Office of Small Failures / Case File 007Claim under review
Dashboard reasoning experiment

Dashboard Lie Detector

Does the chart actually support the sentence people are saying about it? Put the claim on trial before it reaches the next deck.

This tool does not detect fake data or accuse anyone of lying. It checks whether the reasoning around a dashboard metric is stronger than the evidence you have entered.

Enter the dashboard claim

How to check whether a dashboard supports a claim

Dashboard Lie Detector is a free data interpretation tool for checking whether a KPI, chart or reporting dashboard has enough context behind the conclusion being made. It does not verify the source data. It tests the reasoning around it.

Start with the comparison, not the story

A metric moving up or down only becomes interpretable when the comparison is clear. Previous-period comparisons can be distorted by seasonality. Targets and benchmarks answer a different question from matched-period or controlled comparisons.

Check definitions, denominators and segments

A headline KPI can change because tracking changed, the underlying sample changed, or one segment moved while another did not. Before turning movement into a narrative, check whether the measurement itself remained comparable.

Correlation is not the same as causation

A dashboard may show that a metric changed after a campaign, redesign or SEO release. That sequence alone does not prove the intervention caused the movement. Strong causal language needs stronger evidence.

Use the rewrite as a stakeholder-safe version

The tool creates a more careful sentence based on the information entered. It is designed to help teams separate what the data shows from what they suspect may explain it.

Dashboard data interpretation FAQs

What is Dashboard Lie Detector?

It is a checklist for testing whether a dashboard claim has enough context behind it.

Does it detect fake data or lies?

No. The name is playful. It checks reasoning, not honesty or data authenticity.

What does the tool check?

Comparison, time periods, tracking definitions, denominator or sample stability, segmentation, external context and causal language.

Can a dashboard prove what caused a change?

Usually not by itself. Causal claims generally need stronger evidence than a before-and-after dashboard.

Does my free-text data leave the browser?

The assessment runs in the browser, and the claim/metric text is not sent as an analytics parameter.