The promise of targeted online advertising powered by big data and algorithms is, at last, the evidence will be there and advertising can be more supported by the numbers.
The current concerns over Facebook and social media demonstrate challenge that society as much as business needs to face. We get many services for free funded by advertising. That’s not new, of course. TV has been supported by that model for more than 50 years. Newspapers, both free and paid for, have always been supported by ad revenue.
The arguments over fake news and election / referendum outcomes being subject to manipulation of social media doesn’t look like going away any time soon. For ad-supported companies it is in their self-interest to claim that targeted adverts work, but also that they cannot be abused to undermine democracy.
Personally I am sceptical of the evidence to date that the activities of some organisations changed the outcomes. Part of that comes from experience of the ability to use the data effectively.
Last month I had an advert asking me if I was happy with my bra fitting. If that is the best they can do with all the data they about on me and the powerful algorithms that crunch it, then the idea that an election can be swung by these techniques is stretching credibility.
One former colleague had it suggested to him that a Fitbit might be a suitable Mother’s Day gift for his 94 year old mother. Conspiracy theorists may well suggest that these ‘occasional’ lapses are there to make me sceptical. I remain to be convinced.
In workshops over the last few years, I have, in a number of sectors, picked up a growing belief that advertising is becoming less effective. This is not just across TV, print, radio and digital shifts, but overall. A particular issue is the extent to which brand building is supported by ads. Social media is proving very good at trashing brand reputations when service is not delivered but does it work the other way round?
I know a number of small businesses who have told me that advertising on local radio generates revenue and profit far more effectively than social media, which generates lots of hits and likes, but proving business outcomes is far harder. Perhaps we are becoming so overwhelmed by adverts that we’re just switching off.
We do seem to be good at inventing new ways to spend ad money. Take, for instance, ‘sentiment analysis’. At a recent talk, the speaker gave this example: ‘thank you Brand X for casting a cloud over my son’s birthday’.
All the tools he tested said that was positive about the brand. But, how do you interpret this example: ‘one of life’s minor victories is never having to deal with Brand Y again’.
I’m not arguing that it can’t work in principle, but are today’s algorithms out of the kindergarten yet?
Another is ‘influencer marketing’. Does it work? This small example appeals to me because they are trying to learn what really works, rather than hype the idea. I will confess - like many men - to being ‘influenced’ by Scarlett Johansson, but I’m not really sure what I’d buy on her recommendation. Correctly, you may observe that I am not, sadly, in her demographic. In a world where the money is with the older generation we still seem to target the younger generation more than perhaps we should.
The major ad companies are among the most shorted companies on the stock market. The issue is whether the current model is broken or whether advertising is broken per se.
There are an increasing number of sceptical articles in the media about whether big data is generating the actionable insights claimed at the current state of deployment.
For the last year I have tracked my credit card spend on all sorts of goods and services from holidays and gifts to music, film, home and garden. Of nearly 300 items, over half came from me searching. Around a quarter came from people I like in music, books and film. There are around 10 items which came from friends recommendations, so I couldn’t say. I can’t find a single item I’ve bought from a targeted ad at me.
The only thing that works for me is vouchers offering discounts for things I already buy. I haven’t changed brand on the basis of a voucher though.
Much of the focus of the recent claims about Cambridge Analytica are around alleged wrong doing. We will see if those arguments stack up, I am sure. Much more concerning is the sustainability of the ad supported model over the economic cycle. In 2007-8 the digital businesses were still small relative to the overall market, but not going into the next recession, whenever that arrives.
Back in 2012, I attended a lecture where an economist suggested that if targeted ads were more effective than traditional then you would need less advertising to generate the same outcomes. However, he suggested then that the stock evaluations were priced on more, not just better advertising. To sustain that would need higher trend growth. That hasn’t happened this decade. The prices are much higher than in 2012.
Half my posts are wrong, I just don’t know which when I write them.
About the author
Chris Yapp is a technology and policy futurologist. Chris has been in the IT industry since 1980. His roles have spanned Honeywell, ICL, HP, Microsoft and Capgemini. He is a Fellow of the BCS and a Fellow of the RSA.