The value of data in advertising

The value of data in advertising

Legal professional with strong interest in antitrust, strategy and new technology.

We are not unfamiliar with the concept of paying for certain goods and services by giving away intangible things, whether it’s our eyeballs scanning a newspaper page or the signing away of our data to the internal databases of companies both large and small. This model helps certain types of businesses solve a chicken-and-egg problem, whereby their product/service depends on getting two distinct groups to come together (content viewers and advertisers). Without getting both sides on board, the businesses can’t grow and thrive.

There has always been this notion of paying with our eyeballs. Viewing advertising that we don’t like is a proxy for paying with our bank balance. Hence why some newspapers are given away for free and include lots of intrusive advertising, making it purposively hard to find the contents page and other interesting stories. The Evening Standard and City AM are good examples from the U.K. On the other hand, others (e.g. The Economist) charge a fee and include less intrusive advertising. This two sided-market was disrupted by the growth of online advertising, where user-data crunching has rendered the advertising we see more and more relevant to things we like seeing. We now derive positive value from lots of the advertising we’re subjected to (though this is a sweeping generalization), and that positive value makes our eyeballs even more valuable to advertisers. In the context of targeted advertising, we don’t pay with our eyeballs; we pay with our data.

On the other hand, there is the notion that overly relevant advertising is just creepy. It turns my imagination away from the subject of the ad, towards a repulsed puzzlement of “How the hell did you know that?” So there must be a point where the value we derive from targeted advertising stops increasing with the amount of data the advertiser (or intermediary, e.g. Google’s AdWords product) has, and starts decreasing.

Something I threw together to illustrate:


At point A, let’s say, the advertiser has garnered enough information to broadly target you individually: your sex, your age, where you live. It collects more data, like where you work (from Facebook), where you went to school (from Facebook), who your friends are (from Facebook), what you’re into (from Facebook), and then it knows to show you a pretty convincing replica lightsaber because you checked in at the cinema last Thursday and have a solid disposable income, working at McKinsey. But by the time it starts tracking your movements and the websites you visit in realtime (past point B), and then immediately shows you an ad based on that information, your positive value turns slowly into disgust.

The point is that data-collection can not keep increasing in value over time. There will come a point where we are handing over data that possesses no value in and of itself in the context of targeted advertising. Large firms that managed to collect the most data first will see the value of their data mature as their services reach as many people as possible, who have handed over as much useful data as possible.

The decline of targeted advertising.
Targeted advertising is in decline because the growth of supply of advertising reduces the value of individual ads, and because we are getting better and better at avoiding irrelevant ads. But this second point goes against the whole notion of targeted advertising – why would we skip past something that is relevant to us? It must either be that it’s not relevant at the time of browsing (i.e. we’re there for the content, not the ads) or that it’s too relevant (see above). Adblockers exist to combat the popups one receives when browsing shifty websites that Google refuses to admit to its publishing network (e.g. illegal streaming services, porn websites). When they also block targeted advertisements, they really throw the baby out with the bathwater.

As we switch to mobile and apps, the space we are eyeballing has no room for the traditional shapes of ads that have contained targeted advertising (banners, boxes and skyscrapers, both text and image-based). If these ads are declining in value, it is because no one is looking at them anymore. That’s why we’ve seen the rise of native advertising, of which BuzzFeed is the prime example. These websites offset the decline in value of targeted advertising by, instead, including posts that, despite having the appearance of original stories, are actually just one big ad (and have attracted criticism in the process).

Google’s bread and butter is its revenue from targeted advertising. But Google isn’t just Google anymore. It’s a conglomerate of futuristic tech companies, from self-driving cars to balloon-based Internet connectivity. The long-haul strategy of this sprawling acquire-first-ask-questions-later company, I suspect, is to preempt the Schumpeterian Gale that sweeps the rug from under data-based advertising, by grabbing a foothold in the technologies that have the potential to be the technological behemoths of the future. In this sense, the next Google is Google.

What’s next for our data?
Data-collectors will be milking targeted advertising for years to come. But there will (/has) come a time where collecting more data for the purpose of advertising represents a fruitless effort, either because we don’t value creepily relevant advertisements or because we just ignore ads altogether and make up our own minds about where we want to browse. The concern is what our data will be used for next. Improvement of app usability? Great. Anything that improves marketplace information is surely a bonus (though may involve a transfer from customers to firms, as in price discrimination). What I find much more harrowing is the alignment of corporate and government interests regarding data collection, and the potential scope for sharing between them. As data-collectors look beyond advertising as a means of monetizing the data they collect so easily and frequently, we must be alert to the fact that it could end up in more dangerous hands.

Data Google Online Advertising Privacy Two-Sided Markets