Funnel Bidding...Adnetik's Approach to Media Metrics & Pricing

I’ve been reflecting on a recent metrics presentation I saw at the IAB Mixx Canada Conference in March. I had a great discussion with Nathan Woodman, Global Managing Director of Adnetik, a Havas Digital company that is attempting to change the way we valuate online media.

I was curious after hearing his ultra logical methodology, why it’s so difficult to adopt change despite consensus that our current methodologies for media evaluation, pricing and inventory are archaic.

Nathan’s presentation outlined the short history of online media models. He walked through the initial bridge from traditional media (CPM) and eventually got to the performance-based models we are seeing now. I think the quantum leap was in jumping from a model that pays for a placement to one that pays for an individual that is in a specific consumer mind-set.

As marketers we must look to the purchase funnel to gain a deep understanding of the distance between awareness and purchase. Each business may have its own particular staging criteria for their respective sales cycle and Woodman was talking about applying a value associated to each of these stages so that the advertiser could bid accordingly.

Google search tells the automotive industry that the value of the “new car” keyword is ridiculously high due to competitive market bids but those dealers or manufacturers that have employed cross platform/media analytics know that they may be much wiser to invest in 3 exposures of the cherry red mini on a vertical site because that exposure is trending to be an indicator of further depth (yes, further than search) down the purchase funnel. All things measured (and bought) equally, the marketer should be able to reallocate funds to bid higher on the display ads.

Deep breath…”So as a marketer, I can start to allocate funds to various stages of the purchase funnel based on my internal intelligence of the likelihood of the prospects changing to customers”…Exhale.. Is that so complicated?

It has the flavour of behavioral targeting yes, but this is not invading privacy as much as it is employing common sense and technology that is available here and now to start buying according to real value. Analytics have become much broader in scope. Optimization is not just about pumping funds into keywords that are working or into display ads that are getting clicked on.

The logic is beautifully simple but the horror sets in when you try to explain the underpinnings of the system and how the values are calculated etc. I guess the answer to the original question is that change is difficult when it’s not easily explained. The more equations you show to prove validity, the more you lose the audience.

My opinion of this model is that it is fair and that it works in a network environment that can provide this volume of inventory. I think it will take a long time for publishers to jump on the band wagon unless there are some case studies produced that show equal or greater earnings potential to the respective properties. It’s interesting to think that inventory that was previously thought of as second tier, could once again be deemed premium (in a micro-transaction kind of way).

It’s all so deliciously complex. Nathan is on to something but I’m not sure the apple hit as many heads as it ought to yet…