Product recommendation
Overview
Recommending the right products is at the core of Jolimoi’s experience. Yet for many Stylists, especially new ones, this is where friction starts.
This project was developed in close collaboration with Product, Tech, and Data teams.
As Product Designer, I owned the end-to-end experience: structuring the flow, designing the interfaces, and making sure the feature felt simple, actionable, and fully integrated into existing usage.
The recommendation logic was defined by the Product and E-shop teams, and the content for each routine step was provided by the E-shop manager. My role was to translate this into a clear and intuitive experience that Stylists could actually use in real situations.
The problem
Through NPS feedback and discussions with Stylists, several recurring pain points emerged. Stylists reported difficulty navigating the catalog, a lack of confidence in their product knowledge, and uncertainty around how to build a complete routine for their clients.
As one Stylist put it:
"I don't know what to recommend."
Even when they found a relevant product, the next step wasn’t obvious. They had to figure out what comes before, what comes after, and how to turn individual products into something coherent for their client.
The issue wasn’t access to products. It was the gap between knowing products and actually recommending them.
Reframing the Problem
Based on these insights, we shifted the approach.
Instead of helping Stylists know more products, we focused on helping them know what to recommend next.
A product should not be the end of the journey — it should be the starting point of a routine. This became the foundation of the experience.
Designing the Experience
The feature is embedded directly into the product page, as a natural extension of the browsing experience.
When viewing a product, Stylists can immediately see how it fits within a routine and what is missing to complete it. The product becomes the starting point of the recommendation, creating a smooth transition from browsing to recommending.
Routine preview on product page
Routine preview on product page
From there, they can access the full routine view, where the experience becomes more guided. The routine is organized into clear steps, each with one main product and up to three alternatives when relevant. This makes it easy to build a complete routine without having to search or compare extensively, while still keeping flexibility.
The goal was not to automate decisions, but to guide them just enough to move forward confidently.
Full routine view
Full routine view
A key challenge was to keep the experience efficient. Stylists need to act quickly, not configure complex selections.
The interface allows them to select products within each step, see the total update in real time, and add the full routine to the basket in one action.
Everything is designed to reduce the gap between decision and action.
Integrating into Real Usage
Another important aspect was making sure the feature works in different contexts.
We designed a flexible system:
  • when client data is available, recommendations are more tailored
  • When it’s not, the experience still provides a structured routine based on product and category logic.
This ensures the feature is always usable — and never blocks the Stylist.
Personalized routine
Personalized routine
Impact
This feature directly addressed one of the main friction points in the experience: moving from browsing to recommending.
Stylists now have a clearer path to build routines, which reduces hesitation and speeds up the recommendation process. It also helps them better understand how products fit together, especially at the beginning of their journey.
More broadly, it strengthens Jolimoi’s positioning around assisted selling by making product expertise more accessible directly within the app.
Next steps
This project laid the foundation for a more guided and scalable recommendation system.
Next steps would focus on making the system more adaptive and connected to real client behavior — for example, by personalizing routines based on past purchases, or adapting the experience depending on the Stylist’s level.
This would allow Stylists not only to recommend more easily, but also to build more consistent and relevant follow-up with their clients.
Conclusion
This project helped turn product recommendation into something more concrete and easier to use in practice.
By structuring the experience around routines, Stylists don’t have to figure everything out on their own anymore. They can move faster, with more clarity, and feel more confident in what they recommend.
The challenge was less about adding a feature, and more about making it fit naturally into how Stylists already work — without slowing them down.