This was originally published in ExchangeWire
Elsewhere in marketing, the user is front and centre. User experience, customer engagement, outside-in marketing – all are buzzwords that remind us that we’re selling to real people, not robots. Yet, an equal and opposite force is also at work. Alongside discussions of UX and CX, marketers drive towards increased automation, programmatic strategies, and measurement derived from mind-bendingly complex algorithms. Attribution typically operates in this latter paradigm.
This is the wrong approach. Marketing attribution should embrace people and platforms. In fact, users – or, more specifically, user ID mapping – separate limited attribution measurement from holistic analysis.
User ID mapping takes all of those measured activities and connects them back to a single coherent user profile. This was historically accomplished primarily through browser cookies, so it was relatively easy; marketers could claim a decent understanding of how single users acted as they navigated the digital landscape. It was also overly simplistic, lending itself to early attribution models like last-click. We knew most of what a user did, so we could make a decent guess at why they converted.
Recently, however, creating comprehensive user profiles has grown more challenging. Cookie-based mapping breaks down across multiple devices. Blocking mechanisms render traditional tracking tools useless. An attribution platform, like ad servers, can provide their piece of the puzzle; but isolated slices of data provide a narrow view of the world. Users are doing more, across more channels, than ever before, and single-stream profiling can no longer keep up.
Marketers need to understand behaviour from the user perspective. It’s critical to omnichannel success and marketing effectiveness. But outdated methods aren’t the only issue. A vast amount of data, and the resources to properly crunch it, do exist – but they live behind the walled gardens of ‘The Big Guys’. Google, Facebook, Verizon, and the like, know a tremendous amount about what everyone does, but they keep that information very close to the chest. That means the rest of us must come up with innovative ways to better understand our users.
Enter the user identity graph. User ID graphs connect the science of attribution to the art of customer-centric marketing. They compile all types of identifiers into a single profile, across devices, browsers, websites, and platforms. From server identifiers and device IDs, to login details and third- and first-party cookies, every action that a single user takes on his or her journey to conversion is meticulously accounted for.
[Side note: If you think this sounds creepy, you’re not wrong. But while companies can, and do, collect vast amounts of data on what each of us does all day, that data remains completely anonymous. No information that is not controlled by the user – via privacy settings in Facebook, for instance – is available. So, while they do know everything about you, they don’t know who you really are.]
Back to the graphs. Once the system gathers each piece of disparate data, it connects the dots. Think of a user graph like your network on LinkedIn. Some data points can be immediately synced via obvious overlap – your first-level connections. Others must be added to the overall puzzle through second- or third-level connections. This is where the complicated data science comes in. The algorithms interpret the data in a way that makes a million separate actions cohere into an engagement timeline of a single person.
But back to those walled gardens. Most systems don’t have access to all the data, all the time. Holistic user graphing often results from a network of partners coming together to share expertise. One may provide details on social networks, for instance, while another focuses on particular devices. This ecosystem further bypasses traditional, one-platform attribution by relying on the most relevant data from the most trustworthy sources. It also provides a unique, and, in my opinion, equal-if-not-better, alternative to The Big Guys’ user behaviour analysis.
The result of all of this data crunching, the user ID graph, is meaningful on multiple levels. As a view into a single user path, it allows for real-world validation at each step of the omnichannel marketing campaign. Like an interview or focus group, it gives marketers a specific, targeted window into how a real person – not a cookie – makes decisions and takes action.
The graphs provide equally powerful insight when rolled up into aggregate. User behaviour can be analysed by myriad dimensions – channel, publisher, time of day, device – while retaining the accuracy and breadth of the entire conversion path. This 30,000-foot view delivers high-level insight, particularly relevant at scale, when single user analysis is no longer realistic or actionable.
Comprehensive user ID graphs are one piece of the complicated attribution whole. But, in a data driven marketing system, they provide valuable perspective. We humans are unpredictable, dynamic, and have lived omnichannel lives long before that word was a twinkle in a marketer’s eye. Only a holistic view into our purchase paths, not our robot proxies’, offers the information and insight we need to make the smart decisions that drive differentiated results.