Developing a Point Based System to Attribute Conversions
|CATEGORY||ROI CONSULTANCY AWARD|
|CAMPAIGN TITLE||Developing a Point-Based System to Attribute Conversions|
|CLIENT||All (with the focus being Honda, Fitbit, Aldi and Tennis Australia)|
|BRAND||Aviva Motor Insurance|
|TEAM MEMBERS||Hamish MacDonald
Car insurance is one of the most competitive industries in the UK, and reaching potential customers online has become more difficult due to fragmentation of media channels, and highly contested digital audience buying.
Aviva provides around 31 million customers worldwide with insurance, savings and investment products. They are the UK’s largest insurer and one of Europe’s leading providers of life and general insurance. In 2016 they wanted to grow market share by making their digital display marketing budget work more efficiently with high ad viewability levels harnessed as a real-life performance indicator.
Not all ads are viewable, for two main reasons: first, some ads are served below the fold and the viewer doesn’t scroll the webpage down far enough to see them; or second, the ad didn’t load quickly enough before the user moved elsewhere. This can result in major marketing investment wastage, and if ads aren’t viewable, how can they increase Aviva’s market share?
The internal Zenith Programmatic team were challenged to move away from the standard CPA model that had been used for years, and factor in viewability as a real-life performance indicator, while retaining efficiencies at scale. This ultimately meant that total campaign conversions would now be discounted by the campaign viewability factor.
There was a reason no-one else was doing this: CPA and viewability have an inverse relationship. Publishers reserve their most viewable impressions for buyers that are willing to pay a premium. To achieve an effective viewable CPA and drive positive ROI, it is not always possible to pay the premium that publishers are demanding.
With viewability becoming a more reliable measurement tool for programmatic campaigns, it was important for Aviva to incorporate this into their marketing measurement. If they did not, this could result in higher budget wastage, and more importantly, basing future marketing investment on targeting that does not reflect true media performance.
The Zenith Programmatic team looked at the lowest common denominator – price – and determined that if they were able to buy the right audiences and the right impressions, rather than simply the most viewable impressions, for the right price, an upward trend in both CPA efficiencies and viewability would be possible.
By analysing Aviva’s first party audiences based on conversion path, search intent and recency, combined with multi-dimensional analysis of historical data, the Programmatic team were able to segment audiences based on the propensity of a user to convert, as well as delivering in the right context. It was essential for the Programmatic team to know the propensity for a user to convert to determine the optimal bid strategy given the likelihood that an impressions was deemed to be viewable. By balancing the likelihood that an impressions was deemed viewable and the propensity for a user to convert, the Programmatic team were able to use these signals to inform bidding strategy, ensuring an effective viewable CPA.
Comparing performance in the January-May periods in 2015 and 2016, the Programmatic team were able to improve the efficiency of cost-per-viewable sale by 31%, increase viewable sales volume by 83%, while also increasing viewability by more than 32%. This amounted to a saving of more than £111,000 year on year. The strategies that were implemented in this campaign now form an integral part in the always-on remarketing strategies, not just in the car insurance category, but across all Aviva’s programmatic campaigns, and is starting to be rolled out across all programmatic remarketing campaigns.
“We’re very pleased with the progress we’ve made in the programmatic space, and the optimisations that have driven a stronger result. We’re now rolling this optimisation methodology out on the remainder of our DR campaigns alongside DMP audiences, and have every confidence that we can continue to improve performance.” (Helen Graham, Senior Marketing Delivery Manager, Aviva)