Skip to content
Technology transformation expertise firm · ARDIANE Group Innovate, secure, succeed

Glossary

Data storytelling: definition, components and limitations

Domain

Data

In brief

Data storytelling refers to presenting the result of an analysis in the form of a narrative built for a specific audience, which combines data, visualisations and commentary. A term of common usage, data storytelling aims to make a quantified finding understood, along with the decision it calls for.

Page contents

Definition of data storytelling

Data storytelling refers to presenting the result of an analysis in the form of a narrative built for a specific audience. This entry distinguishes three elements in it: data, visualisations and commentary that orders them. The term is one of common usage, and its boundaries vary from one author to another.

The word “story” does not refer to fiction. It refers to an order of presentation: a context, a finding, an explanation, then what follows from it.

What narrative adds to a figure

A metric displayed in a report says how much. It does not say whether the value is normal, why it has changed, or what ought to be decided. The body reading it must reconstruct this reasoning itself, and two readers may draw two conclusions from it.

Data storytelling does this work upstream and sets it out. It answers four questions: what is observed, compared with what, for what reason, and with what consequence. Its result is judged by the decision the audience can take at the end of the presentation.

Data, visualisation, commentary: three components

On the data side, the narrative relies on a small number of figures, whose definition, scope, period and source are displayed. A metric calculated in two ways in two presentations weakens the narrative. Its single definition falls to the semantic layer.

Each visualisation carries a message. The type of chart is chosen according to the comparison to be shown: a trend, a breakdown, a gap against a target. An annotation points to the point that matters, and the scale does not distort the gap.

The commentary orders the views and states the message of each one. A title such as “Delays are concentrated on two sites” says more than “Delays by site”. A language model can draft a first version of it; the figure and the causal link remain for the author to check.

The audience governs the whole. The same analysis is told differently to an executive committee, which arbitrates, and to an operational team, which is looking for where to act.

Data storytelling, data visualisation, dashboard, reporting: not to be confused

Data visualisation is the graphical representation of data. It is a component of the narrative.

The dashboard is a monitoring instrument, consulted repeatedly, which displays the same metrics. The narrative is one-off and carries a conclusion.

Reporting distributes figures at fixed intervals, in a stable format. It informs without arguing.

Exploratory analysis looks for what the data contains. Data storytelling comes afterwards, to set out what has been found. In self-service BI, a business analyst can carry out both.

Nor is a narrative a proof. It selects, and this selection can mislead: figures chosen to serve a thesis, truncated axis, correlation presented as a cause. Showing what contradicts the conclusion and separating the finding from the hypothesis limit this risk.

Example: a response time that is getting longer

Take a customer service centre whose response time to complaints is getting longer. The dashboard flags the gap against the service commitment, without explaining it. An analyst prepares a four-view presentation for the steering committee.

  1. The finding: the response time curve, compared with the commitment, with the date from which the gap appears.
  2. The breakdown: response time by complaint reason, which isolates billing disputes.
  3. The explanation: since a change of procedure, these cases go through a second approval.
  4. The proposal: two options costed in workload, and what the analysis does not yet explain.

The committee has a finding, a probable cause and a choice to make.

Defining metrics, the semantic layer and designing dashboards around the decisions to be taken are presented in our page business intelligence and analytics.

FAQ

Frequently asked questions

Another question? Ask us

What is data storytelling?

Data storytelling consists of presenting the result of an analysis as a narrative intended for a specific audience. It combines data, visualisations and commentary that links them, from the finding to the proposed decision. The term is one of common usage, and its boundaries vary from one author to another.

What is the difference between data storytelling and data visualisation?

Data visualisation is the graphical representation of data: line chart, histogram, map. Data storytelling uses it as a component and adds to it an order of presentation, a message and commentary suited to an audience. A chart presented without a message is data visualisation alone.

Does a dashboard do data storytelling?

A dashboard is a monitoring instrument: it displays the same metrics at each consultation and leaves the interpretation to its user. Data storytelling produces a one-off presentation, built to carry a finding and a proposal before a decision-making body. The two complement each other: a narrative can start from a variation spotted on a dashboard.

What are the limitations of data storytelling?

A data narrative selects what it shows, which exposes it to three drifts: keeping only the figures that favour a thesis, distorting a comparison through the choice of scale, and presenting a correlation as a cause. Displaying the scope, the period and the source of the figures, and including the elements that contradict the conclusion, limits these drifts of data storytelling.

A project to bring to completion?

A first conversation is enough to understand your challenge, identify the right people and tell you frankly what we can bring.