Dominic Trigg is Managing Director Europe at Rocket Fuel. Here Trigg discusses the evolution of real-time bidding and why it is slowly changing the fundamentals of the media buying world.
The online ad marketplace has evolved from an industry dominated by indiscriminate forward contracts on bulk ad space negotiated directly through publishers, to today’s increasingly widespread exchanges. Today, real-time bidding allows parties to use highly complex, integrated strategies based on every element of each individual online ad impression to calculate the most accurate and effective bid.
In theory, the creation of a perfectly liquid market like this, allowing for complex arbitrary bidding algorithms, should have resulted in huge value creation for advertisers. In practice, the invisible hand of the market is currently putting an artificial cap on this value by using manual bidding, but this will inevitably change as more people adopt the tools available.
Stuck in a rut
Old habits die hard and we are still seeing a large proportion of exchange bidding being placed in very round numbers – indicating that manual bidding is still rife.
The reason for these round bids is that agencies often lack the means to decide how much an impression is really worth, and which ads will convert. This leaves advertisers overpaying for some inventory while missing out on other key opportunities. The difference of just a few pence per CPM can have a vast effect on agency margins or the inventory purchased for a given budget.
Participating in a real-time exchange manually is like having a Ferrari and only using it to drive to the grocery store – you can brag about it to your friends, but it’s not really adding any value. This doesn’t indicate a failure of real-time exchanges but demonstrates the massive opportunity for sophisticated technology to run rings around manual bidders. And of course, those using the leading technology gain a huge competitive advantage both in terms of agency performance and the ultimate reach of a campaign.
Certainly, the online advertising industry has all the raw materials to create a ‘perfect’ liquid market. However, it is still relatively young and many of the players haven’t yet acquired the technology needed for sophisticated valuation and buying.
Where automated bidding is being used, advertisers are often still hanging onto first generation technology platforms. These advertisers will never be able to leverage the full benefits offered by automated bidding and real-time campaign optimisation.
To bid the correct price for an ad impression, advertisers need to access more sophisticated technology and more detailed models that give the ability to take immediate action when an impression shows potential. It’s also critical to participate in all of the auctions, not just a limited few.
A new approach
What unlocks the full value of online advertising for both advertisers and publishers is an ‘intelligent arbitrageur’, with a twist. In finance, an arbitrageur adds value by creating wealth for itself and equalising prices to ‘fair’ levels; but otherwise arbitrage is a zero-sum game.
By contrast, in online advertising, value is created by matching demand (ads) with supply (ad space), which fulfils the “right ad, right user, right time” promise. By doing this well, everyone wins – publishers are more fairly rewarded, advertisers can win more ads and achieve better results for the same spend, and users see more relevant ads.
To be a great arbitrageur, an online advertising partner must not only allocate the advertiser’s budget in an optimal way, but must do so over time, over the course of an entire ad campaign.
Over delivery and under delivery are only the most obvious symptoms of a badly paced campaign. Wasting money in the wrong hours of the day or on the wrong days of the week can happen to all but the very best of these online players.
To be most effective, the arbitrageur must look at the cost of each individual impression and calculate an accurate value of how much it would be worth to show that impression right now. The science involved in doing this well is a combination of a number of components such as predictive modelling, adaptive control, and portfolio optimisation.
Predictive modelling means using all the information about an impression to calculate the probability of it contributing to performance goals. Intelligent systems can then compare the data to actual conversions and calibrate the model to continually improve predicting conversion rates. Similar technology is used in many marketing contexts from direct mail to Netflix movie recommendations.
Adaptive control is like a helicopter automatically hovering in space in the face of uncertain winds buffeting it about. An adaptive control algorithm can help an online advertiser buy the right amount of inventory over the course of a day to ensure that it doesn’t spend all its money at once or too early.
Finally, portfolio optimisation considers an array of potential investments and allocates assets optimally to achieve a desired risk/reward blend, and as the name implies, is similar to its financial analogy.
Brave new world
This kind of efficient, targeted buying can offer many benefits for advertisers but they need the right partner; one that can value that inventory in a fine-grained way instead of treating it as one large undifferentiated segment.
It may seem self-serving but, in my opinion, the solution is intelligent automated bidding and if all advertisers used it, they would be significantly better off.
Publishers will also benefit, as advertisers bid higher for the most valuable inventory. In order to take advantage of this, they need to share audience data with smart ad platforms and start selling audiences as opposed to putting different prices on different “types” of web page content (e.g. news, politics, fashion).
As the future heads this way, it will be exciting to see the second generation of technology and bidding processes take their rightful place at the centre of the online advertising stage. Even more exciting will be the win, win, win results for all parties involved.