It was announced today that Twitter will open up its self serve ad platform to around 10,000 business clients. With numerous studies revealing how poorly brands typically do at engaging with fans on social media it got me thinking as to how effective all of this will actually be. Whilst search advertising often produces good results, banner blindness is commonplace on the web. Will ads segmented based upon your interests do better at producing responses?
New research suggests that it just might do. The paper suggests that ads that closely align with the values and identity of each individual work incredibly well, even if users are barely even paying attention to the advert. Place the same ad on another web page however and its impact plummets. So, they suggest, the alignment with our interests that is possible due to the huge amount of personal data the social networks have on us all could prove hugely valuable to marketers.
The authors of the report call this approach implicit self-referencing. In other words, people don't need to own or have purchased a product from a brand in order to associate with it. You don't need to have owned a Mercedes in other words to identify with the luxury car. Therefore placing an ad next to our personal interests is a potent combination.
How they tested the hypothesis
The researchers conducted three experiments to test their theory. In the first they presented participants with a set of made up brand names on a computer screen for approximately 30 seconds. Whilst the branding was on screen, terms relating to either themselves or someone else were subliminally flashed onto the screen. After they had done some random tasks they were then asked to categorize and identify the brands. The brands linked to the self identity received much better responses, with participants suggesting they would be much more inclined to buy from those brands.
The 2nd experiment expanded on this idea to explore the role of self-esteem in this process. This experiment was identical to the first but recorded the participants self-esteem. The results showed that the user-brand connection was much stronger in people with high self esteem. Of course, research has suggested that Facebook users are particularly narcissistic, so this may be a boon for marketers.
The final experiment applied these findings to the real world by testing them on actual social networking sites. Participants were asked to compare Hi5 and Facebook. They were asked to compare either a generic version of a page on Hi5 to either a generic Facebook page or one with personalised information and ads. The ads on every version were for the same fictitious car company, with each site displaying a unique car from that brand.
After answering some questions on each site, they were then asked some questions about the brands promoted in the adverts on each site. As with the earlier experiments, participants were much more likely to choose brands on the personalised Facebook page than on the generic version of either site.
The researchers suggest that our subconscious awareness of ads on sites such as Facebook is incredibly powerful if the advert is aligned with our interests and identity. That this happens often without us realising could help marketers to overcome traditional banner blindness.
How valuable is the subconscsious?
All of which sounds great, but in reality it begs the question of how valuable is this subconscious desire to marketers? After all, research last year suggested that banner ads on Facebook achieved half the click through rate of banners displayed on other websites. The click through rate on Facebook ads was a paltry 0.05%. This figure was slowly getting worse, dropping 0.01% from 2009, which suggests that any efforts by Facebook to target ads better was not delivering results to marketers.
So the question would appear to be, are these ads simply not targeted sufficiently well, or is it the case that subconsciously observing an ad will benefit the brand, just not via a direct click through to their website?
Neurology would suggest that is quite probably the case. How you measure the ROI on that though is anyones guess.