We know from previous research with GroupM that content optimised for Twitter* is significantly more impactful than running original TVC versions. Together we found that optimised ads generate a +33% increase in emotional engagement and +19% in ad recall vs. original TV ads.
Having established that optimised content makes a difference, we wanted to delve into placement. We discovered that the same ad doesn’t always perform in the same way across platforms; the environment can be a critical factor in how well it cuts through.
In partnership with Kantar, we used ad replacement technology to test the same ads in users’ feeds — across Twitter, Facebook, Instagram, and YouTube. Respondents then took a survey. They did not know they were being exposed to test ads and they were asked to browse the platform as they normally would.
Twitter was more likely to drive video views; 77% watched an ad for 2+ seconds vs. an average of 59% for other in-feed platforms. Twitter was also more likely to achieve completed views.
Twitter also drove more brand impact with higher uplift for top-of-mind awareness; +24% vs. other in-feed platforms.
These findings are consistent with the way people use different platforms. Twitter is a very different space to other in-feed platforms; the leaned-in state and discovery mindset people have when using Twitter impacts how they respond to ads. There is more openness to seeing ads in this environment and ads on Twitter are cited as being more relevant to them versus ads they see elsewhere.
In summary, advertisers should not assume that all mobile platforms are the same and will yield the same results. Optimisation of video ads is crucial and advertisers need to consider the relative strengths of different environments.
*Optimised for Twitter — video ads that are optimised for Twitter include three main rules: short & focused, persistent & prominent branding, audio agnostic (sound-off strategy).
Kantar, commissioned by Twitter, Environment Matters, 2019, UK. Study used ad replacement technology to test how 10 different video ads performed on Twitter, Facebook, Instagram, and YouTube. Base: weekly users of each platform, n=4,200 (3,200 exposed and 1,000 control).