Today I went back and reread the "Absorbing Data" chapter from Smarter, Faster, Better, and how it applies to designing data infrastructure for digital products. The book points out that while we're surrounded by more data than ever, it's easy to fall into "information blindness" — where too much data makes it harder, not easier, to make decisions.

The key insight is that we learn best when we interact with information, breaking it into smaller pieces and actively engaging with it. Here are some creative ways to apply these lessons to a low-touch digital product:

  • Introduce Disfluency: Make users interact with data by building custom reports or asking questions, not just passively viewing charts.

I've always felt this was the real issue with simple analytics / business dashboards. They skip the important step of handling the data, and just show you the output which is why I've always preferred to export some of the data and handle it in a spreadsheet.

The act of handling the data gives you a better understanding of the data and the questions you should ask.

  • Chunk Down Data: Break complex dashboards into simple widgets or focused comparisons.
  • Encourage Interactivity: Let users drill down into graphs and tables to reveal more detail.
  • Data Storytelling: Use visual narratives to guide users through key insights, like user journeys or conversion funnels.
  • Automated Insights: Highlight anomalies or trends automatically so users don't have to sift through everything.
  • Educational Layer: Add tooltips, help articles, or walkthroughs to teach users how to interpret metrics.
  • Hypothesis Testing: Let users form and test their own ideas about what drives behavior or conversions.
  • Custom Alerts: Allow users to set alerts for important metric changes.
  • Feedback Loops: Let users input their assumptions and see how they match up with real outcomes—gamifying learning.

Takeaways:

  • Data is only useful if users can absorb and act on it.
  • Interactivity, chunking, and storytelling help overcome information blindness.
  • Design data systems that encourage users to ask questions, test ideas, and learn by doing.