An operational dashboard is not a smaller version of a management report. It exists so that somebody on shift changes what they are doing in the next hour, which is a much harder test than looking informative.
This guide covers how to choose the three or four figures that pass that test, how to agree a threshold and an action for each, and how to prove the whole thing on paper before anybody builds a screen.
Key takeaways
- A tile needs a threshold: A number with no value that triggers action is a trend, and it belongs in a monthly review.
- Find the morning spreadsheet: Somebody already rebuilds the dashboard by hand, and those hours are your first saving.
- Name the watcher: The owner is a role on shift, not the department that commissioned the screen.
- Delete what nobody reads: Removing a tile after people grow attached to it is far harder than never adding it.
Start with the business outcome
This works when somebody changes what they are doing during the shift because of what the screen shows. Write down which role that is and which decision, because an operational dashboard without a named reader is a reporting exercise with a faster refresh rate.
Find the spreadsheet somebody rebuilds every morning. It is the dashboard the business already asked for, it has survived contact with real use, and the hours spent rebuilding it are the first saving you can quote.
Four areas to define before choosing features
1. Definition
Write the calculation out in full, including what it excludes, until two teams would produce the same number from it. Most dashboard disputes are definition disputes wearing a technical disguise, and they surface in the meeting rather than in the data.
2. Owner
Name the role that watches this during the working day, not the department that asked for the dashboard. A tile with no individual owner is looked at during the first fortnight and ignored afterwards, which is a slower and more expensive way of deleting it.
3. Threshold
Agree the value at which the number stops being informative and starts requiring action, and record who set it. A threshold agreed in the abstract gets challenged the first time it demands something inconvenient, so write down the reasoning as well as the number.
4. Action
Describe the specific thing the owner does when the threshold is crossed, including what they do when they cannot fix it themselves. A tile without that sentence is a strategic metric that has wandered onto an operational screen.
Decide what to measure
Choose a small set of measures tied to an operating decision. A metric deserves space in the product only when someone owns it, understands its definition, and can take a clear action when it changes.
Manage the most likely risks
| Decision area | Risk to make visible | Practical safeguard |
|---|---|---|
| Definition | The calculation is disputed | Write it out in full, filters included, until two teams produce the same number from it. |
| Owner | Nobody owns the tile | Name the role that watches it during the day, and remove any tile nobody will claim. |
| Threshold | The threshold was never agreed | Ask the owner at what value they would act, and record the reasoning beside the number. |
| Action | There is no action to take | Describe what happens when the line is crossed, including what to do when they cannot fix it themselves. |
Build a roadmap around evidence
1. Watch what people check today
Spend two shifts noting which numbers supervisors look up and where they find them.
2. Keep only the tiles with thresholds
Drop anything the owner cannot attach a value and an action to, which is usually most of the list.
3. Publish on paper for three weeks
Send the surviving figures daily by the quickest means and record who reads them and what changed.
4. Build what proved useful
Automate the tiles that prompted action, and leave the rest in the monthly review where trends belong.
For how these operational numbers connect to a wider digitisation case, see Digitalization in Business: Benefits and Roadmap.
Plan adoption and operating ownership
Dashboards fail at handover more often than at build. Decide before launch who maintains the definitions when the business changes, who answers when somebody disputes a figure, and what happens to a tile that stops being looked at. Without those three answers the screen is accurate on the day it ships and quietly wrong six months later.
Rehearse launch as an operating change
Plan a day-in-the-life rehearsal for an operational KPI dashboard. Use representative accounts, realistic data, the actual staff roles, and one common exception. Move through Definition, Owner, Threshold, and Action while observers record confusion, missing access, unclear ownership, and support questions.
This approach treats adoption as part of the product. It gives the business a practical route to connect each KPI to a definition, owner, threshold, and response workflow while protecting customers and staff during the period when the workflow is still becoming familiar.
A 30-day validation plan: Operational dashboard
Days 1 to 5: watch what people check today. Spend a few shifts noting which numbers the supervisors actually look up, where they find them, and what they do next. Most of a first dashboard already exists as a spreadsheet somebody rebuilds each morning. Find that spreadsheet, because it is a more honest requirement than any workshop output. Note who rebuilds it and how long it takes them, because that time is the first saving and it is easy to defend in a budget discussion.
Days 6 to 15: agree the thresholds before the design. For each candidate tile, ask the owner at what value they would change what they are doing. If they cannot answer, the tile is a trend and belongs in a monthly review rather than on an operational screen. Keep only the ones that survive that question, which is usually three or four rather than twenty. Write the threshold next to the owner's name, since a number agreed in the abstract tends to be disputed the first time it demands an awkward action.
Days 16 to 30: run it on paper and count the looks. Publish the surviving numbers once a day by whatever means is quickest and record who reads them and what changed as a result. If a figure prompts no action in three weeks, delete it. Building a dashboard around tiles that already proved useful is far cheaper than removing tiles after people have grown attached to them. Keep a note of the ones you deleted and why, so the same tile is not requested again by someone who was not in the room.
A practical worksheet: Operational dashboard
An operational dashboard earns its place by changing what someone does in the next hour. Take the three tiles you would put on the first screen and complete a row for each before any of them is built.
| Decision area | What to record |
|---|---|
| Definition | The exact calculation behind the tile, including what is excluded, written plainly enough for two teams to agree on it |
| Owner | The named role that watches this tile during the working day, not the department that commissioned it |
| Threshold | The value at which the number stops being informative and starts requiring action, and who set that value |
| Action | The specific thing the owner does when the threshold is crossed, and what they do when they cannot fix it themselves |
| Measurement | How often the tile is actually looked at after the first fortnight, which is the only honest test of whether it was needed |
A tile with no threshold and no action is a strategic metric that has wandered onto an operational screen. Move it to a monthly review, where a trend is useful, and give the space back to something a shift supervisor can act on.
Related guides
- Where this gets expensive is covered in Digital Transformation Roadmap: A 7-Step Business Guide.
- The groundwork for this is set out in Paperless Workflow: How to Digitize Paper-Based Processes.
If Operational Dashboard: KPIs, Examples, and Design is the start of a roadmap someone has to own, Discuss Your Digital Roadmap.
Ali Boran Gazel