At the dawn of the digital age, the volume of data being produced online was seen as the antidote to billion-dollar advertising decisions led by “experience” and a healthy dose of hope in
traditional media.
After years of trying out targeting, marketers have now realized it takes much more than a large mass of data to make the right decisions. It takes
intelligence.
So the industry turned to analysts and folks with titles like “decision scientist,” who were employed to make sense of all of the data and hopefully get
better at targeting. Hundreds of ad-tech companies emerged with presumably better and better targeting. But, despite all of that, digital advertising ROI was still not improving significantly.
The key turning point in the use of data occurred when the idea of intent was introduced into the equation.
Rather than assuming every male between the ages of 25 and 40
might be interested in buying sneakers, why not target only males we know are in-market for sneakers? Dubbed retargeting, this was the beginning of a broader and more compelling use of data in
figuring out the intent of users and personalizing content to them. With personalization, we can go even further to show this individual the actual color of the sneakers he may likely purchase, and a
pair of sneakers tailored to the kind of physical activity he may enjoy.
And, more than just targeting people who are showing intent, this approach allows for prospecting where large
audiences could be grouped into much smaller “micro-clusters” of users who share common interests, preferences, and behavior, to message them precisely.
This intent data,
combined with audience data, gives us a powerful arsenal of data and a level of precise messaging never been seen before.
But before we go about popping bottles of champagne
celebrating our digital marketing victory, it’s worth pausing to think about it some more, since we have now run into another problem: humans.
Human intelligence does many things well
— such as dreaming, hoping and innovating, which are integral to creating consumer engagement. It is we who create the big ideas and forward-thinking strategies. It is also we who invent
the machines and the software that collects and processes that data and harnesses it to target and personalize content. However, what human intelligence is not very good at is processing massive
amounts of data in real time to come up with precise ways of targeting and messaging consumers.
Machine intelligence, or AI, does this quite well—and can do so in real time. AI is
also able to detect patterns and make correlations about consumer behavior.
In addition to the ability of humans to imagine and craft emotionally appealing visual designs and communication,
humans also have an important role to play in ensuring AI is applied correctly. For example, a machine-learning algorithm may incorrectly blacklist a site as a terrorist property because it referred
to jihad when in fact it was a site about the history of Islam. Insufficient or incorrect data inputs could lead to wrong outcomes. This is one place where humans step in to better train the machines
to make the right decisions.
Striking a balance is not only an imperative, but also well within our grasp.
Humans create strategic goals, the vision of making personal connections at
scale. Machines predict the right collection and combination of individuals and content most likely to result in a purchase over and over again, millions of times.
It’s this kind of
balance we need to realize the full potential of all the data we collect.