An independent study reference written by Dr Phuc V. Nguyen. It is not official subject material — for assessment requirements always follow your subject outline and vUWS.
Data storytelling
Data storytelling is the practice of building a chart, or a sequence of charts, around a claim and the decision it supports, rather than around whatever data happened to be available. Three parts do the work. A structure that leads with the finding instead of the method. Annotation on the chart itself, so the reader is told what to notice. A stated implication, naming what should change. Segel and Heer describe a spectrum running from author-driven presentations, which are tightly ordered, to reader-driven exploration, which lets the audience roam. Most business communication sits somewhere between the two.
Why it matters
An analyst who has spent a month inside the data knows where to look. The audience has ninety seconds and no map. Storytelling is the work of transferring that orientation. It is not decoration, and it is not persuasion bought at the cost of accuracy. It is the difference between handing someone a chart and handing them a chart with an arrow, a sentence, and a reason to care.
A results presentation opens with data sources, then method, then cleaning steps, and reaches the finding only in its twenty-sixth exhibit. What is the main cost?
Worked examples
An analyst has a thirty-exhibit pack covering every cut of the customer data, and twelve minutes with the leadership team.
Reorder around the claim. Open with the one sentence the analysis supports, for instance that second-month drop-off accounts for most of the lost revenue. Follow with a single exhibit that shows it, annotated so the second month is marked and the size of the gap is written on the chart. Add one exhibit that rules out the obvious alternative explanation. Close with the decision being asked for and what it would cost. The remaining twenty-seven exhibits become an appendix, available if challenged and not walked through.
A public-facing report has to serve both a general audience and analysts who will want to check the detail.
Use the martini-glass structure described by Segel and Heer. Begin author-driven, with a short fixed sequence that establishes the question, the headline finding and the caveats. Then open into a reader-driven view where the audience can filter by region, period or segment. The narrow stem guarantees that everyone leaves with the same core reading, and the wide bowl lets sceptical readers test it. Opening straight into a free-form explorer, with no stem, usually means the general audience leaves with no finding at all.
Common mistakes
- ✗Storytelling means making the data more persuasive than it deserves. Narrative structure decides what is shown first and what is explained, not what the numbers say. A story that survives its caveats is stronger than one that hides them, and hiding them is a different failure with an ethical name.
- ✗A story needs a beginning, middle and end in a fixed order. Narrative visualisation covers a spectrum. A tightly ordered sequence suits a live presentation, and a guided-then-open structure suits a document readers explore at their own pace.
- ✗Annotation clutters a chart. Untargeted decoration clutters a chart. A short label pointing at the one feature that matters reduces reading effort, because it removes the need for the audience to search the graphic for the point.
- ✗The story comes at the end, once the analysis is finished. Deciding early who the audience is and what decision is on the table shapes which analysis is worth doing at all. Leaving it to the end usually produces a set of findings with no owner.
Revision bullets
- •Lead with the finding, not with the method
- •Annotate the chart so the reader is told what to notice
- •Name the decision the analysis is asking for
- •Author-driven to reader-driven is a spectrum, not a binary choice
- •Martini glass: fixed opening sequence, then open exploration
- •Structure is about ordering, never about overstating
Quick check
A results presentation opens with data sources, then method, then cleaning steps, and reaches the finding only in its twenty-sixth exhibit. What is the main cost?
Which change to a chart is storytelling rather than distortion?
Connected topics
More in Visual Communication
Sources
- Segel, E., & Heer, J. "Narrative Visualization: Telling Stories with Data." IEEE Transactions on Visualization and Computer Graphics, 16(6), 1139-1148, 2010.Source of the author-driven to reader-driven spectrum and the martini-glass structure.
- Kosara & Mackinlay (2013)Kosara, R., & Mackinlay, J. "Storytelling: The Next Step for Visualization." IEEE Computer, 46(5), 44-50, 2013.Argues that presentation is a distinct stage from analysis, with its own design requirements.
- Knaflic (2015)Knaflic, C. N. Storytelling with Data: A Data Visualization Guide for Business Professionals. Wiley, 2015.Practitioner treatment of annotation, decluttering and structuring an analytical message.