Optimising the digital shopping experience for a home improvement retailer

We audited the same journeys four ways, then pulled the findings into one ranked map

Within each category, we modelled two journey shapes on both desktop and mobile: people browsing for inspiration, and people who know what they want and need it done by Saturday.

I scored every screen against Nielsen heuristics with a severity scale (Minor, Moderate, Severe, Unusable), logging 153 ranked issues by end of week three. Logging issues without severity rankings produces a list with no signal, where 153 problems all look equally urgent on the page and the client has no way to start. The four-point scale meant the 16 Severe and 2 Unusable issues stood out from the 66 Minor and 69 Moderate ones, and the team could open the prioritisation conversation without going back through the list themselves.

In parallel, the team layered competitor benchmarks (direct rivals like Screwfix and B&Q, plus category specialists like Lick, Dulux, Magnet, Wren), six user testing interviews, and conversion-funnel data from Google Analytics.

The analytics work, owned by a colleague on the team, agreed with where the heuristic findings clustered: add-to-basket conversion was running at 5.9% against an industry average around 11%, and the most severe heuristic issues sat on the same screens the funnel was leaking through. Two lenses landing in the same place was a much stronger signal than either lens alone.

Each strategic question came with Figma variants the team could pull apart

Findings needed to translate into something the team could act on, so I built Figma design variants against each of the four "how might we" questions we landed on. A diagnostic engagement ends with a deck and a list of recommendations, which usually disappear into a backlog. The deck did its job, but the variants were the piece I cared most about, something the team could pull apart, mix, or reject on sight. A written recommendation says "do this". A variant says "consider this", which is what a client team actually needs to make their own call.

Hick's Law on the basket

The existing flow asked shoppers to commit to delivery or click-and-collect on every item before checkout, slowing the whole journey, so the variant moved that decision into checkout where it belonged.

Mega menu

Interior Paint was buried mid-list despite being top-three for pageviews, so the variant put it first.

Paint PDP

The variant bridged inspirational landing pages to product pages with a layout that surfaced both.

Appointment booking

A streamlined form replaced the existing booking path's friction.

A sequenced starting point with strategic questions to anchor next steps

The team got a prioritised problem map across three categories, four strategic questions to anchor the next phase, and Figma design variants to choose between. I don't have visibility into what got picked up after handover, but the engagement was scoped to give them somewhere to start, and it did.