With rising media costs and flat Performance Max (PMax) returns, B&Q sought a step-change in targeting, personalisation and automation. The goal was to restore profitable growth, scale marketplace visibility backed by vendor media, and improve efficiency.

The proposal replaced manual rules and fragmented workflows with an adaptive, data-led operating model that responded to real-time demand at a product level.

Upp.ai delivered unified control of bids, budgets and product prioritisation, combining data engineering and enrichment with workflow automation. The platform aligned product visibility to live market context, learning which items to promote and when, and orchestrating within PMax to maximise relevance. 

The transformation was phased-in gradually to protect continuity, with early wins used to build trust and embed new ways of working across trading and merchandising. Outcomes included stronger return on ad investment, materially higher product visibility in core and seasonal lines, and substantial revenue growth in first-party and marketplace portfolios.

Judges noted the campaign demonstrated a strong, structured case with clear challenges, execution and outcomes. It was praised for restoring meaningful control and insight within PMax and improving performance while easing operational burdens. The strategy was viewed as an interesting and timely application of AI to maximise brand spend and accelerate success for new ranges.

One judge said: “An increasing issue with the emergence and growth of PMax has been the loss of control and insights into paid search activity and exactly how Google is delivering the performance that PMax campaigns are suggesting.

“This looks like a great way of taking back some of this control and additional insights in a way that actually helps the bottom line.”

Read more about the work with B&Q here.