First party data. Email. Surveys. Panels. Small data. Emails. Surveys. Smaller segments.
When "big data" took center stage a few years ago, many of us in the insights, research and analytics industry became enthralled with the the idea that data work needed to be focussed on massive data sets from passively generated sources. The talk was third-party databases, DMP's, modeling and look-alikes, programmatic media, AI, and Machine Learning. We wanted millions of rows and columns and be able to use AI and Machine Learning to pull stories from it. IBM Watson ran expensive ad campaigns to promote around this idea that insights would be magically unveiled even if we were not sure how.
As a result, things like email, surveys, segments of a few hundred people were considered a bit "old school" and boring. Although people who knew the game knew that small permission based segments were good seeds for building models, still it was not the lead part of any data discussion.
And you can see why: having more access to existing data that was more granular and based on behaviors seemed like a good thing. Good for businesses that wanted to be more relevant and more resourceful. And good for consumers, who would be able to login to places more easily and get a more relevant and less-paywalled internet.
[ Note: no matter where you stand on the targeting vs privacy debate, I think we can all agree it is not as good as it is promised. All you have to do is look at the way ads re-target you after you have already bought the product online, it still doesn't feel optimal. That said, there is a lot of value in the idea and concepts of more intelligent web ecosystem for both users and platforms to agree on when and how to pay for content, either with money, data or some combination. ]
But now Google, Apple and others are saying hold on: let's change how browsers and the plumbing of the internet work. In the name of privacy, GDRP, and maybe some competitive advantages, they want to go back to prioritizing consent-based, first party data.
The pendulum has swung back as data policies and ecosystem fight for better privacy controls but also better business models (and control). "Out" now is the idea that a business should build profiles attached to third party cookies. And "In" are approaches that use more direct permission first party data techniques, often with registration and email at its core.
The result of that is making email, surveys, panels, and what was once small data...big again. Some people have been saying that email - good old fashioned email - may be the new universal ID (say in a Gandolf voice: "one ID to rule them all") as is the sentiment of Ron Jacobson in this thoughtful AdExchanger piece from Sept 2020.
And how to replace the third-party cookie has become a major industry debate, as reported in Mediapost on March 4th 2021:
Google Triggers Mostly Backlash After Rejecting Plans To Replace Cookies With Personal IDs", there are many reasons why this is happening: including consumer concerns, changes to privacy laws as well as the corporate strategy of bigger platforms increasing their value (and their data) in the ecosystem.
Per the article: "The advertising industry shared mixed thoughts Wednesday about Google's news confirming that once the advertising giant phases out third-party cookies it will not build alternate identifiers to track consumers as they browse the web, or use them in its products."
And so for smaller publishers, it was ALREADY important to gather first party data for increasing reader revenue and paid subscriptions (see NiemanLab 2019). And it still is. But now it is doubly important for publishers to do that work of getting users to opt-in to cookies, to register and become a known user who accepts that the site will know who you are, not only to support the audience revenue but potentially to fill the data void left by deprecating cookies in the ad revenue side. While appending data will still be useful in some cases, it will take solid first party data to create the core of the profile, and that likely means email, and a first party cookie. So small data is big again.