Woman applying mascara to her eyelashes looking side eye to a mirror in her hand
Case Study

Charlotte Tilbury Beauty reduces CPA by 29% with Custom Bidding

Charlotte Tilbury Beauty, the cosmetic line from the celebrity makeup artist and British female entrepreneur, has quickly become one of the UK’s biggest luxury beauty brands. Charlotte believes makeup is about enhancing natural beauty and empowering women and men to feel their best selves everyday.

Central to the growth of the brand is its investment in digital advertising which the marketing team uses as their primary way to boost the online purchases of their high-end goods. Preserving the privacy of its customers is also a top priority for them. With this in mind, they selected Display & Video 360 as their preferred media buying partner to further develop their business in the UK and expand to the US.

Optimizing for conversions

The beauty brand’s marketing team started by turning to Display & Video 360’s automated bidding solutions to optimize the performance of their campaigns. This tool automatically bids on impressions with a high likelihood of leading to a conversion, allowing the brand to significantly increase the number of people who purchased their products.

But soon after, Charlotte Tilbury Beauty ran into a catch-22 that’s well known by retailers selling high-value goods – they could not grow their media spend above a certain plateau without also growing their cost per customer acquisition beyond profitability thresholds. That’s because brands selling high-price point products drive fewer conversions at a higher price. This gives bidding algorithms less data to learn from.

Turning to Custom Bidding

To reach more customers profitably, the Charlotte Tilbury Beauty team connected with Scibids, a technology company that creates customizable bidding algorithms. They built an algorithm that was made just for the brand to power-up Display & Video 360’s standard bidding solutions.

“Display & Video 360 provides the perfect framework to unlock the power of Scibids’ customizable artificial intelligence. Our algorithms take the brand’s specific goals into account to design a personalized bidding strategy that can help them grow media efficiency and scale.”

Remi Lemonnier, Co-founder & President, Scibids

To account for the fact that the cosmetic brand had fewer conversions to optimize off of, Scibids designed a tailored full-funnel algorithm that factored in relevant event data higher up in the user journey. For example, they started valuing visits, clicks, and ad viewability as new input signals.

The beauty brand didn't have to wait for conversion data to gradually trickle in. From the start of the campaign, the custom bidding strategy provided Display & Video 360's machine learning technology with more events to learn from. This made up for a compressed learning period which saved the brand precious time and limited their media waste. From there, the algorithm kept dynamically informing Display & Video 360’s bidding tool with accurate conversion weighting, refreshed every few hours.

Experimenting to assess the impact on ad spend and reach

Charlotte Tilbury Beauty’s team needed an unbiased testing framework to verify the custom-built algorithm was working as intended. They used the experiment & lift feature in Display & Video 360 to test how the introduction of the custom bidding strategy impacted the campaign’s performance.

Within three weeks, the portion of the campaign using the tailored algorithm started exceeding the performance of the control campaign. And in no more than two months, the brand observed a 29% reduction in acquisition cost and a 60% increase in conversion rate.

“Scibids’ unique application of Display & Video 360’s custom bidding solution across our campaigns generated a step change in performance. We are excited to leverage this partnership to further scale our programmatic effort.”

Andrew Leung, Global Head of Performance Marketing, Charlotte Tilbury Beauty

Finally, giving more weight to upper and mid-funnel events in the bidding process allowed the cosmetic brand to double viewable reach. Thus avoiding repeatedly reaching the same viewers with the same ad which is a common remessaging pitfall.

Looking ahead

With these learnings fresh on their minds, Charlotte Tilbury Beauty is looking ahead to expand even further the efficiency of custom bidding. The brand is doubling down on their first-party data strategy and planning on using those insights to enrich Scibids’ learning database. As for Scibids, they’re anticipating helping an increasing number of marketers willing to take their bidding strategy to the next level.

You might also like: