Corvidae AI helps RS Group lift Paid Search revenue 6% on 5% less spend

A measurement problem first

According to the release, the project started as a data problem rather than a bidding one. RS Group PLC moved to a strict "reject all" first-party cookie consent model. As a result, platforms such as Adobe Analytics, Google Ads and Microsoft Ads had to model and estimate the customer journeys they could no longer observe.

Joshua Bamsey, Senior Manager of Performance Marketing Effectiveness at RS Group PLC, said the company had a long-term ambition to move away from last-click attribution but "lacked a robust tool to execute it". He said the consent change accelerated the transition and meant traditional platforms relied "heavily on modelling and guessing based on significantly less data."

What Corvidae says it did

Corvidae AI says its patented technology reconstructs customer journeys from raw first-party eventstream data using predictive AI, with no cookies and no persistent identifiers. The company states that the approach is compliant with GDPR and CCPA by architecture.

Over the same four-month test period, across all RS Group PLC properties worldwide, Corvidae reports that it tracked £193m more revenue than Adobe Analytics and covered over 95% of the active customer base.

The reconstructed data also changed the picture of how customers buy:

• Journeys were 46% longer in the UK and 68% longer across the EU than the platforms had reported.
• 58% of all sales touched a Paid Search ad.
• 72% of those sales also involved an Organic Search interaction, an overlap that platform reporting did not show.

Rather than deliver this as a report, Corvidae fed the reconstructed journeys directly into RS Group PLC's bidding algorithms and into Power BI via API.

Results and rollout

The controlled pilot ran in RS Group PLC's Italian operation. It produced the 6% increase in Paid Search revenue on 5% lower ad spend, along with the 12% improvement in return on ad spend. On the strength of those results, the process is being extended to all other regions.

Paul Morris, Senior Director, EMEA Performance Marketing and Web eCommerce at RS Group PLC, said that even at the company's size it needs tailored solutions so data does not sit in a "walled garden" but integrates with its own systems, namely Power BI and the Google stack. He said the board and Executive Committee see Data-Driven Attribution, and by extension Corvidae, as a core part of the company's analytics evolution. According to Morris, it makes the business less reliant on bottom-of-funnel channels and paves the way for "far more aggressive growth budgeting."

About Corvidae and what it means

Corvidae AI describes itself as an ad performance automation platform. It was founded in 2008, is based in Edinburgh, and lists RS Group PLC, QVC, Menkind, Macmillan and Groupe SEB as clients. The company has raised £5m, led by the Archangels business angel syndicate alongside the Scottish Investment Bank. Its technology is covered by patents held by parent company QueryClick Ltd, including US 12,461,984 and EP 3888325.

CEO and Founder Chris Liversidge argued the issue extends beyond RS Group PLC. In his words, "Every advertiser buying Google and Meta is bidding on incomplete data." The results are reported by the vendor and its customer in a press release, and come from a single-market pilot, so wider rollout will show whether they hold across regions. For advertisers working under strict consent rules, the case points to feeding reconstructed first-party journey data into bidding, rather than relying on platform modelling alone.