A data story either walks readers through a fixed sequence or hands them the controls to explore

What it is

Data narrative structure falls into two basic modes: in author-driven mode, readers follow the author's set sequence, the message unfolds layer by layer, and readers do not control the path; in reader-driven mode, readers get an open data view and decide for themselves what to look at and in what order. Neither mode is superior—they fit different scenarios: author-driven suits occasions with a clear conclusion to convey (presentations, news data stories); reader-driven suits occasions where conclusions must be discovered or verified by readers (analysis reviews, transparency reports). The classic symptom of mixing them up is "a full filter panel embedded in a slide" or "preset narrative conclusions inside an exploration tool."

Why it happens

The modes' difference is rooted in the direction of information flow and the reader's role. In author-driven mode, information flows one way (author → reader), readers are recipients of conclusions, and the design goal is maximum delivery efficiency: a linear order eliminates the decision cost of "what to look at first," each step serves one point, and concentrated control makes pacing, emphasis, and setup all manageable—at the price that readers cannot verify, only choose to believe or not. In reader-driven mode, information flows both ways (data → reader's query → data's response), readers are producers of conclusions, and the design goal is maximum discoverability: openness lets readers ask questions the author never anticipated, at the cost of high cognitive effort—without guidance, most readers do not know where to start. The two modes correspond to two ways readers build trust: author-driven relies on the argument's persuasiveness, exploration on one's own verification. An author-driven narrative lacking credibility signals (data source, methodology) breaks trust early; an exploration tool with insufficient openness (key dimensions locked) converts failed verification into distrust just as fast.

Where it stops holding

Mode choice is not the author's preference but is determined by three situational variables: whether readers have the will and ability to explore (executives are time-poor and low-willingness—author-driven; analysts are willing and able—exploration works), whether the narrative goal is delivering a known conclusion or supporting discovery (quarterly report vs problem diagnosis), and how contested the conclusion is (contested conclusions require readers to verify themselves; exploration mode is a necessary condition for trust). Pure modes are idealizations—real artifacts are almost always hybrids, and the key to hybrid design is the switching point.

Applying it

• Declare the narrative mode (guided / exploratory / hybrid) when the artifact is scoped; team consensus on mode precedes concrete design.
• Author-driven artifact check: linear sequence, one point per step, no stray interactive controls.
• Exploratory artifact check: key dimensions freely filterable, no preset conclusion text, a starting guide (where to begin).
• Verification: ask target readers to describe how they used the artifact; "I just went through it in order" (guided) or "I looked at it from a few angles myself" (exploratory) matching the declared mode means the design landed.

Related

• Same group: U9.04.2 Hybrid mode needs an explicit switching point · U9.04.3 Scrollytelling must preserve graphic context
• Nearby: U9.01.1 Exploratory visualization serves the analyst themselves · U9.01.2 Explanatory visualization serves readers and needs explicit conclusions
• Search terms: author-driven · reader-driven · narrative visualization