Privacy, Real-Time Bidding and the Uncertain Market Size for Display Advertising
While I have posted estimates of the size of the display advertising market, those numbers are nothing more than a shot in the dark. The size of the market (like any market) is sensitive to both price and quantity. Moving forward, price and quantity of display advertising are variable to display advertising effectiveness. If it becomes more effective, prices for a display ad can go up multiple fold. Higher effectiveness also will draw in more advertising dollars from offline mediums.
A recent Economist article explores “real-time bidding,” a new model for selling display advertising. The article describes the response to a slump in display ad prices due to low click-through rates and general ineffectiveness in reaching the targeted audience for the ad. For example, an advertiser, seeking to reach young men for its product, buys ads on ESPN, but also ends up paying for the women who visit ESPN, not the advertiser’s targeted audience in this example. With real-time bidding, in the span of a heartbeat, an auction for the display ad takes place when a set of eyeballs visits a page. The auction leverages information collected about the visitor, utilizing “digital traces” about where the users has visited before. This allows the middlemen auctioneers, who sell 60-80% of the ad inventory on a content site, to aim ads at more relevant eyeballs, using better user profiles, to reach better prospects such as those who have previously visited the advertiser. So, as an example, Zappos could pay to serve an ad to a reader of the New York Times, who has visited the Zappos website in the past, to tempt that user to go on a shoe-buying detour.
There are two things going on here: (1) the ability to buy ads on an impression by impression basis and (2) the ability to make that decision based on information about the user.
Google says “real-time bidding” also known as “retargeting” can raise click-through rates by a factor of five or ten. There are a variety of effects. The value and consequently the price of a display ad goes up as it delivers more value by bringing more potential customers. It potentially disrupts the value of ads on certain publishers (Facebook?) to the extent that one can reach the same customers by purchasing inventory on a cheaper publisher. It also potentially moves more value to the middlemen auctioneer to the extent that the middlemen can build up better profiles about internet users.
This is a description from Google on the benefits of real-time bidding from a blog posting, announcing the acquisition of Invite Media last year:
“We’re big believers in the benefits and future of this type of display ad buying. But we’re all just at the beginning. It can, and will, be much bigger than it is today, which will benefit the entire ecosystem – advertisers and agencies who can run more effective campaigns, demand-side platforms who offer real-time bidding services to advertisers and agencies, publishers who will get higher prices and more competition for their ad space (while controlling what space they make available, and to whom), and users who will see more useful ads that load faster.”
All of these effects would seemingly increase the size of the display advertising market by orders of magnitude by increasing the value of a display ad and by drawing more advertising dollars in as internet advertising delivers bigger value. It will also redistribute those ad dollars in different way.
But, the big unknown is privacy law. Depending on where the FTC and legislative bodies end up, the ability for the intermediaries to collect and use information about users (particularly to the extent even anonymous profiles can be de-anonymized) will determine how effective the targeting will become, and how big the market becomes. Ultimately, there is a tradeoff between privacy and the size of the display advertising market.