensive the data is.

Five design principles

Fewer metrics

Five to seven. This is the single hardest constraint to hold because everyone in the organization wants their metric on it. The board wants EBITDA per unit. The controller wants prime cost. Marketing wants NPS. HR wants turnover. Add all of them and you have the 47-metric dashboard I inherited, and nobody opens it. The discipline is that a metric only earns its way onto the daily dashboard if a general manager can move it in the next 24 hours. Everything else goes on a weekly or monthly view, or nowhere.

Unit-ranked, worst to best

Every metric on the dashboard should show every unit in the region, ranked from worst to best on that line. Not alphabetical. Not by store number. Worst to best. This one design choice does more work than any coaching program. Nobody wants to be at the top of the labor variance list. Peer visibility is a stronger motivator than manager pressure, and it works without you having to say a word.

Phone-first, one click

The general manager is on the floor. They are not at a desk. If opening the dashboard requires a laptop, a VPN, a login, and three clicks, it will not become a morning habit. It will be opened once, filed as too much work, and forgotten. Design the dashboard for the phone first. One bookmark. One tap. If your data platform cannot serve that, either fix the data platform or export a simpler view that can.

Updated daily, ideally by 7 a.m.

A weekly-refreshed dashboard is a stale dashboard by Tuesday. The general manager needs yesterday's number this morning so they can make a decision this afternoon. The moment the dashboard is more than 24 hours behind, it becomes something you look at for context, not something you use to steer. Steering data has to be fresh.

Every metric has an owner

Each of the five to seven metrics should have a clear owner at each unit. Labor variance is the general manager's. Food cost variance is the chef's or kitchen manager's. Comps and voids are the front-of-house lead's. The owner is named on the dashboard next to the number. This is what turns a metric from a report line into a job responsibility.

Fig. 1 · Same P&L. Different design. Different behavior.

What belongs on the daily dashboard

Here is the short list that has worked for me across concepts, at Hana Group's 21 franchise units across 6 states and at Zareen's five Bay Area locations. Six metrics. Not more.

Labor as a percent of sales versus plan, yesterday. The single most important operator metric. Not weekly labor. Yesterday's labor. If it is off by two points, the general manager should know why before lunch.
Food cost variance versus theoretical, yesterday. Requires a real theoretical, which is work to build, but once built it is the metric that catches waste before it becomes a monthly write-down.
Sales versus plan, yesterday. Not sales in the abstract. Sales against the number the general manager committed to on Monday. Variance is the story.
Comps and voids as a percent of sales, yesterday. The metric that catches guest recovery being used as a work-around for a service problem or a kitchen problem.
Top three selling items, yesterday. Not for accounting. For the kitchen. If the mix has drifted, prep has to catch up before the shift.
Sales per labor hour, yesterday. The productivity ratio that makes labor variance interpretable. Two units with the same labor percent can have very different sales per labor hour, and that is the tell.

That is the daily dashboard. Anything else you want to see, put on a weekly view. Anything you want quarterly, put on a quarterly view. Do not put it on the daily.

What to leave out

The list of what should not be on the daily dashboard is longer than what should. Some of it will hurt to leave off. Do it anyway.

EBITDA per unit. A quarterly metric. Not something the general manager can move in a day. Belongs on the monthly P&L review, not the daily dashboard.
NPS or guest satisfaction score. A weekly metric at best. A general manager cannot influence today's NPS with today's shift. It belongs on the Monday standing meeting dashboard, not the daily.
Turnover. A monthly view. Do not confuse it with the daily view.
Marketing spend versus plan. Not a unit-owned metric. Leave it off entirely.
Any metric that requires interpretation. If a general manager cannot look at the number and immediately know whether it is good or bad, the metric is not ready to be on the daily dashboard.

What makes a metric operator-owned

Every metric is either operator-owned or it is not. There is no in-between. An operator-owned metric passes three tests, all three of them.

The operator can name their number from memory. Not roughly. Exactly. If you ask a general manager what their labor percent was yesterday and they say "somewhere around 28," they do not own the metric. They are watching it.
The operator can name what moved it yesterday. "Two cooks got called in on Sunday when we should have flexed down for the weather." That is ownership. "It was busy" is not ownership.
The operator can name what they are doing about it today. Not next week. Today. Ownership means the metric has consequences in the operator's calendar this afternoon.

A metric that a general manager cannot name, cannot explain, and cannot act on is not a metric they own. It is a report line the regional director is watching, and it will not move.

The bad dashboard I inherited, and what we changed

The 47-metric weekly-refresh VPN-required dashboard I mentioned at the top of this piece was, in fairness, a well-intentioned effort. Someone had spent months building it. It was pretty. It had heat maps and drill-downs and a filter panel with 12 dimensions. In the boardroom, projected on the wall, it looked like an operations program. In the field, it was invisible.

Here is what we changed, in order:

Deleted 41 of the 47 metrics. Kept the six that a general manager could move in 24 hours.
Moved from weekly refresh to daily refresh, by 7 a.m. This required rebuilding one data pipeline. It took three weeks. It was worth it in the first week.
Ranked every unit worst to best on every line. Removed alphabetical sort. Made rank visible without a click.
Built a mobile view. Not a responsive version of the desktop view. A separate view designed for a phone, with the six metrics on one screen and no need to scroll.
Removed the VPN requirement. Put the dashboard behind a single sign-on link that opened in the phone browser.
Made the dashboard the source for the Monday standing meeting. This is the mechanism that pulled the general managers into the habit. If the meeting used the dashboard, they had to open the dashboard.

Within a month, the dashboard went from one opener to twenty. Within a quarter, general managers were referencing their unit's rank on the standing calls without being asked. Within two quarters, the labor variance across the region had compressed measurably. The dashboard did not cause the compression. The dashboard just made the compression visible to the people who could act on it.

What changes when the general managers own the numbers

Something quieter happens once the dashboard becomes a daily habit, and it is worth naming. General managers start calling each other. Not through the regional director, not on a group chat, directly. The unit ranked worst on labor variance will text the unit ranked best asking what they are doing differently. That is the payoff that no other operating tool produces. Peer-to-peer coaching starts happening without you scheduling it, because the ranking on the dashboard makes it obvious who to ask.

The second change is that the standing Monday meeting gets shorter. When every general manager has already opened the dashboard on Sunday night, the meeting stops being about surfacing information and starts being about deciding what to do with it. At Hana Group, once the dashboard was working, our Monday meeting went from 75 minutes to 45 minutes with more decisions moved. The information was already shared. The meeting could do the harder work.

What I would tell an operator building their first dashboard

Three things.

Start with six metrics. Not sixteen. You can always add. Adding is easy. Removing is the hard part, and removing is where trust gets damaged, so start with too few and let the general managers ask for more.
Build for the phone before you build for the monitor. The general manager is not at a desk. If the phone view is an afterthought, the dashboard will be an afterthought too.
Tie the dashboard to the meeting. The habit forms because the meeting requires the dashboard. Without that mechanism, adoption dies at week three.

The point

A dashboard is not a data problem. It is a behavior problem. The metrics matter. The refresh cadence matters. The click count matters. But the thing that separates a dashboard that changes the region from a dashboard that decorates a boardroom is whether the general managers open it, own their numbers, and can name what they are doing about them.

Fewer metrics. Ranked units. Phone-first. One click. Daily refresh. Tied to the standing meeting. That is the whole design. Everything else is negotiable. Those six choices, held with discipline, are what turn a report into an operating tool.

Rebuilding your operator dashboard?

I work with multi-unit groups whose dashboards have quietly become boardroom decoration. If you are trying to figure out which metrics have earned their way onto the daily view and which ones need to come off, I am happy to compare notes.

Get in touch

Questions people ask

What is a regional KPI dashboard?

A regional KPI dashboard is a single view that lets a multi-unit operator and every general manager see how each unit is performing on a small set of key metrics. In a working version, it is updated daily, opens on a phone in one click, ranks the units worst to best on each line, and includes no more than five to seven metrics. The point is not comprehensiveness. The point is that a general manager can look at it in ninety seconds and know exactly what to work on today.

How many metrics should a regional dashboard include?

Five to seven. Fewer than five and it is not covering enough of the P&L to matter. More than seven and nobody reads it. The tell that a dashboard has too many metrics is that the general managers stop opening it. If a metric has to be included but has not been opened in three weeks, it does not belong on the daily dashboard. It belongs on a monthly one.

Which metrics belong on a daily regional dashboard?

Labor as a percent of sales versus plan, food cost variance versus plan, sales versus plan, comps and voids as a percent of sales, and the top three selling items yesterday. In some concepts you add ticket time and guest count. Every metric on the dashboard has to be one a general manager can influence in the next 24 hours. Metrics that only a controller can move do not belong on a daily operator dashboard.

Why do most operational dashboards go unread?

Three reasons, in order. They have too many metrics, so the general manager has to hunt for what matters. They are updated too rarely, usually weekly, so the number is stale by the time it lands. And they require more than one click from a phone to open, which means they get opened once, filed as too much work, and never opened again. Fix all three and the dashboard becomes a daily habit. Fix only two and it stays a weekly report.

What does it mean for a metric to be operator-owned?

A metric is operator-owned when the general manager can name their number from memory, can name what moved it yesterday, and can name what they are doing about it today. If the metric lives only in a report the regional director reads, it is a regional metric, not an operator metric. Operator-owned metrics are the only ones that actually move, because ownership is what turns a number into a behavior.

How do you get general managers to actually open a dashboard?

Make it one click on a phone. Make it updated by 7 a.m. every day. Rank the units worst to best on every line, publicly, so peer accountability does the pulling. And build the weekly standing meeting around the numbers on the dashboard. If the dashboard is the source for the meeting, and the meeting matters, general managers will open the dashboard on Sunday night before the meeting Monday morning. That is how the habit forms.

What is a real example of a bad regional dashboard?

The one I inherited at one job had 47 metrics, was updated every Wednesday for the prior Sunday-through-Saturday week, required three clicks and a VPN to open, and was only opened by the regional director. It looked comprehensive on a monitor in a boardroom. It moved nothing at the unit level, because no general manager had ever opened it and none of them could have named a single number on it. We deleted 40 of the metrics, moved the rest to a daily refresh, and made it one link on a phone. Open rates went from one person to twenty within a month.