Many of the industry’s best marketers, data scientists, and digital advertising technologists were present to share the latest developments on the topic. Needless to say, none of us left with definite answers because the subject is still too abstruse, and ad technology and analytics platforms are just beginning to catch up with the multi-screen universe. However, it was a great opportunity to start asking the right questions, which is how we will ultimately arrive at real solutions in the long term.
As a one day event, Omni.Digital was packed with subjects varying from Cross-Device Data Collection & Attribution, to a case study presentation on the influence of SEM on brick and mortar sales, to the impending age of marketing with the Internet of Things.
However, the underlying theme really focused on the increasing complexity of the customer journey. No longer can we be satisfied with a linear channel approach, where 1 impression on 1 device leads to 1 click which leads to 1 conversion.
The above attribution model, in truth, never properly reflected reality. Very few people actually respond directly to one ad on a single device to make a purchase decision. Recent studies have shown that 90% of Internet users now use more than one device to accomplish a task over time.
In recent years, analytics vendors tried to evolve us away from last click/touch by developing a multi-channel approach to measuring the conversion path. As a result, digital media pundits fervently killed the sacred cow of last click attribution (and that’s a very generous statement because the fact remains that most advertisers still function on the good ol’ post-click conversion in channel silos) and moved to the multi-touch model offered in GA, Adobe, and DCM, or through specialized attribution solutions like Visual IQ, Adometry, and C3 Metrics.
That was a good start, but as more and more consumers began to use their mobile devices (smartphones and tablets) for both transactions and product information, multi-channel attribution (MTA) was no longer enough to gain a proper single customer view.
That’s how the whole conversation about cross-device data and attribution started in the industry. How do we follow the customer from one screen to another and attribute the right score to the right device on the right channel? Throughout the conference, this was the BIG QUESTION on the panelists’ and speakers’ minds.
Now, what I appreciated about the event was that just about all the presenters were humble enough to admit that no one really has the answer. At best, we have new technologies to allow us to test out models that will eventually take us there.
The big question is therefore, what are those new technologies? In that respect, AdExchanger (the organizers of the event) did a pretty good job of assembling some of the key start-ups emerging on the market to fill the new cross-device data and attribution niche. It so happens that Adviso was already working with some of these new players, but a few of them were new to me.
So the moment I came back from Chicago, our team began to reach out to all these new and interesting players. We’re currently in the process of mapping the new range of options and possibilities that stem from their various offerings, and so far it looks very, very promising!
Having said that, please feel free to ask us questions and contact us if you need more details or want an in-depth consultation for your organization on this murky topic. We know these technologies will have to connect with your backend data warehousing, CRM, and digital analytics infrastructure, and of course that’s when things can get complicated. We understand and we’re here to help 😉
Now, coming back to the question of “what are those new technologies?” Simply listing out companies without context probably won’t be very useful. Instead, we’ll share with you three important buzz terms surrounding the new world of omni-digital and cross-device/channel marketing that you should start familiarizing yourself with. Soon, it’s all you’re going to be hearing about, like the word programmatic three years ago.
There are two ways of matching people to multiple devices. The first, called deterministic, is the simplest. Basically, if you’re Facebook or Google for example, you possess a very large database of users who have to log in to your services in order to access their account (i.e. Gmail, Facebook Profile).
You can then keep track of which desktops (via cookies) and which smartphones (via device ID) your users are accessing your service from, and simply merge them into a unified ID. In fact, more and more DMPs have enabled deterministic matching and fusing across devices, namely Adobe Audience Manager and Tealium AudienceStream. From there it is easy to retarget those people across devices. AOL 1 is another DMP that has an abundance of 2nd party deterministic data with lots of matched devices, even in Canada!
Arguably, as I mentioned in a conference I gave on native advertising in 2014, Facebook, Twitter, and LinkedIn are the first pure cross-device marketing players on the market, because they have so much deterministic data.
As far as the probabilistic model is concerned, that’s where data scientists and very highly paid engineers come in, because this model needs both a tremendous volume of big data and an extraordinarily complex mathematical model to arrive at what we call a proprietary probabilistic ‘device graph.’
Luckily for marketers, you need not go through all this pain. A few new companies are emerging who specialize in the probabilistic matching of people and devices. The media team at Adviso started working with some of them two years ago, but a streak of new players have emerged since. If you need more details on which of these you might want to work with, feel free to contact us to find out about our consulting services.
This buzz term is essentially the output of the two previous ones. That is to say, once you’ve done your matching, either probabilistic or deterministic, you’re left with a cross-device identity. As the browser cookie becomes more and more obsolete, this new term is fast taking center-stage in the audience data game. Both cookies and device IDs can help us in the cross-device world, but only if they’re combined with other variables to assist in mapping a single identity or customer view. This single view is the CDI, which forms the basis for an identifier in the device graph mentioned earlier.
In the good old days, we were satisfied with setting a target frequency capping goal across the various channels of our media plan. Today, much of that is becoming difficult to measure when we take into account what’s really happening across multiple devices and screens.
If we set a frequency cap on a mobile ad placement and another on a desktop placement and another on a tablet placement, how do you know which users are connected to which devices across all of those placements?
Using a silo approach, you don’t. But using a device graph and multi-screen targeting as a primary data layer on your campaigns (mainly on programmatic for now), you can build truly people-centric campaigns instead of channel-centric ones. Hence, you’re able to make use of new targeting tactics like cross-device messaging for example. To quote a cross-device data partner of ours on this topic:
[Cross-device data] makes for more effective media planning and reduced media wastage. For example, true cross-device reach would allow you to optimize a media plan by removing mobile placements that aren’t delivering incremental reach to desktop placements.
In conclusion, the Omni-Digital world is the new frontier for marketing, and the arms race is on among technology and media vendors and advertisers alike to reach a viable and scalable model for targeting and attribution. To ignore this transformation in consumer behaviour and the communications challenges it brings could make or break your business in just a few years (especially at the pace of technological changes and advances). On the other hand, those who embrace this difficult challenge and pioneer the conquest of this new frontier could earn a cutting-edge competitive advantage, and unexpected players could end up on top overnight (think Google and Facebook when they appeared on the market out of the blue). The question is, where will you be?
If your organization needs consulting on this subject, don’t hesitate to reach out to Adviso for help. Our focus over the last two years, and particularly in 2016, has been to figure among the first digital agencies in Canada to reach and master this far-reaching frontier, and we’re well on our way.