Data storytelling still matters—arguably more than before.

The craft is shifting from purely hand-built charts and meetings to a partnership where human judgment frames the narrative and AI accelerates exploration, drafting, and iteration. The skills that have always been at the core of strong stories—clarity, context, and domain knowledge—remain the foundation. AI changes the pace and breadth of what we can evaluate leading up to the story.

What Hasn’t—and Shouldn’t—Change

1. The audience decides.
   Reports don’t create alignment—people do. We still need to meet stakeholders where they are and anticipate their questions. A CFO hears “variance, risk, runway.” A sales leader hears “pipeline, opportunity, next actions.”
2. Context beats data points.
   Why something moved matters more than noticing that it did. Seasonality, pricing, incentives, and operational realities shape meaning. AI won’t know those the current market nuances unless you bring them to the surface.
3. Trust is the moat.
   Trust comes from clear definitions, transparent logic, and a steady cadence of accurate work (and honest corrections). No model can fix undefined terms, shaky data quality, or vague KPIs.
4. Structure carries the message.
   A simple arc—setup → insight → decision → action—still builds the momentum to help people commit. Most remember the path to agreement, not the chart that got them there.

What Has Changed

1. Speed and breadth of exploration.
   When we are focused on the internals of the business, AI can help us think outside of the box to suggest alternative segmentations, and flag “you might have missed this” anomalies. You can test ten story angles in the time it once took to test two.
2. Expectations for personalization.
   Stakeholders now expect role- and region-aware versions. AI makes it practical to turn one core finding into several audience-specific summaries without losing the thread.
3. A higher bar for visuals.
   If anyone can generate a chart quickly, the human role is choosing the right visual for the moment—and tying it to the decision at hand.
4. Greater scrutiny of assumptions.
   As AI contributes drafts, people will ask: “Where did this come from?” Strong storytellers will need to show lineage, definitions, and constraints—often in an appendix or definitions page.

Where AI Helps

• Idea generation: e.g., “Show top drivers of margin compression controlling for seasonality, product mix, and customer segment.”
• Scenarios: quick what-ifs (supplier +2%, segmented price changes, comp plan tweaks) expressed in plain business terms.
• Story variations: one core story adapted for executives, sales, ops, and success teams—each with a tailored “so what” and “now what.”
• Translation: converting findings into a slide, email, scorecard, or meeting notes.

Where AI Can Mislead— Simple Guardrails

Hallucinations and overreach.
When sources are thin, AI may fill gaps with confident but weak assumptions.

Guardrails:

• Limit inputs to sources you trust; require citations for key claims.
• Ask for counter-evidence and failure cases alongside the main argument.
• Keep a short “assumptions & exclusions” box in the deliverable.

The New Blend: Human Story + AI Assist

• Frame the decision.
  Be explicit about the outcome, the trade-offs, and what “good” looks like.
• Curate data and define terms.
  Agree on entities, windows, and measures. Publish definitions with the story.
• Use AI to explore, not to decide.
  Have it propose views and counter-narratives; keep notes on what you tried and why you set options aside.
• Draft, then human-edit for logic.
  Let AI write the first pass. You tighten reasoning, remove filler, and connect to strategy.
• Instrument the story.
  Label visuals in plain language. For each insight, attach a recommendation, an owner, a commitment, and a follow-up date.
• Close the loop.
  Track actions and outcomes in KPIs. Feed what you learn back into prompts and the model.

A Simple Checklist for Today’s Data Stories

• Decision: What are we asking the room to do?
• Definitions: Are KPIs and terms visible?
• Drivers: Did we test multiple causes?
• Counter-story: What would a skeptic say—and did we address it?
• Commitment: Owner, date, lead measure, and lag KPI.
• Lineage: Can someone else reproduce the result?

The Payoff

People still trust people and make decisions. People need clarity, context, and trust. AI doesn’t replace that; it amplifies it—speeding exploration, tailoring messages for different audiences, and surfacing blind spots earlier. Invest in the craft—audience understanding, crisp definitions, and honest logic—and AI becomes a multiplier, not the narrator. The most persuasive story is still the one that sparks collaboration and motivates action.

Thanks for reading!

Mark