Sponsorship Gap

A gap in the market for sport

Planning, allocating, delivering and reporting LED sponsorship campaigns was slow, confusing and hostage to stakeholder bottlenecks. Cortex already had an LED playout product running on the boards themselves — but the layer above it, where deals were sold and assets allocated, was still being done in spreadsheets.

I led the design and ran a series of workshops to define where the product needed to be in two years. The first thing I needed to know was where teams actually were today.

Tearing down the current workflow

Before designing anything, I mapped the existing process end-to-end — every spreadsheet, every email, every handoff between partnerships, sales, operations and the LED playout team. The bottleneck pattern surfaced fast: deals were being sold faster than they could be operationalised, and every step that wasn't in the playout product was reliant on someone remembering to copy a cell into another file.

“The deals weren't the problem. The space between ‘deal signed’ and ‘asset on a board’ was the problem.”

Ethnographic research with SportFive

LED campaigns are sold in a unique way, so we spent two days conducting ethnographic research — shadowing members of the partnership team at the UK's leading LED sponsorship agency, SportFive. I watched live deals being scoped, planned and negotiated; I sat in on internal stand-ups; I logged every tool opened and every spreadsheet touched.

By the end of the two days, I had mapped a complete end-to-end journey and collected a stack of working artefacts to analyse back at the office.

Document analysis

The artefacts I brought back from SportFive became our second discovery dataset — pitch decks, allocation spreadsheets, internal comms, client reporting templates, ops handover sheets. I ran a tagging pass on every document: the fields that recurred, what information moved between teams, where the same value was re-entered manually three times.

What surfaced: across every document, the same dozen data points kept appearing under different names. That observation defined a big chunk of the product's data model before we drew a single screen.

Nuances across the globe

Beyond SportFive, I interviewed agencies across other markets. How teams planned, sold and delivered LED campaigns with stadiums varied — sometimes driven by the sport, sometimes by team preference, sometimes by local regulation. The final solution had to flex for all of them without forking the codebase or the mental model.

Atomic research to lock the themes

Usability testing the early concepts with real partnership teams was non-negotiable. I used atomic research to break results down — Experiments, Facts, Insights, Opportunities — and surfaced the themes that would shape the rest of the product. Four categories emerged, mirroring the campaign process itself: campaign proposals, campaign creation, asset allocation, and reporting.

Reviewing in the open

I ran an all-hands workshop to put the early concepts against the user goals and requirements. Sales, operations, design and engineering in the same room — no opinion went unargued, and the workshop's output was a prioritised list everyone had signed off on.

Defining the architecture

Based on the four themes, I structured the product to follow the campaign process from proposal through delivery. The IA mirrored how users already thought — proposal → assets → allocation → reporting — rather than imposing a new mental model on top of an already-busy workflow.

Internal Sponsorship SystemUser IA

The allocation UI

I focused on simplifying the asset-to-game allocation — bulk apply across multiple events, and a status system so the Ops team could see at a glance what was ready to hit the LED boards. Status was the bridge between sales saying “done” and ops saying “live”.

Competing with the speed of spreadsheets

In the old process, users could copy/paste sponsor allocations across spreadsheets in seconds. The product had clear advantages elsewhere, but if it was slower than Excel for the most common operations, it would never replace it. So I prioritised features that matched that speed.

Autofill logicInline editingCopy across matches

I worked with the backend team on intelligent configurations that auto-generated sponsorship-allocation playlists for each event. One autofill click sent a ready-to-use playlist straight to the LED operator — the last spreadsheet step absorbed by the product.

What I'd do differently

The ethnographic + document analysis combo gave us a sharper data model than any number of stakeholder interviews would have. If I ran this again I'd build that into the kickoff template for every product team — two days in the field plus a structured artefact analysis, before any concepting starts. It compresses months of stakeholder back-and-forth into a week of grounded research.

I'd also engage LED operators earlier. We focused heavily on partnership and sales teams during discovery; ops feedback came later than it should have, and a few patterns in the allocation UI had to be revisited once we saw the workflow on the operator side.