Performance dashboard

Helping traders self-diagnose and reduce support load

Redesigned CMC’s performance dashboard to shift from vanity metrics (win rate, P/L) to behavioural insight and coaching.

CMC Markets is a London-listed global online trading platform providing retail and institutional clients access to leveraged products including CFDs, spread betting, and FX across multiple asset classes.

Problem

Customer Problem

Tracking performance was a tedious, manual process. Traders struggled to distil complex trade data into meaningful insights and often didn’t analyse their behaviour consistently.

The dashboard showed results - but not patterns.

Business Problem

Customer retention averaged 11–12 months. Analysis showed many traders were liquidating accounts due to poor risk management and lack of behavioural awareness.

Low insight → Poor decisions → Account losses → Churn.

Product hypothesis

Offering customers a way to view their performance and understand what they are doing wrong, will improve their trading strategy and prolong their engagement with the platform - leading to increased profits for CMC Markets.

Solution

A dashboard that shows customers their stats across multiple time frames and instruments.

Approach

My role

Strategy
Discovery
User research
UI design
Usability testing

What I did

Analysed 100+ support tickets to identify recurring pain points.
Conducted 30+ discovery interviews (including open-ended questions and ranking exercise) across novice and experienced segments.
The main objective was to understand what metrics they track and what they focus on in understanding their performance.
Built 3 prototype iterations that used customers' actual past trading data - making feedbck much more genuine and intuitive.

Information architecture research

Brainstorming metrics

Stack ranking

User interview

User interview

Design iteration

Key Product Decisions

Shift from reporting to coaching
Rather than surfacing static metrics, I introduced behavioural nudges (e.g., “You tend to lose after 3 consecutive trades”) to encourage reflection and adjustment.
Progressive density across devices
Mobile prioritised quick diagnosis and pattern recognition.
Desktop enabled deep analysis and time-frame comparison.
Trader-native language over finance terminology
Rewrote performance explanations using language traders already use, reducing interpretation friction..