Our industry is changing very rapidly. For those of us that have been involved with online advertising for a decade, the recent pace of advancement has been truly remarkable. And one of the changes that’s having a huge impact across the entire industry is occurring within traditional exchanges and publisher aggregators. Although traditional exchanges have been providing lots of value for years, they are rapidly transforming and enabling advertisers and agencies to find their audience in real-time anywhere on the web, and do it in a much more cost effective way. All without impacting brand safety. The change I’m talking about is real-time bidding (RTB).
Exchanges have been playing an important role in digital advertising for awhile, and Turn has been working with them on behalf of our clients for a few years. They gave us great reach and 2nd price auctions, but there were still barriers that prevented us from maximizing the value of exchange inventory to agencies and advertisers. First, it was almost impossible to optimize bids on a traditional exchange, since exchanges had very static bids on each possible buy. Keeping exchanges in sync with an external system like ours also has its challenges. Changes take some time to take effect in a traditional exchange, so if you have a campaign that needs to stop delivering for whatever reason, it may be a while before the exchange actually take the changes into account. Budget optimization was also hard to achieve across multiple exchanges, as some exchanges didn’t even have an API, and emails or phone calls had to be made to affect any change. In short, traditional exchanges have been great for reach, but some have been just too manual, too slow, and too static to give our clients all the benefits they are looking for.
Virtually every single traditional exchange and inventory aggregator (and even some large publishers) is now deploying or testing RTB. Demand side optimization platforms like Turn are integrating with their systems, either directly or through 3rd parties. With RTB, we’ve been able to use every bit of our technology in optimized bidding, in budget allocation, targeting, forecasting, etc.
Turn has already been very successful buying on the Real-Time Bidding Exchanges that have deployed. We have been measuring dramatically increased metrics (CTR, ACR, eCPA) for advertisers, and we’ve also seen increased CPMs for publishers. What is responsible for these gains? Well, we are simply able to find more perfect audiences for our advertisers, giving them the biggest bang for their buck. As for publishers, Turn is bidding higher for the impressions that really matter. These exchanges also support transparency, so brand safety is not compromised. The theory is that every impression is valuable to someone a lot more than some network average bid. Everybody wins. It’s beautiful.
The benefits are there, but so are new challenges. Traditional exchanges shielded the buyers from the complexity of bidding and scale required to support auctions. The scale here is what really makes techies like me excited. In milliseconds, a platform like ours will bid on clients’ behalf in the most optimal way possible, taking into account everything that is known about the particular bid request that is being auctioned off. In Turn’s case we are a bit different from competitors: we evaluate every bid request totally and completely. We come up with a new bid based on the parameters of the auction, as opposed to some kind of average based on just some of the information. This gives superior results to our clients, but the volume of computation that has to go on is truly massive.
We have been tuning our systems and deploying additional server capacity for a while, and I keep reminding my team that come next year, when volumes of inventory available for real-time bidding will grow, we are going to bid on all remnant traffic out there. I am talking about many billions of bid requests daily. Now that’s a lot of needles in a very large haystack that we can find for our customers.
Senior Director, Platform and Applications