Fitia Andriana

Mendi redesign across three views: customer login, mender job-type selection, and mender order list

Every user group acting on its own

Company

Mendi

Industry

Fashiontech

Tasks

Product Strategy

UX Design
User Research
User Testing
Prototyping

Product Strategy

UX Design
User Research
User Testing
Prototyping

Context A two-sided marketplace connecting customers with a network of menders across Norway for clothing repair and alteration. I owned the design direction for three user groups within one ecosystem: customers, menders, and admin. As the company's capacity narrowed, I owned the product priorities and scope as well.

Challenge Peer-to-peer means the flow only works when both customer and mender act. A mender needs enough information about a job to decide whether to accept it. A customer needs to feel safe enough to complete an order with a stranger. On the old platform, neither group had what they needed to act on their own, so admin stepped in manually to keep each order moving. The platform worked, but only because admin manually filled the gaps.

Production

Hvor mange knapper skal sys på?

–

199 kr

Opptil 3 knapper

Fortsett

Hvor mange knapper skal sys på?

–

199 kr

Opptil 3 knapper

Fortsett

Hvor mange knapper skal sys på?

–

199 kr

Opptil 3 knapper

Fortsett

Process and insight I ran qualitative interviews with active menders and continuously observed the mender community, documenting the findings for the development team. Two patterns stood out.

On the mender side, customer service worked as a middleman because the flow depended on it. Communication opened only once a mender had already accepted a job, so they had to commit before getting clarification. The information customers submitted was scattered rather than collected in one view and admin had to study each case and relay it onward.

On the customer side, I found a trust barrier. Customers chose pick-up points because that was the free option. But it carried a hidden trust cost: they felt uneasy leaving high value garments without confirmation.

I backed the qualitative work with systematic QA: documented issues across several test rounds, which made the work iterative and traceable.

Process and insight I ran qualitative interviews with active menders and continuously observed the mender community, documenting the findings for the development team. Two patterns stood out.

On the mender side, customer service worked as a middleman because the flow depended on it. Communication opened only once a mender had already accepted a job, who had to commit before getting clarification. The information customers submitted was scattered rather than collected in one view and admin had to study each case and relay it onward.

On the customer side, I found a trust barrier. Customers chose pick-up points because that was the free option which carried a hidden trust cost as they felt uneasy leaving high value garments without confirmation.

I backed the qualitative work with systematic QA: documented issues across several test rounds, which made the work iterative and traceable.

Rebranded

Rebranded

Rebranded

Design decisions Each decision follows directly from the findings. I reworked the flow so the mender sees all customer information in one view and can assess an assignment before accepting it, which lets the mender act independently and frees admin from the flow. I let customer and mender communicate and arrange delivery and pickup directly, independent of pick up points, which meets the trust barrier head on instead of leaving it behind a free option. For the admin group, I designed a new dashboard that brought order detail lookup down from 4 clicks to 1, turning a manual workaround into a fast, purpose built tool.

Internal tool The admin ran on Jetadmin, where opening a single order took four clicks and the information that mattered, status, communication and payment, sat across separate screens. I rebuilt it in Retool, wired directly to the database with no extra integration, on a surface I can adjust myself whenever something stops working. Now everything the team needs sits on one screen, and opening an order dropped from 4 clicks to 1. A manual workaround became a fast, purpose-built tool.

Prioritization and scope Stepping in as interim manager meant I owned what shipped and when, not only how it looked. I triaged documented issues by severity and by impact on the core two sided flow, fixing what blocked customers and menders from completing a job first, and leaving polish for later. I made deliberate cuts: work that fell outside the critical path was scoped out of this release so it would not delay it. Holding that tension between an ambitious vision and real constraints was as much a part of the role as the interface itself.

To push the critical path forward, I facilitated two internal hackathons where the team took large parts of the ordering flow from concept to near-complete. Time-boxing the work kept the team aligned on what mattered most and turned prioritisation into shipped progress.

Result I delivered a conversion-oriented platform, built to ship in stages, with a new visual identity and design system. Early feedback from menders described the new communication flow as clearer and better structured, and the redesign reframed customer service as an edge-case tool rather than a step the flow depended on. The admin dashboard turned a slow manual process into a fast one. Together, the changes were designed to let each user group act on its own, the goal the old platform never reached.