Objectives
LifeSearch is an insurance brokerage specialising in life insurance, income protection, and critical illness cover, amongst other products. Since 1998, they have safeguarded over three-quarters of a million UK families and businesses. They take pride in their commitment to selecting the most suitable cover for you and your family.
In addition to a Kaizen display ads campaign aimed at enhancing brand awareness for LifeSearch, we were simultaneously running a retargeting campaign with the primary focus of generating leads.
Lead generation is infamously challenging for display campaigns, but we relish an opportunity to test and learn. With the nuances of lead gen in mind, we created an experiment to A/B test a new bespoke bidding algorithm with the goal of increasing leads and improving conversion rate while maintaining a similar CPA to the control.
- Increase brand reach into relevant audiences
- Increase leads
- Increase Conversion Rate
Our Approach
Because the aim of the re-targeting test was purely lead generation, we created a bespoke algorithm focused on value-based bidding. Value-based models leverage a type of AI called machine learning (ML) to allow the DSP (Demand Side Platform) to factor granular first-party conversion data into bids. In this case, we factored in data about site visits, user engagement, form starts, and form completes.
We assigned a progressive weighting model to various site activities, with user interactions assigned proportionately higher weights the closer they took the user to submitting a lead form. We deliberately excluded certain other forms and interactions, such as phone number clicks, to ensure the DSP was not unintentionally learning to bid more for impressions likely to lead to non-KPI interactions.
Additionally, to account for the click-based attribution models used in analytics platforms like GA4, post-click visits were weighted higher than post-view visits. Because of this, we expected the model to increase ad clicks as well as conversions. However, ad clicks were not the primary goal of the retargeting campaign and were not explicitly coded into the algorithm for this strategy.
Results
As predicted, the A/B test results indicated that the bespoke bidding algorithm delivered a notable uptick in clicks over standard bidding strategies. With our bidding algorithm, clicks were 3.12% higher compared to standard bidding strategies. This increase in clicks signifies a stronger resonance with the target audience, translating into increased traffic and engagement.
In fact, the most compelling evidence of our algorithm’s superiority lies in its impact on overall site engagement. When considering all user interactions on the site, including visits, form submissions, CTA clicks, etc., our bespoke algorithm lifted engagement by an impressive 11.37%. This really demonstrates how efficient machine learning and tailor-made algorithms can be at identifying the members of your audience who are most likely to engage with your brand.
The ultimate success came in the form of an uplift in leads, with form submissions rising by a similar margin to overall engagement. The A/B test delivered a 12% increase in form submissions when compared with standard bidding models, translating to an 8% lift in conversion rate. This emphasised the effectiveness of our bespoke algorithms at driving desired actions from the audience, resulting in a higher conversion rate and improved campaign efficacy.
Overall, the A/B test showed that advertisers can achieve better results with bespoke bidding algorithms without incurring significantly higher costs, harnessing the power of machine learning to maximise campaign ROI.
- 1.5 million retargeting impressions delivered
- Clicks lifted 3.12% compared to standard bidding strategies
- 11.37% boost to overall site engagement
- Stable CPA
- Leads up 12%
- 8% improvement to conversion rate
Danielle Moore, Senior Digital Marketing Manager at LifeSearch said:
“Venturing into display advertising for lead generation was unchartered territory for us at LifeSearch. Yet, with Equimedia’s expertise, we felt confident in their capable hands. The bespoke bidding algorithm played a crucial role in delivering our lead generation targets, helping us to continue to protect the UK properly.”
Increase in CTR of 3.12%
8% increase in Conversion Rate
12% increase in leads generated
Interested in seeing what we can do for you? Get in touch and let us know about your requirements today.
