After 10+ years of building tools for wetlab scientists, I’ve learned that making complex genomic data accessible isn’t about dumbing it down—it’s about smart design.

Here are my 3 core principles:

  1. Start with the Question, Not the Data

Don’t ask “What can we visualize?” Ask “What decision are you trying to make?”

I once spent weeks building a beautiful multi-omics dashboard. Scientists used it twice… Why? It answered questions they weren’t asking.

Now I start every project with a variation on: “Walk me through your Monday morning.

What would make you say ‘Finally, I can see what’s happening’?”

  1. Make the Common Case Trivial

80% of your users need 20% of your features. Make those 20% effortless.

Example: Scientists don’t need to see every QC metric—they need to know “Is this sample good enough to trust?” One green checkmark beats 15 charts. Save the deep-dive complexity for the 20% who need it.

  1. Show Impact, Not Just Data

Raw numbers are intimidating. Context is powerful.

Instead of showing “Expression level: 2.3 FPKM,” show “3x higher than baseline” with a simple arrow. Instead of p-values, show “Strong evidence of difference.”

Scientists are brilliant—they don’t need protection from complexity. They need their cognitive load focused on insights, not interpretation.

The goal isn’t to make scientists into data analysts. It’s to make data analysis invisible so they can focus on being scientists.

What’s your experience building tools for non-technical users? What works (or doesn’t work) in your field?