Improve Visualization for Metrics with Large Scale Differences in Evolution Graphs

[tsteur]

When plotting multiple metrics with significantly different value ranges in Matomo evolution graphs, smaller-scale metrics become visually unreadable and impossible to see trends.

For example:

• Visits: typically in the range of 50,000+
• Actions per Visit: typically between 1–10

When these are plotted together, the “Actions per Visit” line appears flat (close to 0), even though there are meaningful variations (e.g. from 3.0 to 3.2). This makes the metric effectively unusable in combined visualizations.

This is the case in many or all evolution graphs. Particularly, also when wanting to view for example the amount of AI Chatbot requests and the amount of acquired visits from chatbots at the same time. It's impossible to detect any trends around the acquired visits.

Current Behavior

• Metrics sharing the same unit (e.g. counts) are plotted on the same Y-axis.
• This causes smaller-scale metrics to be visually compressed when paired with large-scale metrics.
• As a result, trends and changes in smaller metrics are not visible.

Existing Partial Solution

Matomo already handles this well when metrics have different units:

• Example: Visits (count) + Bounce Rate (%)
• In this case, a secondary Y-axis is used, making both metrics readable.

Expected Outcome

• Smaller metrics (e.g. Actions per Visit) become readable and meaningful when plotted alongside large metrics
• Users can better analyze correlations and trends across metrics
• Improved usability of evolution graphs without requiring users to split views. It's extremely tedious to regularly switch manually between the same metrics every day.

Example from demo.matomo.cloud