The main takeaway was that the speed of change continues unabated. Details follow.
There is increasingly less distinction between “data” and “measurement.” That’s important, because data used to be any stuff that was represented by bits, and measurement used to be the numbers we used to decide what media to buy, along with other stuff. Now the two are merged — but clarity and rigor go out the window. When someone hands you a conclusion, you’d better know when and how the data was collected, and the motivations of every player in the chain.
Mobile-location data is becoming a distinct supply chain. There is a supply chain of specialists — many competitors — in cleaning, parsing, assigning meaning, syndication, etc. This was exactly the evolution of browser data. The key challenges now involve how we ascribe meaning to location. From the Ramp Up stage, one company publicly complained that they can see when someone with a mobile phone goes to a multiplex movie theater — but they cannot tell which movie the consumer’s watching! It’s tough all over.
The use of location data improves consumer understanding by a lot — and that understanding is applicable across media. It’s a whole new lens, not an incremental improvement. Location data is more interesting in terms of a holistic understanding of people than it is for targeting.
Still, if you thought seeing an ad today for a thing you searched for yesterday is creepy, just wait. For example, your location might suggest you spent too long in the bathroom. May I offer you some stool softener? (Made that up, but — it’s quite a plausible example. Ugh.).
Related to my point above, the privacy debate rages on. With mobile, you might say privacy is dead. But wait! The opposite might be true. We may be reaching a point where there is so much data, and so little signal, that economically viable use cases are scarce. Most people’s movements are so mundane that even if you know about them, who cares?
One group who seems to care are retailers. For example, mobile data will let retailers see competitive activity in the context of a consumer. Did Polly abandon Think Coffee in favor of Starbucks? Her mobile data will tell us. Perhaps we can lure her back with a ….
Attribution is under attack. Brands and agencies want models that show actual sales lift as a function of marketing. However, there are so many factors driving brand sales that it has been historically impossible to reliably say which factors contributed how much.
We fight the attribution gap on two fronts: 1) Magical math and modeling (Think Bayesian, multivariate, etc.; and, 2), predictive surrogates for sales, like attention and audience composition.
In a complex world, #2 has legs. It’s better to gain an understanding of how simple and accurate surrogates drive your business, mainly because the surrogates are tied to actions, while magical models are often backward-looking or unreliable.
A good KPI is like an old-fashioned compass: It’s always wrong, but never by much. Am I going in the right direction? The magical models, on the other hand, can be very accurate, or stone-cold wrong. But which?
Everybody’s looking for truth sets to make their segments better. However, the media industry prefers reach to accuracy. “Truthiness,” as a concept, is frightening.
GDPR (The new EU data privacy regulation) is scaring a lot of people. Where you stand depends on where you sit. Walled gardens will be advantaged since they control the collection and distribution of their data, and thus compliance. For this reason, the regulation is likely to support Google, Amazon, and Facebook at the expense of smaller publishers and ecosystem companies.
I struggle to see how European interests are hurt by failings of privacy, but it’s easy to see how they are hurt by giant American tech. So, the regulation may hurt young companies that bring innovation, while helping the American tech behemoths.
Nice going, EU. Narcissistic, yet wrong. You are catching up!
If you have ever wondered what data scientists talk about over coffee, I feel sorry for you. But if you agree that data is driving the business world (aka “The New Oil”), maybe you should wonder. If you do, I hope this report has been helpful.