It’s hard to imagine a world where, if you went into a shop and bought something, the credit for the sale would go to the last ad you saw, some point-of-sale material, perhaps, or something on the side of a bus. But that’s the way online treats advertising.
Looked at like that, it was no surprise that Microsoft announced a couple of weeks ago that it was beta testing something called Engagement Mapping, which aims to provide a more complete view of how people respond to online advertising. Or that immediately afterwards Google confirmed that it, too, was working on a similar project.
“Last click wins”, as the current attribution practice is known, has become the norm, partly because there is an apparent causal link. If someone clicks on an online ad to go through to complete an action, whether that’s to make a purchase or to book a test drive, it seems reasonable to assume that the ad they clicked on played some part in the decision to click. They responded to a particular advertising message. It’s much harder to prove a relationship between a purchase and the ad on the side of the bus passing by as they walked into the shop.
But that link is clearly far from the whole story, particularly in complex purchasing decisions. The main reason the last click dominates is because it’s easy to measure. It’s clearly much harder to track the customer journey online, taking into account all the ads the customer might have seen. It’s even harder if you want to add in all the information sites that person might have visited, such as those carrying reviews or price comparisons. Try to complete the picture with information about all the offline advertising and information sources the customer might have been exposed to or consulted, and it becomes all but impossible.
The online advertising industry has been concerned about last click wins for a while now, mainly due to the continued perception of online as a direct response medium. Those people with an interest in bringing brand advertising money into the industry have recognised that last click wins is of little use to people in charge of those budgets.
Brand advertisers want to know about reach and frequency, neither of which the current system is able to deliver. As Google and Microsoft square up for a massive struggle for online ad spend, display looks like becoming the key battleground.
What everyone ultimately wants, of course, is a measurement currency that enables all aspects of a customer’s journey to be tracked, offline as well as on, that allows the effectiveness of all channels to be compared, and therefore makes it possible to optimise campaigns across all those channels. However, the problems involved in that are huge and, in some cases, probably insoluble. What Microsoft and Google are working on is little more than the next step. It will include data about other ads seen by customers before their eventual click through, but it won’t include data about the other sites they may have visited for information, which don’t carry relevant ads. And it clearly won’t include any offline elements.
The reason Microsoft is able to take this step at all is its $6bn (£2.9bn) acquisition of aQuantive and its online ad-serving technology Atlas last year. Google, on the other hand, is expanding from its traditional search base into display and video, and finds itself needing to address the role of search beyond the final click. The search giant’s planned $3.1bn (£1.5bn) purchase of online ad-serving and tracking firm Doubleclick, announced last May, would augment its technology in the same way Atlas will Microsoft. The deal was approved by the Federal Trade Commission in the US in December, but remains subject to approval by the EU.
One of the reasons for the EU’s concern has been the privacy implications of one company owning so much data about people’s behaviour online – concerns that these latest moves will throw into sharp relief. As Web analytics expert Matthew Tod explained to me when I interviewed him for the NMA podcast, they represent a shift from a campaign-centred model to an individual-centred model, which also has the possibility of linking previously anonymous online behaviour to purchases, and therefore to an individual’s identity.
Previous concerns about online privacy have been addressed in two ways. The first has been that the data collected doesn’t allow individuals to be identified. The second is that users can delete the cookies that make the tracking and data collection possible. Up to now, the online industry has worked on the basis that the first argument is sufficiently reassuring that only a few people will opt out, and that collection of meaningful amounts of data will remain possible. But once people realise that Google or Microsoft know who they are, and where they’ve been online, that assumption looks a lot more shaky.
Michael Nutley is editor-in-chief of New Media Age