Share of visits and revenue
This group accounts for three times its share of visits in revenue. Search matters, but encouraging additional search use does not necessarily reproduce that effect.
01 — Context
Commerce Platform
Over four months, I designed navigation, a marketplace and a partner dashboard for an established e-commerce company. My work covered usage-data analysis, comparison of design variants and detailed user flows. The CEO and team leads were the internal stakeholders.
My role: analysis, product concepts, information architecture and interface design. NDA / confidentiality: client names and original designs are withheld. The placeholders contain no project imagery.
30-SECOND PROJECT READ
The project at a glance.
- Challenge
- Design search, marketplace and partner tasks as a connected product.
- My role
- Analysis, product concepts, information architecture and interface design.
- Team
- Company project with the CEO and team leads as stakeholders.
- Timeline
- Four months · 2026
- What shipped
- Floating header released; marketplace and partner dashboard developed as designs and prototypes.
- Outcome
- +5.6 percentage points in onward navigation from the entry page, according to project feedback.
Design system
I developed recurring search, account and navigation patterns, including mobile states. Headers and footers needed to remain recognisable across the three areas, with actions suited to the current task.
Working with stakeholders
The CEO and team leads were my points of contact. I contributed analysis, variant comparisons and detailed flows. The public account covers roles and design decisions.
02 — Working with stakeholders
Make the decisions concrete.
My points of contact were the CEO and team leads. I worked on which tasks needed priority in each area and how the interface should support them. Usage analyses, header variants and detailed flows made those choices concrete. Search provides one example: the analysis showed a high share of revenue associated with a relatively small share of search visits. That gave the search field a clear claim on header space. Comparing variants showed how much room remained for categories, accounts and service.
03 — Analysis · Search & navigation
Search deserves more space.
The first analysis connects search use, purchasing behaviour and the journey beyond results. It shifts the priority from displaying as many navigation links as possible to making search visible and understandable.
Orders per 100 visits
Search visits show roughly three times the order rate of category-only visits. Differences in purchase intent may contribute to this association.
Mobile: roughly half the purchase rate
Purchase rate among search visits, desktop = 100. Meanwhile, 38% of mobile visits end on the results page, compared with 28% on desktop. Searches without a click differ less: 41% versus 39%.
What this means for the header
3 %
of clicks go to header categories in the homepage heatmap. Search is the more heavily used entry point. Categories remain accessible while search receives clearer priority.
The analysis also identifies bot traffic in exit data. A sudden increase is therefore not automatically interpreted as a deterioration in UX.
Source: supplied A4 search and navigation analysis. Baseline observations, not a measurement of redesign impact. Mobile comparison: June/July 2026. Absolute visit and revenue totals are omitted.
04 — Navigation
Give search space; keep other tasks within reach.
I compared header variants for category visibility, laptop widths and use across other areas. The early recommendation used a continuous area switcher; the released version was V4 with a floating header. The design work covered open account and language menus, as well as mobile states. Footer content followed the task: discover products, create a listing or continue in the personal area.
05 — Marketplace
Take a listing through to publication.
The marketplace needed a flow for providers to enter details, add images and review a listing. I worked out the steps and the feedback for missing information. A quality summary points back to the relevant step. For changes to existing listings, the design distinguishes between updates that can be applied directly and those requiring another review.
06 — Partner workspace
Bring outstanding tasks into one workspace.
The dashboard brings together open enquiries, listings, orders and other partner tasks. My work covered the information hierarchy and the paths from an overview to handling an item. Responsibility and status belong with the task: who needs to reply, what awaits approval and what can happen next? The team inbox keeps the related communication together.
07 — Analysis · Personal customer area
Created does not mean used.
The second analysis examines use alongside economics. Empty profiles and entry after a purchase are particularly relevant to product design: the personal area needs a clear initial benefit and an obvious next action.
81 %
of created customer profiles are empty
A reason to examine what happens after registration.
3 %
of purchases start in the customer area
A small direct entry route into purchasing.
64 %
first open the area after purchasing
Most first visits happen after a purchase.
Customer comparison: each reference = 100
Revenue per existing customer over the following 180 days. Existing customers without the personal area form the reference of 100 in each group. Customers with the personal area are at approximately 1.2–2.1×. Self-selection and differing customer relationships are possible explanations; the comparison does not isolate a product effect.
More visits do not establish more value.
The overall shop trend provides context. It cannot be attributed to the personal customer area as either success or failure. Its contribution requires separate measurement of activation, return visits and subsequent purchasing.
The cost model depends on assumptions.
0.4–4.0×
Modelled benefit relative to estimated development costs across the combined scenarios in the analysis. These extremes describe sensitivity, not a return forecast. Full costs and incremental benefit are not established by this model.
Product direction from the analysis
The personal area makes the next step specific: add a product, find relevant parts or open the seasonal checklist. The footer responds to the current context rather than repeating the same purchase prompt everywhere.
Whether these offers actually activate people and bring them back is a separate evaluation question. The observed usage shares establish priorities for that evaluation.
Source: supplied A4 analysis of the personal customer area. Financial amounts, group thresholds, salaries and team sizes remain hidden. Indices and ratios were calculated from the supplied values and rounded. Observations, estimates and the header launch outcome are presented separately.
08 — Impact
More visitors continue beyond the entry page.
+5.6 percentage pointsOnward-navigation rate — after introducing the floating header
Outcome reported by the project. Baseline, measurement period and sample are not documented in this public version. The change alone establishes neither higher purchase conversion nor additional revenue.
09 — Project status
What shipped, and what still needs evaluation.
The floating header shipped. The footer, marketplace and partner dashboard reached design or prototype stages. The reported change in onward navigation belongs to the header launch. The next evaluation should track search-result clicks, completed customer profiles and handled enquiries. These measures answer different questions and need separate analysis.
