Timeliness is becoming the determining factor for online success, but the data systems required to achieve it are creating a digital divide.
Data is the great obsession of interactive marketing. At first, unprecedented measurability was the great claim of internet advertising. More recently battles have been fought over how much data people are prepared to share with marketers and on what terms; and in recent years we’ve seen the rise of new forms of targeting based on new ways of collecting and manipulating data. But only recently was it suggested to me that not every marketer has equal access to data.
Interviewing Eddy Swindell of email marketing provider ExactTarget for last month’s Digital Strategy supplement, he said that “there’s a divide being created in the industry between the data haves and the data have-nots”.
At first this seems counter-intuitive. After all, the ability to know who has seen an online ad and how they responded to it is the basis of internet advertising, and as a result, everyone has access to far more data than ever before. But what Swindell was talking about was not so much the amount of data available to individual marketers, but their ability to make use of it.
The American writer Dorothy Parker once quipped that the rule for making jokes in conversation was better never than late, and we’re finding that this is also true of online marketing. It’s become common practice to target car insurance ads at people who’ve just bought a car, or to place ads for investment brokers at the end of share tips rather than before them. But timing is much more sensitive than that.
Take retargeting, for example. It works by targeting people who have browsed an e-commerce site but not bought anything from it with ads for the things they looked at. Because you’re advertising to hot prospects, the returns can be spectacular. But, like any form of targeting, it has to be done carefully or it starts to annoy people. One of the key parameters turns out to be the time between the initial browsing and the follow-up ad being served. Do it promptly and people see it as a useful reminder; leave it too late, or carry on for too long, and it becomes at best irrelevant because the person may have bought the item elsewhere and at worst damaging to people’s perception of the brand.
One of the key parameters of online targeting turns out to be the time between initial browsing and the follow-up ad being served
The retargeting example also highlights another aspect of the problem, which is the need to marry up data from a number of different sources. This issue is well-recognised; it’s the much-discussed need to have a 360-degree view of the customer, and it’s the force behind the breaking down of the silos that have built up within companies around the different online marketing platforms.
This is a challenge, because there are so many types of data, coming from so many different sources. There are response rates to online advertising; open and click-through rates for emails; clicks on both paid and organic search terms; browsing and purchasing behaviour on the website; and sentiments expressed in social media.
But the challenge might be better described as the need for a four-dimensional view of the customer, because of the speed at which people’s status can alter. It’s the people whose systems don’t allow them to respond to those changes quickly enough that Swindell describes as the data have-nots.
To make things even more challenging, ’quickly enough’ in this context is coming to mean ’close to real time’. This is another impact of social media. The days of being able to do a data run overnight and merge different sources of information have gone, because customers’ attitudes are being shaped as much by the response to the TV ad in the social space as by the TV ad itself.
So if the first stage of the data problem was collecting it and the second was integrating it, the third stage, the one companies are now beginning to perceive, is being able to update and call on that data in real time. And of course, each level of complexity adds another level of cost, which is where the divide is opening up.
The implication of this divide is that it becomes harder for small and medium-sized companies to compete in these more targeted, higher return areas of advertising. Big organisations, with big data management budgets, can do it. Smaller ones will struggle to afford it.
This may turn out to be a temporary blip rather than a permanent state of affairs. The cost of processing will continue to fall as the market grows. And for many companies, the challenge of the 360-degree customer view will continue to be the most pressing. But when the most important component of relevance is timeliness, marketers will ignore the fourth dimension of data at their peril.