Skip to content

Radically Better Results

  • Featured in Econsultancy—Best practice case study

Challenge

eCommerce brands are increasingly facing challenges of a sector moving away from first-party cookies and a consumer base fragmenting their shopping experience across multiple sessions, mediums and devices. It soon became evident to market-leading Lovehoney that last-click attribution models didn't offer anything like the level of sophistication needed to truly understand customers' often complex journeys.

Add this fact to an inexorable shift towards the 'black boxing' of proprietary algorithms and added macroeconomic complexities such as a global pandemic, the sexual happiness brand felt that not only were they not seeing the whole truth with last-click attribution, but that they weren't even entirely trusting of the partial truths they could see.

As a result, Lovehoney set out to find an attribution provider that offered first party integration, multi touch attribution and total transparency in how the insights were derived.

They discovered Cubed. A SaaS solution large enough to serve as the super brain behind their marketing efforts, yet small enough to still feel very much like an extension of Lovehoney’s own team.

The brand’s data is now held in their own ecosystem, enriched by Cubed propensity modelling capabilities and secure in its entirely explicable methodology. Lovehoney have securely reposited, fully validated and entirely reliable first-party data - the foundation for all true insight.

With this accurate and reliable data, Lovehoney wanted to optimise paid search campaigns to Cubed data as they believed it would benefit multiple channels, not just paid search, and drive overall business growth.

Action

Using Cubed full model attribution data, Lovehoney worked with their PPC partner to run a robust geo-split test. Historic data was used to determine two comparable (99% correlation) location groups. One group was kept the same - used as the control - with the other used to bid to Cubed revenue data, instead of Google tracked revenue data.

The team then measured the impact of bidding/not bidding with Cubed revenue data across all GA channels.

Spend was comparable across test and control segments.

The insights derived include:

  • Lovehoney spent around 200% more across all seasonal query terms, in some instances over 700% more on some subcategories, in comparison spend on competitor bidding was reduced by almost 50%.
  • Lovehoney spent approx. 40% less on upper funnel query terms which were previously believed to be of higher value.
  • One area where Cubed confirmed Lovehoney thinking, is placing more value in non-branded campaigns, valuing generic campaigns 5% higher than GA and consolidating generic-focused strategies.

Change

  • Additional £734k marginal profit per year for the UK territory alone
  • The campaigns optimising to Cubed data drove an incremental £59.5k of revenue across all (GA) channels
  • Lovehoney will be implementing this strategy across all territories, so we expect this figure to increase considerably

We'd love to hear about your marketing goals