In one of our earlier entries we talked about the importance of fill rate in Mobile Advertising (more on that here)
Over the last few months, Adiquity has been looking beyond fill rate and exploring the possibilities of Real Time Bidding (RTB). RTB is programmatic buying of ad impressions in real time, based on sophisticated targeting models. According to eMarketer, RTB display market is expected to grow 72% in 2013 and account to 19% of all digital display ads sold.
So far, media buying in mobile has been dominated by traditional Ad Networks. The classic Ad Network approach is to sell publisher inventory to advertisers’ for the best price. In most cases, best would mean bulk inventory access to the advertiser for cheaper price. Ad Networks are judged on basis of higher CPCs, whereas in reality, their inventory is being sold at much cheaper rate than their actual worth. Of course one can find exceptions to this statement where certain ‘premium’ publishers are able to claim much higher rates and minimum delivery guarantees from the networks. But lesser mortals will have to settle for lower prices. Similarly from an advertisers’ point of view, quality publisher traffic means clean (with no malicious content) traffic with high engagement from the users (Click Through Rate or Conversions).
Let’s break down this model and take a deeper look. High engagement is always good news for the advertiser and they readily pay more for such traffic. But that does not mean that the entire inventory from a category or a publisher gives the same engagement for a particular Ad.Simply put, not all users will be looking to buy a particular product/service which is advertised. Every user impression coming through a site will have its own ‘personality’. Wouldn’t it be better if the advertiser can target only those user impressions with the ‘personality’ which converts the most, and not waste impressions on the rest of the traffic? The term personality here refers to different dynamic profiles of user impressions (comprising of past behavior, location, purchase intent etc) which are modeled by sophisticated data analysis, and not to be associated with any data collection technique which violates user privacy laws.
Unfortunately typical ad sales and publisher monetization happens on a category or channel level where the publisher’s traffic as a whole is bucketed and sold at a predetermined price. This approach may not be the best considering the fact that advertisers are clearly willing to pay more for certain profiles of inventory which fits their bill. Publishers are also losing out since different traffic profiles have different appeal to different advertisers.
This is where ad exchanges like Adiquity breaks the mould and offers a win-win solution for everyone – the advertiser, the publisher and also the user. An exchange operates as a ‘market place’ where buyer focused platforms – known as Demand Side Platforms (DSPs) bids for traffic which is sold by publishers directly and/or through Supply Side Platforms – SSPs (aggregators of publishers) through the exchange. Exchanges with Real Time Bidding (RTB) capability facilitates automated media buying for DSPs who use real-time, rich data to purchase ad inventory.
An ad exchange operates pretty much similar to a stock exchange where shares of different companies are bought and sold. In this case, the ‘shares’ are user impressions i.e the ad requests from the publishers, and the buyers are the buy side platforms who can be traditional ad networks, agencies and DSPs. This technology driven approach enables an RTB platform to auction EVERY single ad impression to the highest bidder, thereby giving more returns to the publisher than what he would realize by working with agencies or ad networks.
DSPs are built around sophisticated algorithms which can practically evaluate every impression, calculate its value for the advertiser and then buy the matching ad impression for that value. In essence RTB enhances the value of the publisher’s inventory by its ability to map each impression with an advertiser who is seeing the best conversion for that profile of traffic, and then selling it to him for a premium price.
Let’s consider a typical gaming application/s with a million downloads. Each user impression has a different value to different advertiser depending upon the device, location, and certain demographic data which is willingly shared. Now imagine a local Chinese restaurant being able to target ads (say discount coupons) to users during lunch time who are known to prefer Chinese food. It would be better for the restaurant chain to target only those set of user impressions which are relevant to them rather than targeting the entire app’s user base. These kind of advanced targeting and profiling can give insights into the publisher’s user base on their location, demographic details, purchase intent etc. Imagine publisher inventory being sold on a level where the exchange can tell the bidders that a particular site has a new user/visitor who is a 25 year old male from Brigade Road, Bangalore, India at lunch time with preference for Chinese Food. How much are you willing to bid for this user impression?
Programmatic Buying models learn and build different profiles of user impressions which are relevant to different advertisers, and then buy those impressions at an optimum price. This is a classic win-win scenario – the advertiser paying more for the perfect audience, publishers getting higher eCPMs and users getting the promotions or ads they will engage with.
As recent developments show, there is a paradigm shift towards programmatic buying in digital and mobile (more on that here) with marketers and agencies now open to look past traditional media buying. We believe the ultimate success of RTB relies on sophisticated data modelling and targeting capabilities which can shoot up user engagements and bring down cost of acquisition.