Despite RTB’s accelerated adoption in the past two years, it has yet to reach a point of universal benchmarking, and standards. There’s no comScore or Nielsen report for RTB activity – yet – perhaps we’ll start seeing the beginnings of that before the year is out (are you guys listening?). Most RTB industry discussions typically revolve around audience data that is driving buys, whether that be first or third party from a data provider (eXelate, BlueKai, etc). What industry insiders are now learning is that there is a growing set of data behind each auction that doesn’t get much attention, yet tells a fascinating story.
With nearly a year’s worth of auction data having been generated on our RTB platform (Casale Media officially launched it as Index Platform recently), we took a step back to study the data behind the millions of auction happening every day. In creating a real-time activity graph for Index Platform, we generated insight into what and how advertisers are actually buying through RTB. While these data points really represent the initial scratching of the surface of data yet to be analyzed, a key takeaway is that there are early signs of the RTB model quickly maturing.
Learning 1: The true value of an impression
RTB does not stand for “race to the bottom,” as some critics snidely suggest. Simply put, the data behind RTB auctions does not support this argument. RTB’s evolution in pricing draws some parallels to the evolution of the paid search model – there was once a time where you could buy “credit card” or “loan” keywords on Overture for 5 cents a click. Those were the early days, but now those same searches would cost upwards of $10.00 per click. In RTB we’re already starting to see this happen in certain key audience segments. There are significantly high bids in a variety of categories (market auto intenders being one!), some of which approach triple digit CPMs. If advertisers want to buy cheap, they won’t get value, as their competitors will be increasing the set prices on audience, environment and media that is of value.
Learning 2: Delivering on its promise
RTB promised to create competition. We commonly see marketers in the same industry bid for the same impression – e.g. two financial or automotive advertisers bidding on the same impression for the same cookie at the same time. Our conclusion is that brands and their DSP or agency partners are looking at similar data sets and their bidding algorithms are finding the same optimal users. This is proven when disparate platforms using different technology identify and intelligently bid on behalf of different clients for the same audience (or impression). First or third party data is putting these advertisers in the same neighborhood, and it’s the DSP’s job to knock on the right door. Competition is driving up the prices – provided that the buyers have the data to intelligently bid. The innovation of RTB is legitimate.
Learning 3: RTB spend activity is becoming a predictor
RTB is beginning to gain a predictive nature through its activity. We sampled billions of auctions that happened across Index Platform between Q4 2011 and Q1 2012. Spend in Q4 was led by the retail, apparel and home categories (30% of all spend), which then saw a marked decline in these same three categories in Q1 when household budgets are typically light. Conversely, in Q1, travel spending was up 33% over Q4 – Q1 is generally a time for heavy marketing spend in the travel industry. Another example we’re seeing is that in every period we observe automotive (nameplate) spend increase in RTB, auto insurance spend increases with it — the two are operating in a perfect tether. The data in RTB is quickly becoming reliable enough to make predictions in advertising trends.
RTB is maturing very quickly. This is driven by the continued learning from the hundreds of thousands of auctions happening every second. Yet, RTB is still under-developed (roughly 25% of all digital display ad spend), which means there is more to come. Imagine the possibilities for the rest of the year.