Meta platforms and other social media channels offer advertisers a breadth of options to tailor campaign execution, across multiple dimensions and levels within a campaign (such as objective, targeting, creative etc.). These controls do not operate in isolation, which means that campaign effectiveness is determined by the combination of many different variables, all working together to drive success.
With social media adspend forecast to reach US$177bn in 2022, overtaking television at US$174bn, we conducted this research to empower participating clients and Publicis agencies to maximise their return on brand advertising investment within Meta platforms, by identifying those variables that were most influential in driving brand lift. We asked our agency partners and some of Publicis’s largest global advertisers to identify the variables and outcomes they needed to understand, and then used their answers to define five key hypotheses and align them with different campaign attributes. Our purpose was to guide strategic best practices by testing these five hypotheses and finding the most effective combination of tactics to lift brand impact. For example, is it better to target a broad audience or a more defined audience? Which objectives achieve the most cost-efficient brand lift? What percentage reach is optimal, and what weekly frequency is sufficient?
The methodology used to answer the hypotheses was a combination of three different methods that complement each other: Correlation (from Meta-Analysis), Regression Analysis and Random Forest.
To understand brand impact, only campaigns which achieved a statistically significant brand lift impact as measured by a Meta brand lift study were analysed. Our analysis was based on 229 campaigns, 5 different industries and 11 countries. To answer the hypotheses, we divided the results into two parts: the first part focuses on understanding how to optimise brand lift cost efficiency of our clients on Meta platforms (reduce the cost per incremental person impacted); the second part focuses on how to maximise the number of people who are impacted by Meta advertising (increase the brand lift volume), if cost is no barrier.
Overall, we found that whether the goal is to reduce the cost per incremental person impacted or to increase the number of impacted people, the most important strategic areas to focus on are the targeting and creative strategies. Selecting the most effective combination of tactics in these two areas will predict at least 50% of the brand lift outcome.
Some of the results show how a broad target strategy performs better than other strategies. This becomes even more powerful when combined with custom audience exclusions. The results also show how the choice of campaign objective plays different roles for efficiency and volume of people impacted.
When it comes to creative, shorter videos played with sound-off were found to drive cost-efficient brand lift, though longer videos are better if the focus is to increase brand lift volume. One of the research conclusions was that viewability could not be used as a proxy for brand lift performance.
There are many factors that impact brand lift. Some of them are included in this analysis and some others cannot be measured or analysed. However, we can control a subset of the extensive list of variables predicting the cost of brand lift and the volume of brand lift.
The insights from this research should provide readers with a broad understanding of best practices for brand lift, validated across many brands, countries and industries.
The full white paper is available on our Global Intelligence Hub and will give you access to the complete analysis, and results.