Excited?

Let’s Work Together!

Enter your email address below and a member of our team will reach out right away.

Latest news

Start a Business Dashboard With the Decision It Supports

Plan a business dashboard around a real user decision. Clarify the question, data meaning, level of detail and next action before choosing charts or a platform.

Updated

Illustrative physical chart panels beside a branching blue decision path and a square action block on a planning table

Before commissioning a business dashboard, name the reader, the question they need to resolve and what they can do with the answer. Then decide whether they need a summary, a way to investigate, or a working queue that supports action.

That decision comes before the chart style or platform choice. A dashboard can be useful for exploration as well as operations, but the software brief should explain which kind of use it is meant to support.

Key takeaways

  • Define a real reader and question before collecting chart requests.
  • Separate monitoring, investigation and action instead of treating them as one screen.
  • Give each measure a clear meaning and a visible reporting context.
  • Match detail and refresh expectations to the task rather than promising real-time data by default.
  • Accept the reporting tool against a user decision, not only the appearance of its charts.

A dashboard is a proposed surface, not the requirement

"We need a dashboard" describes a possible interface. It does not explain the work the interface must help someone do.

A founder reviewing a monthly summary, a manager investigating a change and a coordinator handling an individual request may all look at related information. They do not necessarily need the same level of detail or the same controls.

Ask each reader what they would do after seeing an unexpected value. Sometimes the answer is to investigate. Sometimes it is to discuss the result at a meeting. Sometimes it is to open a record and complete a task.

Those are all legitimate uses, but they imply different software. For custom web applications, a useful buying conversation starts with those distinctions rather than a gallery of attractive charts.

Follow the observe, explain, act path

Use this original Essential Designs decision path to challenge the initial request. The branches can lead to separate views within one product; they are not a rule that every dashboard must fit a single category.

First ask: does the reader need an overview of a known subject? If so, define what they are monitoring, which reporting period matters and what comparison gives the result meaning. A concise summary may be enough.

Next ask: does the reader need to find out why something changed? If so, identify the breakdowns, filters and supporting detail needed to investigate. A headline number alone will not answer that question.

Finally ask: does the reader need to complete work on an item? If so, define the item, its state, the available action and what happens afterwards. The right surface may be a work queue with supporting measures rather than a chart-led dashboard.

Do not force all three branches into the first screen. Decide which task should be easy to recognise on arrival and which detail belongs one step further in.

Test the distinction with one fictional workflow

Imagine a business reviewing customer onboarding work. This is a planning illustration, not an Essential Designs project or a customer performance claim.

A leadership question might be: "How is onboarding progressing over the reporting period?" A manager's question could be: "Which part of the process needs investigation?" A coordinator may ask: "Which item needs my next response?"

The same source system might contribute to all three answers, but the screens need not be identical. A period summary, a breakdown view and an actionable list each have a different purpose.

The useful brief identifies the question and the next step before deciding how to visualise the information. It does not assume that adding a chart to the coordinator's page makes the work easier.

Nor should an exploratory view be dismissed because it does not immediately trigger a task. If investigation is the purpose, describe what a useful investigation looks like and what context the reader needs.

Make the measure understandable before making it prominent

A measure needs a definition that the buyer and delivery team interpret in the same way. Name what is counted, which period is covered and which records are included.

For the fictional workflow, "onboarding complete" could mean several different things unless the organisation agrees on the relevant state. A visually correct chart cannot resolve that disagreement.

Also distinguish the period described from the time the data was last updated. A fresh retrieval of last month's information is not the same thing as a current operational view.

Put these definitions in the software requirements. Keep the explanation proportionate and close enough to the view that a reader does not have to guess what the number represents.

Choose detail that helps the next question

Avoid treating every available field as a filter request. Ask what question a proposed control helps the reader resolve.

A manager investigating a process might need a meaningful breakdown. A coordinator working a queue may need the record's current state and the next permitted business action. A monthly review may need a stable summary with explanatory notes.

The goal is not to minimise information at any cost. It is to give the user enough context for the intended task without making unrelated detail compete for attention.

Where several audiences use the product, identify the shared definitions and the different views. Consistency of meaning does not require identical presentation.

Do not make real-time updates the default promise

Ask how current the information needs to be for its intended use. A frequent refresh may be important for an operational task and unnecessary for a periodic review.

Then ask what happens when a source has not supplied a new update. The reader needs an honest state, not a confident-looking number presented without its age or limitations.

This is a scope question for the delivery team, not a claim that one refresh interval is best for every organisation. Data sources, integrations and operating needs can differ.

The brief should state the expectation the project is actually prepared to support. "Live" should not be used as a decorative adjective when its meaning has not been agreed.

Plan for the reporting tool to remain useful

The UK Government Analysis Function's dashboard guidance asks teams to establish user need, consider alternatives and allow for support throughout the dashboard's life. It also distinguishes interactive dashboard uses from situations where another format may be more suitable.

That public-statistics context is not a private-sector requirement, but it offers a useful challenge: why is this particular reporting surface justified?

For a custom business tool, assign responsibility for definitions, changing questions and decisions about obsolete views. The need may change even when the software still renders successfully.

Discuss software support in terms of keeping the tool usable for its purpose, not simply preserving every chart indefinitely.

Ask for a demonstration of the decision

Before accepting the reporting feature, give a representative user a realistic question and observe how they use the proposed view. Can they identify the relevant period, understand the measure and reach the intended next step?

This is not a promise that a short demonstration proves adoption or commercial value. It is a focused way to check whether the delivered surface supports the question in the brief.

If the user must export everything and reconstruct the answer elsewhere, investigate what is missing. The answer may be a clearer definition, a better detail view or a different reporting format.

Discuss the decision your reporting tool should support with Essential Designs. Bring the reader's question before a list of preferred charts.

Sources and scope

The bounded primary reference is the Government Analysis Function's building and managing dashboards guidance. The observe, explain, act path and fictional onboarding questions are original Essential Designs editorial recommendations, not a client case or a measured usability study.

No platform ranking, productivity gain or financial result is asserted. Published and reviewed October 30, 2026. Review the source, service links and decision examples by January 30, 2027.

Excited?

Let’s Work
Together!

Enter your email address below and a member of our team will reach out right away.