Digital channels have been a boon to the marketing ecosystem over the past twenty years.
They’ve brought much vitality and innovation to the practice of marketing across all sectors of the economy. And one of those crucial innovations was the development and wide adoption of multi-touch attribution (MTA), a key tool to help marketers unravel the mystery of the consumer’s step-by-step journey along the path-to-purchase.
The current environment, however, is much more complex than when MTA was first introduced:
Privacy regulations are hitting fever pitch around the world, media fragmentation is
accelerating, and third-party cookies—the backbone of traditional MTA systems—are now
clearly on their way out, along with mobile ad IDs (“MAIDs”) and other perishable forms of consumer identification.
To top it all off, walled gardens are living up to their names and are actively restricting
marketers’ access to granular user-level data on their platforms.
For some industry observers, this is all too much of a perfect storm for MTA to survive. But
I don’t share that view. The media buyers and sellers I interact with in my role at Neustar are very proactive about shaping a new, more efficient ecosystem for all. They’re much more engaged with one another than they’ve been in the past, and they’re helping us design robust identity-based solutions to open up new opportunities for themselves and sustain the growth of their business in a post-cookie world.
Differential privacy is one of those key solutions.
Data Science Holds the Key to Privacy-Centric MTA
You’ve probably heard the term “differential privacy.” Like many marketers–even those who specialize in data and analytics–you may not know exactly what it is. That’s OK! The thing is, there isn’t one standard approach or definition of differential privacy. In many ways, we’re constantly optimizing these methods to best suit marketer needs as the ad ecosystem evolves.
Right now, the best data science minds in marketing are finding smart ways to surmount challenges to privacy-safe MTA and measurement. And those very same people are ensuring that these new marketing measurement solutions not only go far beyond the old days of anonymization, but they will also continue working and evolving no matter what tomorrow’s privacy landscape looks like.
What’s important is that today’s marketing data scientists are developing cutting-edge privacy-preserving techniques that incorporate elements of differential privacy. These include employing machine learning and measuring in aggregate to achieve precise measurement with very high anonymity levels.
Advanced processes like these attribute conversions in a privacy-centric manner without the use of third-party cookies or MAIDs. At Neustar, they are built in conjunction with digital platform partners to ensure that accurate campaign reporting and performance analysis works inside those ever-important closed platforms. And, it happens without ever sharing user-level data between the closed digital platform and the measurement platform.
In short, differential privacy methods allow marketers to measure for MTA according to advanced analytics metrics without diminishing the validity of the data.
New Privacy Methods Help Brands See the Full Picture
There are already several brands using these measurement techniques.
Here’s an example. Let’s say a telco or CPG brand wants to optimize their marketing mix. Before these new methods were available, it would have been impossible to see holistic user-level ad exposure inside closed digital platforms.
That’s been a big problem for many brands. Without knowing whether a customer who bought a product in-store saw an ad in a closed social media platform, the brand simply would not have had all the information it needed to optimize properly or understand how that media performed relative to the other media channels in their plan.
Consider this common scenario from before we could take advantage of these new privacy methods: If a brand ran display advertising inside a social media platform, even though a customer who bought a brand of cereal or new mobile device might have seen one of those ads, the brand only would have known they were exposed to an ad if the customer visited the brand site after clicking through from the platform.
That’s a serious gap in important information marketers need to achieve true multi-touch attribution.
Today, with innovative, privacy-centric methods, brands can measure inside closed platforms and compare the return on ad spend across their entire mix, which means they can make that connection. Marketers can see more details about things like ad spend and exposure because of these forward-looking techniques. Now armed with the new information and a more complete picture that is enabled through privacy-safe approaches, marketers can optimize their marketing investments across all channels with confidence.
I’m excited about what data scientists and engineers are coming up with. It means marketers can really understand how consumers are responding to marketing campaigns, even as the privacy landscape changes.
The takeaway? Privacy is not an obstacle to reliable and holistic multi-touch attribution and campaign measurement. It is the very thing that is allowing us to collaborate and innovate within our industry to future-proof marketing and ensure a more secure, and more trustworthy environment for brands and consumers.