Insight
Coca-Cola faced four marketing challenges:
1. As a CPG company, it was difficult to identify where consumers were on the brand affinity spectrum.
2. Non-programmatic media investments were siloed and as a result identifying who were consumers and where they navigate online lacked an actionable insight.
3. Campaigns depended directly on cookies.
4. There was a global guideline for markets to follow more data-led approach. It wanted to move towards more personalised and relevant communication.
Coca-Cola had more than 13 brands with activity in Mexico, each one with different targets, segmentations and investment capabilities, MediaCom needed to create a solution with the usage of the DMP-DSP that would allow Coca-Cola to be “media agnostic”, measuring efficiencies with full transparency, creating knowledge and audience insights for each brand.
Strategy
It knew that programmatic technology could solve these challenges. That is how the “PRESICION WITH SCALE” project was born, where it developed proprietary DMP-DSP technology. Two initial phases were identified, each one with its own results, performance and benefits.
1. Shift to Programmatic: The agency and the technology partner worked together to take existing media plans and buy programmatically by looking at each brand’s goals, how those goals could be achieved programmatically and what would be the resulting KPI in the DSP. Considering campaign objectives and taking into consideration that the knowledge surrounding users in contact with Coca-Cola assets was limited, the strategic decision was to drive users to the Main Digital Experience on a web based landing in order to complete the digital cycle, gathering the data input necessary for the DMP analysis and creation of audience segments. Benchmarks were redefined, moving away from shallow click and impressions KPIs to more tangible goals like full views, visit leads or a completed funnel, making ongoing real-time optimization assuring brand safety and full transparency in costs and delivery.
2. Creating Audience Segments: By shifting towards Programmatic, campaign data collection was possible in a unified platform, and an increase of users directed to Coca-Cola´s first party data sources was achieved, from here it was possible to segment users using a segmentation methodology that was created to specifically suit the needs of the PWS project, this segmentation allowed to understand the wide spectrum of brand affinity where the users were, measuring efficiencies by affinity level.
Execution
In close collaboration with the DSP partner, the agency started to give structure and strategy in programmatic buying. The integration was based in transparency of costs, optimisations, sites, exchanges, performances and targeting.
There were two primary buying tactics:
1. Prospecting: With the increased volume of impacted users that were guided to the brands owned media, the agency gathered the necessary amount of data that allows the discovery of audience insights.
2. Target: Each of the brands targets were effectively impacted with the use of segmentation tactics such as: age, gender, interest and lifestyle segmentation, geo targeting, device targeting, whitelist, day part, category targeting, etc. For example Sprite target was teens between 12 and 19 years old, while Santa Clara’s was food lovers between 22 and 44 years old.
The delivered integrated formats were Display, Video, Mobile and Rich Media. Also, the agency started to develop a private marketplace for video and display - when a site worked very well for a given campaign a private deal was created with that publisher in order to better scale and capitalise on its performance. Coca-Cola brands are storytellers, this is why the programmatic buying unit evolved from traditional campaign KPIs to full views and visits.
A segmentation methodology was created for the DPM, through this methodology audience clusters were created taking into account Brand Affinity. During the second stage these clusters were integrated into buying tactics, this allows the agency to do retargeting according to affinity level, making heavy investments with the low affinity clusters and keeping the involvement of the high affinity clusters. The collected data was used to create the audience profiles in the DMP.
Results
During the first stages of the campaign each brand experienced significant results against their main KPIs, compared to non-programmatic campaigns running during the same time period, Coca-Cola achieved a decrease in Cost per Completed Video View (CPCV) and Cost per Click (CPC) in performance ranging from 60% to 80% and eCPMs decreased by 30 - 70%. Coca-Cola was paying on a CPCV or CPC basis, the shift to programmatic represented a buying efficiency of more than seven million Mexican pesos vs. non-programmatic buying.
A total of 84 segments were created in the DMP showcasing high, medium and low engagement towards all Coca-Cola brands; developed from both media and first party data. Through the proprietary DMP-DSP technology developed, the agency found not previously known insights, nine brand user profiles were detected by making a detailed analysis of the data.
To identify user’s affinity, the agency looked at the frequency and number of page visits on a weekly basis. Current data shows 40-60% of users in the low affinity bucket, 20-35% in the mid affinity bucket and 10-15% in the high affinity bucket.
Ultimately, this has helped the agency to plan for new KPIs and targeting strategies in the future.