How many devices do you currently use to access the web? Most folk would think two, their computer and their mobile, but right now are you reading this post on your PC at work or at home? Are you on your personal mobile or a work mobile? You may even be viewing via a tablet, a games console or perhaps an internet TV service? More than you thought!
The truth is that today there so many different devices with which to access the web, and understanding how users interact between them could become very valuable to agencies.
Current attribution modelling only really discusses attribution upon one device, but we need to look beyond this and start to incorporate multiple devices.
As it is tracking mobile Display presents a significant challenge, with a restriction to first-party cookies removing the ability to track post-impression conversions and the divide between applications and browsers further complicating things. Combining mobile and desktop measurement is tricky if you want to incorporate all online channels. Throw in a tablet and things can get very confusing.
Let’s take an example: you see an online desktop Display ad for a new product a couple of times. You then remember the product and search on Google desktop using a brand keyword, click on a PPC ad and convert. So far, so easy, right? But what if you see an ad on your desktop, then visit via a brand paid search ad on your work PC before finally converting on your mobile after downloading and using the advertiser’s app?
The display ad will not be attributed as an influencing factor and future media budget may be wrongly allocation. There’s also a good chance the user could continue to be served with re-targeting ads, even though they have already converted, and we all know that this can be frustrating for the user.
Rumours are this year could FINALLY be the year of the mobile (again… yes we are all sick of saying that!) and recent IAB figures indicate the same. Combine the growth in mobile with the ever growing tablet market and it becomes obvious that we need to be able to measure the influence of different channels across different devices, and figure out how we can create campaigns that use multiple devices in harmony.
So, what’s the solution?
Well one possibility is a system of unified logins across all devices. One login for phone, your desktop, tablet and web TV service would help bring together your online activity and maintain one online persona, which houses your preferences, settings, favourites and web history all in one place. Google are best positioned for this approach with their multiple desktop services, android operating software and Google TV, although it obviously ignores a certain elephant in the room (Apple, take a bow).
Google are already pushing the integration of Google + with all their other online services, so it wouldn’t be out of the realms of possibility to tie in mobile activity and have the ability to measure everything in one place. Tie it in with their tracking services such as Google Analytics and DoubleClick and you could have a very unique product that many agencies would be desperate to get their hands on.
The potential for this type of integrated approach is huge, if slightly terrifying. Imagine seeing an ad on your Google TV for an Airline, and at the end of the ad it provides you with real-time prices for a flight you searched last week! From an advertiser point of view this may be very cool, but given the recent furore over Google’s attempt to simply unify data on the desktop this might just send the world into a privacy meltdown. What perhaps makes people even more nervous is that Google are collecting this data for a use that is unclear. It certainly isn’t available to advertisers, for the moment at least.
Facebook could also be well positioned for such an approach. Rather that targeting ads based on how you behave online, they target ads to you based on what you have activity expressed an interest in, (i.e. liking a page) and this model could easily transcend into mobile and tablet devices. Of course this would only be limited to when people are actually using Facebook, but with users spending more time on Facebook than any other site, what better place to serve your ads?
There are also rumours that Amazon may one day launch an ad network, and given their reach online and ties into multiple platforms through various aps and services such as LoveFilm, this could prove quite successful.
These are just a few examples, but the truth is anyone who has a decent enough presence online, and across multiple platforms could be in a position of strength over the next few years if they can put together a worthwhile attribution model for advertisers.
Is any of this even possible?
Perhaps. There’s already a suggestion that the work PC could be no more – this would significantly help online advertiser as the less devices a person uses, the easier accurate attribution becomes. But let’s not kid ourselves; multi-device attribution is going to be a massive challenge over the coming years, especially given the somewhat murky privacy legislation situation at present.
There may not be an easy solution yet, but this is a challenge that is definitely worthy of some consideration and investment – the problem of device fragmentation isn’t going away. Whilst it is hard to imagine anyone but Google getting close to a useable off-the-shelf model that incorporates targeting as well as multi-device measurement many advertisers are beginning to look into ways to link media exposure to customer data across devices. This doesn’t make it possible to target advertising on one device based on behaviours on another, but at least it means we know what media is working. Maybe (probably?) Google are secretly at it already, figuring out how to make this all work. I guess the day you get served a desktop ad for a restaurant you searched for on your phone is the day you’ll find out.