As the ‘Mobile Takeover’ continues to infiltrate, reshape, and reenergize our industry, digital marketers are forced to quickly find newer, more holistic approaches to generating and interpreting user data. Just as the advent of the internet revolutionized how marketers get to know their audience (paper surveys and focus groups were quickly rendered obsolete), mobile devices have also triggered an industry shake-up.
Think about the way your customers engage with your brand. Do they visit your website at work? Look up your phone number on their phone? Use your mobile ecommerce platform?
More than likely, they engage in some combination of all these activities and many more, forming a multifaceted, cross-device brand experience. Just like any brand experience, the data you accumulate and interpret will help you build the most efficient and sound marketing strategy for reaching them. While browser-based web tools like cookies (more on them in a moment) have been excellent for monitoring desktop behavior, that is simply no longer sufficient for representing the wide and varied world of user behavior data.
It’s also important to avoid seeing cross-device tracking as merely a duality between desktop computers and mobile phones. In a recent post for Ad Exchanger, Allison Schiff points out why that interpretation falls short: “With the advent, still nascent, of connected TVs, wearables, and the Internet of Things, the concept of cross-device is expanding to potentially include anything that gives off a signal.”
While you may not currently place a high importance on reaching your audience with the right message via “wearables,” it’s important to take an open-minded view of how the ‘Mobile Takeover’ will continue to alter marketing strategies. As your average user’s roster of connected devices grows, it only makes sense to do anything you can to connect them to your message on as many platforms as possible.
Matt Lawson, Google’s Director of Performance Ads Marketing, is responsible for informing the public on how new developments like these will relate to AdWords functionality, and recently wrote a detailed post on how marketers and advertisers can use Google to help develop a cross-device marketing strategy. In it he talks about how Google has been at the forefront of cross-device data accumulation, referencing the 2013 inclusion of cross-device conversions to the AdWords mainframe.
These types of conversions, in his words, “are an important way to understand how your users navigate your site. And they’re not simply bonus conversions on top of your normal performance – they are your normal performance only you’re finally able to track it.”
In other words, cross-device data is exactly that: data. It won’t immediately bring new conversions into the fold; what it will do is illuminate a large swath of previously murky conversions and help you describe, quantify, and target the users involved. It’s a way of increasing the specificity with which you market.
Finding a way to turn that increased specificity into an improvement on your bottom line is a longer-term endeavor, but one that you’ll have no problem finding a strategy for if you’ve ever seen the benefits of buyer personas. The more clearly you are able to see your users’ engagement, the more effective your message to them will be.
Why Cookies Don’t Work on Mobile
Cookies have been the hallmark of internet advertising (and, to a large extent, digital marketing in general) since they hit the scene in the 90s. They are created when a website sends a text file to a browser, confirming that a particular user has loaded the site. Marketers use that process to execute what is known as “cookie profiling,” by which they can track a user’s activity online from a more general standpoint, thereby obtaining information about how they behave on many sites, not just their own. This is clearly a huge advantage for many in our industry – it enables online marketers to spend money on ads that will reach a specific subsect of their audience, one that has already been confirmed as filled with people who would be most likely to react positively and convert into a customer.
Not so much on mobile.
While it’s a misconception that cookies do not exist on mobile, the reason for that misconception is likely the fact that the mobile usefulness of cookies is so forgettable. Because a cookie is limited to the browser it is opened in, a link between user and online behavior quickly crumbles (sorry, couldn’t resist) when multiple devices are being used – almost always within a single day.
It’s also important to note that information obtained from a cookie within an app setting (as opposed to a mobile web browser) can’t be shared with other apps. Without that linkability, the information is made far less useful for marketers.
If Not Cookies, Then What?
Industry leaders are generally in one of two schools in terms of cross-device user matching: deterministic versus probabilistic.
Deterministic matching is, for the most part, relegated to giants like Google and Facebook. They are able to do this because users are so frequently logged in to the platforms in some capacity, that behavior from multiple devices can be linked together from the simple fact that they were signed in on both devices. In order for brands not named Google and Facebook to access these benefits, they have to integrate with one or both of them. This is known as the “walled garden” side of cross-device behavior data because the larger companies have access to this vast “garden” of valuable information, thus further incentivizing smaller brands’ integration with their platforms.
Probabilistic matching is based on developing algorithms that analyze huge amounts of anonymous data points and spits out likely matches. The specific attributes that are analyzed include device type, operating system, location data associated with bid requests, time of day, and others. With this route, there is less of a privacy concern, but the end product is also less consistent. Two of the new sub-industry’s leading names, Drawbridge and Tapad, say they can create matching algorithms that operate with around 70-90% accuracy. (Drawbridge uses no deterministic data to drive its algorithms while Tapad does).
Ultimately, it is still unclear exactly how much time, money, and effort should be spent on cross-device marketing. Because the method is so new and still so much in flux, it would be a mistake to put a large investment into a particular strategy, only to see it quickly fade from industry favor and see the data streams it produced become ineffectual.
There are, however, certain things that we do know. We know that with mobile usage sky-rocketing, the current, cookie-based user tracking methods are not long for this world. We also know that cross-device marketing will be even more important when different genres of internet-connected devices (smart televisions and wearables, for instance) become even more commonplace. The two schools of cross-device thought mentioned here will likely continue to adapt to the market (and each other) until one emerges as a consistently viable option for businesses of all sizes. It would behoove any entrepreneur, marketer, or online advertiser to keep a close watch on that process occurring, and be ready and able to take advantage of whichever strategy wins out.