As we explored in last month’s column, the history of digital media buying and selling has been dynamic and varied. While the early evolution was based primarily on the need for advertisers to find new channels they could use to reach prospective customers, the quick growth of the industry and the number of sites that advertisers had to choose from soon made the idea of direct buying and selling pretty much infeasible. Imagine for a moment that you’re a media buyer tasked with spending a $3 million online budget on behalf of a client. Even if you had a good handle on the target audience for the campaign and a list of publishers that cater to those audiences, the logistics of managing multiple direct buys would make the campaign so administratively “heavy” that it would be almost impossible to measure and optimize. As a result, ad networks sprung up to give digital advertisers more of a “one-stop shopping” option that allowed them to advertise on many sites with only a single media buy.
While the introduction of ad networks took the responsibility of direct sales off of the plates of smaller publishers, it brought with it some questions about the quality of the audience targeting and the value of the impressions that were being purchased. After all, if an advertiser can’t track where the ads are running, then they certainly can’t track if their ads are getting the exposure and placement they’re paying for. So, while ad networks made the process of media buying easier, it can be argued that they didn’t necessarily make it more effective.
Part of the motivation behind this push for increased transparency and effectiveness in online display advertising was the unparalleled success of search. Search advertisers were able to more readily reach the right consumers with the right message at the right time. By comparison, display advertisers started to look like a boatload of fishermen casting lines in every direction hoping to get lucky.
One of the biggest obstacles facing online publishers is that they can’t really know how many impressions they have for sale to advertisers until they have them, since impressions are based on the real-time actions of site visitors. As a result, publishers may end up with a surplus of impressions for a certain month and need to monetize those impressions. For a number of ad networks, this amounted to a nice little side business where they could buy remnant inventory very cheaply from publishers and repackage it as targeted inventory to be resold to advertisers. Unfortunately, a number of these “targeted” buys were really just a cobbling together of remnant inventory from dozens of sites, and didn’t really take into account the needs of the advertisers or the audience.
However, publishers were only too willing to work with the ad networks, because some money for an impression is better than no money for an impression. However, the result of this practice was a steep increase in remnant inventory that quickly caused a glut, and caused inventory prices to plummet – including publisher’s premium inventory. This resulted in less-targeted campaigns being run, which further diminished the value of digital display advertising and the vicious cycle continued.
Unfortunately, campaign results, largely measured using click-through metrics, kept returning lower and lower ROIs for these campaigns, and once again the industry needed to think about ways they could help advertisers get the right message in front of the right consumer consistently.
Because publishers and advertisers were getting better at collecting and digging through site visitor data, they started to get a better understanding of the best sites, pages, content, and times on which to run ads. The result was new ad networks that were driven by audience-targeting technologies.
With more robust behavioral and audience-targeting tools at their disposal, advertisers now had the flexibility to place ads on sites that best reached target audiences at prices that were based on current demand for those impressions and not on a set CPM (define) cost. The result was that advertisers were able to get more “premium” impressions for non-premium prices and publishers were able to sell more of their former remnant impressions using a real-time bidding approach.
In 2005, this drive led to the creation of the first ad exchange. Not unlike a stock exchange, ad exchanges allow publishers, ad agencies, media buyers, and advertisers to bid on available inventory in real time. This means that instead of buying a set number of impressions from any particular publisher, an advertiser could instead focus on reaching their target audience across ad networks while competing against other advertisers for any available targeted impressions. The result for many advertisers was better ROI (define) at a lower price.
As we’ll explore next month, the rise of the ad exchanges lead to demand-side platforms, buying groups, data brokers, and other uses of technology to optimize ad delivery and campaign results.