04
Virtual Business Coach
Turning performance data into clear, actionable guidance.
- AI / 0→1
- Ambiguity
- Ownership
01The problem
Brand Partners could see their performance and targets, but the data did not clearly tell them what needed attention or what to do next.
The challenge was turning information into actionable guidance.
02How I knew
The existing Business App exposed performance data but offered limited support for interpreting it.
Mapping business targets and recurring activities showed an opportunity to connect performance signals directly with relevant actions.
03My role
This was my first project at Oriflame. I led the product design from problem framing and concept development through prototyping, validation and MVP implementation support.
04Key constraints
Complex business rules had to remain understandable at a glance, while Recruit, Activate and Develop goals, progress and motivational elements needed to work as one coherent experience.
05Design process
- Business rules
- user needs
- information structure
- concepts
- prototype
- validation
- MVP
The focus shifted from presenting more data to helping users understand their next move.
06Key decisions
I organised the experience around Recruit, Activate and Develop, connecting performance insights with recommended actions.
Gamification supported progress, but actionable guidance remained the primary focus.
07Trade-offs
More performance detail offered completeness but made the next action harder to identify.
I prioritised actionable information over data density, keeping supporting detail secondary.
08Solution
Virtual Business Coach combined targets, progress and recommended actions, supported by achievements, rewards and rankings.
- Recruit / Activate / Develop
- Insights & Actions
- Gamification
- Rankings / Top 100
09Result
- ≈25–30%
- increase in weekly engagement
Users also showed more consistent follow-through on key activities such as recruitment and activation.
10What I’d do differently
I’d validate the recommended actions earlier, not only whether users understood the interface.
The stronger question would be: “Does this recommendation actually help you decide what to do next?”