Law of Prägnanz

People will perceive and interpret ambiguous or complex images as the simplest form possible, because it is the interpretation that requires the least cognitive effort of us.

The scenario: a monthly KPI dashboard for a small coffee-roasting business. The numbers are identical in both versions — revenue, new customers, churn, and the same product-sales breakdown. Only the presentation changes: one version drowns the data in decoration, the other is already the simple shape your brain would reduce it to anyway.

What's wrong here? A gradient-on-gradient banner, three metrics in three competing typefaces, badges and icons fighting for attention, and a “3D” pie whose nine tilted slices are impossible to compare (is Gift Box bigger than Cold Brew? You genuinely can't tell from the chart). Viewers squint, mentally strip away the decoration, and walk off with a vague — often wrong — simplified impression.
app.beanledger.io/dashboard

✨ June Performance Command Center ✨

🔥 LIVE 📈 TRENDING ⭐ PRO
💰 Revenue
€48,250
▲ +12% vs May
>_ New customers
312
▲ +8% vs May
⚠️ Churn
2.4%
▼ −0.3 pts vs May
🍩 SALES MIX BY PRODUCT 🍩
  • Espresso Blend — 32%
  • Filter Roast — 24%
  • Cold Brew Kit — 18%
  • Gift Box — 12%
  • Decaf Blend — 5%
  • Ceramic Mugs — 3%
  • Hand Grinder — 2.5%
  • Subscription Add-on — 2%
  • Stickers — 1.5%

🔍 What changed

  • The data is identical — same three KPIs, same deltas, same product percentages; only the presentation was simplified.
  • Three typefaces, gradients, ornate borders and badges were replaced by three plain, uniform stat tiles.
  • The tilted “3D” pie of nine slices became four horizontal bars — lengths are comparable at a glance, and the 14% long tail is summarised honestly in one line.
  • Whitespace replaced decoration, so the numbers themselves are the loudest thing on the screen.

💼 Explaining it to stakeholders

“The brain simplifies whatever it sees — that's not optional. If we ship the decorated version, users do the simplification themselves and lose accuracy in the process: a 3D pie literally distorts the proportions it exists to show. Designing the simple version ourselves means the takeaway people leave with is the takeaway the data supports — and it's cheaper to build and maintain, too.”