This week’s guest blog post is an abstract from the original published in The Data Source by Anindya Datta, Founder, CEO, Chairman, Mobilewalla.
Understanding and modeling users behind digital devices are the keys to 1:1 marketing. In this era of personal digital devices (PDD), such modeling enables a consumer to be reached uniquely via his or her PDD. Consumer modeling forms also the basis of computing digital audiences, widely used in brand advertising.
Mobile represents an attractive medium for such modeling by understanding customer movements. By being co-located with the consumer at home, in the workplace, while shopping & in entertainment locations, a smartphone also knows what a customer’s favorite music and video apps are, in addition to valuable insights into more complex concepts such as voting preferences. As a result, signals sourced from a mobile device are widely used to make judgements about the owner of the device. The most widely used such signals are, "ad requests,” generated by app and web usage on smartphones and tablets.
Ad requests are not only information-rich, but are also relatively easy to interpret, given the structure imposed on them by standards bodies (such as the OpenRTB organization). Structured ad requests are most common in the programmatic framework, where exchanges/SSPs aggregate ad requests from a variety of mobile media and push them out as Bid Requests (BRQs) for consumption by DSPs and ad networks. BRQs, accounting for about 50 percent of all US mobile ad supply, represent a key source of data helpful in modeling and eventually enabling "contextual targeting" of consumers.
Context, in a textbook sense, is a combination of place (location), time and event. Products are best marketed in optimal contexts — for example, we are more likely to respond to an offer of ice-cream on a hot day at a beach than on a rainy, cold evening. Because a mobile device constantly accompanies its user and is therefore present in every “context” in which the user finds themselves, it becomes a powerful way of reaching him contextually. Thus, the ice-cream ad can be delivered right to a consumer’s mobile at the most optimal location.
State of the art in consumer addressability on mobile includes techniques that target consumers along all three contextual dimensions — a consumer can be served an ad (subject to his/her availability of course) at a specific time (say between 8-8:30 PM), at a specific location/place (say at the AT&T Stadium in Dallas) during a specific event (say, a Country Music concert by Shania Twain). In particular, much innovation have occurred around the location dimension, enabling high precision targeting at arbitrary locations. However, the same is not true for time, as we will argue in this article.
Time is a key dimension — serving the ice-cream ad at noon on a summer day is more likely to be effective than on a winter night. However, time-based ad delivery is largely limited to ensuring that a consumer is reached at a specific time. This is easily achieved in programmatic scenarios by acquiring BRQs from the consumer’s device that originated within the desired interval.