Our goals included:

• Making the chart instantly readable: users should understand what they’re looking at within seconds.

• Clarity over density: The first screen should be scannable and teach the pattern quickly, pushing complexity into the drill-down.

• Stable context: Switching views or filters shouldn’t feel like starting again.

• Progressive disclosure: Deeper detail (definitions, forecasting, examples) should appear when users ask for it.

• Scalable patterns: Design for additional views and future evolution without one-off solutions.

Making the landscape chart the front door

As a product team we chose a scatterplot for the landscape overview because the core job of the tool is to show two signals at once; how established an attribute is (Prevalence) and whether it’s accelerating or declining (Growth Momentum). Plotting Prevalence on the x axis and Growth Momentum on the y axis lets users interpret the market in a single glance; niche vs dominant, rising vs falling, before deciding where to drill in.

In dense markets, many attributes can share similar values, which causes points to overlap and reduces the chart’s usefulness. To preserve readability at scale, I defined a visual treatment that used reduced point opacity to reveal overlapping clusters, paired with subtle positional jitter to separate coincident points without distorting the underlying data story. Where lowering opacity risked weakening visibility, I reinforced the points with a subtle stroke so they remained legible while still reducing visual noise.

Finally, we intentionally kept the scatterplot format consistent while improving the underlying methodology and the supporting experience around it. This avoided retraining users on a new chart type, preserved the familiar “market snapshot” mental model, and allowed us to focus change where it added most value (interpretation, drill down, forecasting).

Unifying separate tools into one coherent experience

I defined a single product entry point with internal navigation between lenses, instead of making users jump between separate tools.

This reduced the “where do I go?” anxiety and encouraged cross-attribute exploration, which is where the insight value lives.

Balancing guidance for first-time users with continuity for returning ones

A key product decision was to balance guidance for first-time users with continuity for returning ones.

On first use, the experience adapts to the user by pre-selecting filters based on their industry, giving them a relevant starting point without requiring upfront setup. From there, the system remembers their most recent view and filter context, so future visits feel continuous rather than reset.

The intent was to reduce early friction while building familiarity over time. For first-time users, a guided starting point removes decision load. For returning users, restoring their previous state helps them pick up where they left off making the product feel intuitive and reliable from the first few seconds.

A guided starting point for new users, with state persistence for returning users.

Filters as the anchor for context and exploration

The broader goal was a unified experience that lets users switch between lenses seamlessly, rather than treating each view like a separate product.

I anchored the experience around a stable, prominent filter area so the user's context (category/market) was preserved while switching lenses.

This helped users think “what market and category am I exploring?” first. Keeping that stable makes exploration feel fast and reversible.

Overview → drill-down → forecast: progressive disclosure of complexity

Landscapes and forecasts are designed to work together. The chart helps users identify what matters, forecasts help validate whether a trend looks sustainable or short-lived.

This is why the information hierarchy matters. The overview should remain scannable, while forecasting and supporting detail appear in drill-down, only after users have signalled intent by selecting an attribute.

Progressive disclosure: overview first, then forecasting depth on selection