iMoein در حال آماده‌سازی صفحهPreparing your page

When Users View a Dashboard but Take No Action, How Should You Improve the Experience?

Day 14 of 1,000 Product Manager Interview Questions

Imagine an operations supervisor opening a dashboard every morning, noticing several red warnings, reviewing the charts and closing the page without doing anything.

The product team's immediate reaction might be to redesign the interface. Perhaps the charts are confusing, the buttons are insufficiently prominent, or too much information competes for attention.

But there is another possibility: the user may understand the information perfectly well yet have no idea what decision to make, lack the authority to act, or see no reason why action is necessary.

This is where the difference between a dashboard that displays information and a product that supports decision-making becomes apparent.

Day 14 of our series of 1,000 Product Manager interview questions addresses the following challenge:

If users see a dashboard but don't take action, how would you improve the experience?

Although this appears to be a user experience design question, a strong answer requires an understanding of user behaviour, data analysis, problem prioritisation and product judgement.

What Is the Interviewer Really Assessing?

The interviewer is not simply looking for suggestions to change button colours, introduce simpler charts or display additional notifications.

They want to understand whether you can diagnose the problem before proposing a solution.

A convincing answer should establish:

  • Why users visit the dashboard and what they are trying to accomplish.
  • Which user segments experience the problem.
  • Whether the displayed information is sufficient, trustworthy and current.
  • Whether users understand what action, if any, they should take.
  • Whether they have the authority, time and tools to carry it out.
  • How the proposed improvement would be evaluated against meaningful outcomes.

The underlying skill is distinguishing an observed behaviour from its actual cause.

The fact that someone has not clicked a button does not, on its own, demonstrate a usability problem.

First, Establish What Action Is Actually Expected

Before investigating users' behaviour, we need to understand the dashboard's purpose.

Not every dashboard is designed to trigger immediate action.

Some support monitoring, others enable analysis, and some help users identify exceptions requiring intervention.

Consider a finance manager reviewing monthly revenue figures. If performance is within an acceptable range, taking no further action may be entirely appropriate.

A customer support supervisor reviewing unresolved critical tickets, however, might be expected to assign or escalate certain cases.

Before treating user action rates as a success metric, we therefore need to answer:

Which user should take which action, under what circumstances, and what outcome should that action achieve?

This distinction prevents us from treating appropriate inaction as evidence of product failure.

Examine the User's Decision-Making Journey

One useful way to diagnose the problem is to break the experience into a sequence:

Notice → Understand → Recognise significance → Decide → Act → Observe the outcome

This is a practical analytical model for this question, not a formally prescribed product framework.

A breakdown can occur at any stage.

1. The information is not relevant

A dashboard may contain numerous indicators without presenting the information needed for the user's immediate task.

An operations supervisor, for instance, may care less about total sales and more about which orders are at risk of delay.

In this situation, the problem is not necessarily a lack of information. It may be that the product is presenting the wrong information to the wrong person.

2. The information is difficult to interpret

Even accurate data can be difficult to use without context.

Displaying an error rate of 4.2% without explaining the reporting period, expected range, severity or potential consequences leaves the user uncertain about what it means.

Users need to understand what has changed, why it matters and whether any intervention is warranted.

3. Users do not trust the data

A dashboard showing outdated, inconsistent or unexplained figures may not be trusted as a basis for decisions.

Users might consult colleagues or verify the information in another system before acting.

Redesigning the appearance of the chart will not resolve the underlying data-quality problem.

4. The next action is unclear

Sometimes users recognise the problem but cannot determine what to do next.

The dashboard may indicate an unfavourable situation without providing a route to the relevant records, the person responsible or the details needed to investigate.

What is missing is not necessarily another visualisation but a clearer connection between information and action.

5. Users cannot take the action

The primary obstacle may exist outside the interface entirely.

A supervisor might have read-only permissions, require another manager's approval or need to switch to a separate operational system.

Adding an action button without addressing permissions and workflow constraints would not solve this problem.

How Do We Discover the Real Cause?

My recommendation would be to conduct focused discovery before committing to a dashboard redesign.

First, segment users according to their responsibilities and jobs to be done. Senior managers, operational supervisors and frontline specialists may all view the same dashboard but expect quite different things from it.

Then combine three forms of evidence.

Behavioural analytics: Examine which sections users visit, whether they open detailed information, where they leave the journey and whether the intended task is subsequently completed elsewhere in the product.

User research: Observe people completing realistic tasks. Ask what they understand from an alert, what decision they would make and what prevents them from acting.

Operational evidence: Establish whether the underlying issue genuinely required intervention, who was responsible and how the case was eventually resolved.

These sources should be considered together.

Product analytics may tell us what happened, but behavioural events alone cannot reliably explain why it happened.

The GOV.UK guidance on measuring service performance similarly recommends combining quantitative measurements with user research and contextual analysis.

Which Solution Should We Choose?

Once the cause is established, there are several defensible directions. The right choice depends on the failure we have identified.

Option 1: Make Information More Relevant and Prioritised

If users struggle to recognise what matters, redesign the dashboard around the decisions they actually need to make.

Possible improvements include:

  • Showing critical exceptions ahead of purely informational metrics.
  • Highlighting meaningful changes against previous performance.
  • Making data freshness visible.
  • Separating items requiring attention from normal conditions.
  • Providing direct access to supporting details.

This approach can reduce cognitive effort and help users recognise priorities.

However, better information presentation will not be sufficient if the user lacks authority to act or faces an unnecessarily complicated workflow.

Option 2: Provide a Direct Path from Insight to Action

If users understand the issue but must navigate several screens to address it, the priority may be to shorten the journey.

For example, an alert could link directly to affected records, allow a case to be assigned to the responsible person or open the appropriate resolution process.

That does not mean every action belongs directly on the dashboard.

Sensitive operations, such as changing prices, cancelling orders or altering user permissions, may require additional context and confirmation.

Reducing clicks is valuable only when it improves efficiency without introducing unacceptable errors.

Option 3: Improve the Workflow Rather Than the Dashboard

Sometimes the real problem is that users should not have to repeatedly inspect a dashboard to discover urgent issues.

For time-sensitive exceptions, targeted notifications, task queues or automatic assignment to an appropriate owner may provide a better experience.

There is a trade-off, however.

Excessive alerts can become noise, encouraging people to ignore even genuinely important warnings. Severity, timing and responsibility therefore matter.

My starting preference would be the smallest change that addresses a verified, recurring barrier for a specific user group, rather than a comprehensive redesign.

If the root cause is unreliable data or unclear ownership, the first improvement may need to happen behind the interface.

Practical Example: An Operations Dashboard for a Hotel Booking Platform

The following scenario is entirely hypothetical and is intended to illustrate product decision-making.

Imagine a B2B hotel booking platform with an operations dashboard showing reservations that may have inventory discrepancies or confirmation problems.

Operations supervisors review these warnings daily, but a substantial proportion of cases requiring attention remain unresolved.

The design team suggests making the warnings larger and redder.

As the Product Manager, I would not approve that change immediately.

Instead, I would investigate how many alerts are genuinely valid, who is responsible for handling them and what steps users currently take to resolve the underlying issues.

Suppose observation reveals three recurring problems:

  1. Some alerts are no longer valid because the dashboard uses delayed data.
  2. Supervisors cannot easily distinguish urgent cases from less important ones.
  3. Accessing the relevant booking details requires several steps, and responsibility for follow-up is unclear.

Rather than redesigning every chart, I would propose a focused experiment.

At the top of the dashboard, we introduce a prioritised list of verified cases requiring attention.

Each item displays its severity, latest status, assigned owner and a direct path to the relevant details.

Where permissions allow, users can also mark a case as being investigated or assign it to an appropriate colleague.

Why choose this approach?

Because the evidence suggests that the problem is not simply whether warnings are visible.

Users need to distinguish important cases from background information and understand how to progress towards resolution.

The proposed design targets that particular barrier.

Nevertheless, it remains a hypothesis.

If further investigation shows that most alerts are inaccurate, improving data freshness and alert logic should take priority over creating a new interface.

How Should We Measure Success?

The primary measure should not be dashboard visits or button clicks.

Those events can support behavioural analysis, but they do not demonstrate that users are solving the intended problem.

For our hypothetical booking platform, I would define three layers of measurement.

Primary Outcome: Effective Resolution Rate

The proportion of valid, actionable cases that receive an appropriate and correct intervention within an agreed time window.

The denominator should be the number of unique cases that genuinely require action and have been made available to an authorised user, not the total number of dashboard views.

The response window should reflect operational expectations and severity.

A critical booking exception might require rapid intervention, while a lower-priority discrepancy may allow more time.

Diagnostic Metrics

Useful supporting measures could include:

  • Median time from first exposure to a valid alert until the first appropriate action.
  • Number of steps required to access the information needed.
  • Percentage of users who correctly identify the highest-priority case during usability testing.

These measures help explain which parts of the experience have improved.

Guardrail Metrics

We should also monitor incorrect actions, unnecessary escalations, invalid alerts and adverse consequences for bookings.

A design that encourages faster decisions but causes more mistakes is not successful.

Before running an experiment, establish a baseline, define the relevant events, agree on an observation window and determine the criteria for making a decision.

Where feasible, compare the new experience against the previous one using comparable user groups or cases.

If the sample is small or meaningful actions occur infrequently, interpret the results cautiously rather than claiming certainty from limited evidence.

How Do Product Responsibilities Differ?

The Product Manager should establish which outcomes matter to users and the business, decide which barriers deserve priority and determine which change is worth testing.

The Product Designer contributes through research, interaction design and usability evaluation. Engineers help assess data reliability, permissions, technical feasibility and implementation risks.

A Project Manager, where that role exists, generally focuses more on delivery coordination, schedules, dependencies and agreed constraints.

If the organisation uses Scrum, the Scrum Product Owner is accountable for maximising product value and the effective management of the Product Backlog, as described in the Scrum Guide. Scrum does not require a separate job title of Product Manager.

Importantly, a Product Manager should not unilaterally redesign an operational process owned by another team.

Changes to workflows and decision rights require collaboration with the people responsible for carrying them out.

Example of a Natural Interview Answer

A credible spoken answer might sound like this:

"First, I'd check whether users are actually expected to take action after viewing the dashboard. Some dashboards exist primarily for monitoring, so doing nothing isn't automatically a problem.

"If action is expected, I'd narrow the question to a particular user group and investigate what's stopping them. They might not understand the information, trust the data, know what to do next or have permission to do it.

"I'd look at behavioural analytics alongside a few observations of real user tasks and the operational records behind the dashboard. That would help me distinguish a usability issue from a data or process problem.

"Depending on what we find, I'd consider clearer prioritisation, more relevant information or a shorter route to action. I'd test a small change with a defined group before investing in a full redesign.

"And I wouldn't judge success simply by clicks. I'd look at whether genuinely actionable cases are being handled more accurately and promptly, without increasing mistakes. If those outcomes don't improve, I'd revisit the hypothesis."

This response demonstrates a clear progression from problem definition and evidence to product choice, evaluation and willingness to revise the decision.

Follow-up Interview Questions

1. What would you do if you had only two weeks to investigate and improve the experience?

Two weeks would not be enough to investigate every possible cause with high confidence or deliver a substantial redesign.

I would therefore narrow the scope.

During the first few days, I would review available behavioural data and observe a small number of users from one clearly defined segment completing their actual tasks.

Once the most plausible barrier was identified, the team could develop a lightweight prototype or low-cost change and evaluate it through usability testing or a limited release.

By the end of the two weeks, I would look for two kinds of evidence: whether users understood the proposed journey more clearly and whether early indications suggested an improvement in appropriate action.

Two weeks might not provide enough observations to measure the ultimate business outcome reliably.

I would report what had been learnt, identify the remaining uncertainty and recommend whether the experiment should be expanded.

2. What evidence would make you reject your proposed solution?

I would question the solution if users continued to struggle with the relevant decision, if effective resolution rates failed to improve or if incorrect actions became more frequent.

I would also change direction if research established that the main obstacle was data reliability or a lack of user authority.

In either case, further investment in a visual redesign would be difficult to justify.

Rejecting a hypothesis is not a Product Manager's failure. Continuing to invest in an approach despite contradictory evidence would be a poorer decision.

3. What if users understand the information but still do not act?

This is an additional practice question rather than one of the two follow-ups in the original exercise.

I would investigate whether intervention is necessary, who owns the outcome, what incentives or constraints exist and whether organisational processes permit the expected action.

If the barrier concerns responsibility or operating rules, the best solution may sit outside the dashboard.

Common Mistakes in Answering This Question

Starting with a feature list. Suggesting more charts, different colours or additional buttons before identifying the cause may generate development work without solving the problem.

Treating all users alike. Senior managers, supervisors and frontline employees may make entirely different decisions from the same information. Designing one undifferentiated experience can make matters worse.

Optimising superficial engagement. More visits, longer sessions or additional clicks do not necessarily indicate better decisions or outcomes.

Ignoring data quality. A polished interface cannot make decisions based on unreliable information trustworthy.

Assuming users should always act. Sometimes the correct outcome is to review the situation and deliberately make no change.

Presenting certainty without evidence. A competent Product Manager should be able to explain which observations or results would cause them to reconsider their recommendation.

Conclusion

When users view a dashboard but do not take action, the first response should not be to make buttons more prominent or simplify charts.

We must establish why people use the dashboard, what decisions they are expected to make and where the journey from observation to action breaks down.

My recommendation is to validate the underlying problem, test the smallest meaningful improvement with a specific user segment and assess success through the quality and outcomes of users' actions rather than dashboard engagement alone.

If the evidence points to unreliable data, unclear responsibilities or an ineffective operational process, the solution must change accordingly.

A successful dashboard is not necessarily one that generates more clicks. It is one that helps people make better decisions at the right time and, when appropriate, turn those decisions into effective action.

Sources

  1. GOV.UK Service Manual — How to set performance metrics for your service — Guidance on meaningful performance metrics, combining evidence, contextual analysis and using dashboards to support decisions.
  2. Microsoft Learn — Tips for designing a great Power BI dashboard — Guidance on audience-centred dashboard design, information prioritisation and reducing visual clutter.
  3. Reforge — 7 Product Manager Interview Questions From Top Product Leaders — General principles for assessing structured problem-solving and product judgement in interviews; not evidence that this exact question was asked in a real interview.

Share this article

Send this article to someone else.

Moein Ghezelbash Technical product management and digital platforms
Read next Designing a Product to Find a Reliable Exercise Partner: From User Needs to a Testable Solution