Do we need a dashboard, or something else?
A dashboard addresses the last mile, not the first. It makes sense when data exists, is reliable and everyone agrees on what each indicator means. When these conditions are not met, a dashboard will not solve the problem; it will simply display it more quickly and with better graphics. I regularly encounter requests for dashboards, and in roughly two-thirds of cases I recommend starting somewhere else.
What a company really wants when it says ‘we need a dashboard’
This request usually reflects one of three needs:
- ‘I want to see what is happening in the company without having to ask.’ This is a need for oversight, and a dashboard can meet it if the data exists.
- ‘I want everyone to discuss the same figures.’ This is a need for definitions, and a dashboard will not meet it. Without agreement on what an indicator means, the dashboard will become a fourth version of the truth.
- ‘I want people to make decisions without me.’ This is a need for decision architecture. A dashboard is only its presentation layer.
A test before you start building
Before commissioning a dashboard, you should be able to answer these questions:
- Which indicators will it contain? (Five to ten, not fifty.)
- What exactly does each one measure? Write it in one sentence.
- Where does the data come from, and how old is it?
- Who is responsible for ensuring that the figure is correct?
- What decision will be made based on it?
The fifth question is the most important and the one most often left unanswered. An indicator that no one acts on is decoration.
Why dashboards often die in manufacturing companies
The typical cycle looks like this. A dashboard is built and everyone looks at it for the first month. Then someone finds a figure that does not match what they know from operations. It turns out that the dashboard is only partly correct because the data arrives late. Trust falls. Three months later, one person still looks at it. A year later, no one does.
The technology did not kill it. It was killed by the lack of trust caused by the underlying data.
What to do when the data is unreliable
This is the cheaper and more effective sequence:
- Select five indicators that genuinely inform decisions
- Define each in one sentence and appoint one person responsible for it
- Find out where they originate and how delayed they are
- Reduce that delay wherever it can be done cheaply
- Only then build the display
Steps 1 to 4 require no technology, and without them step 5 is money wasted.
Your next step
Whether your company needs a dashboard or needs to address something first is one of the questions I answer during a Digital Business Review.
Book a free 30-minute call →