By now, marketers have realized the old model of advertising is gone for good. “Mad Men” are quickly being replaced by MathMen (and women). However, the options, tools, and tactics of “what’s now,” “what works,” and “what’s next” are as overwhelming as they are inspiring. The fragmentation from mass media to online consumption is happening; but it’s the proliferation of mobile phones that’s causing another disruption, while fueling more opportunities. Marketers are allocating significant portions of their budgets to mobile. According to Adweek, mobile advertising grew 65% in 2015 into what is now a $32 billion market. eMarketer estimates that this figure should more than double by 2019.
Almost as quickly as advertising is migrating to mobile, it’s also evolving to become more “programmatic” – which refers to the use of software to purchase digital advertising, as opposed to the traditional process that involves RFPs, human negotiations, and manual insertion orders. This is an area where our unique, real-time approach to aggregating and processing purchase data from point-of-sale machines will be an opportunity to drive programmatic ad buying in new and meaningful ways.
By combining real-time sales and inventory data from point-of-sale systems, we can programmatically promote specific messages. Here’s an example. Imagine a restaurant has too much turkey on hand and, via predictive analytics, it’s determined it won’t sell before it spoils. We could use that data (predicted sales + inventory) to programmatically control targeted ad units ($1 off turkey burgers) on mobile devices to increase sales.
Beyond mobile and programmatic advertising, there is a rapidly emerging field of “cognitive computing,” where computers are becoming even more intelligent. One of the key drivers to progress in this field is called “machine learning,” which aims to give computers the ability to learn without being explicitly programmed. This opens entirely new possibilities, where marketing becomes not just automated, but autonomous – free of human intervention. One of the key requirements to machine learning is collecting massive amounts of data that can “train” machines to think on their own. You may have heard of Facebook’s plan for the automated assistant “M”; or IBM’s Watson, which beat the human Jeopardy champion just a few years ago. Or, more recently, Google’s victory over a human champion of the game “Go,” where there are estimated to be more moves possible than there are atoms in the universe.
Reaching a milestone of 3 million SmartReceipt transactions in one day is an example of how Mobivity is amassing millions of real-world purchase transactions daily and marrying many of those transactions to consumer identification points, such as mobile phone numbers. We are well on our way to processing more than half a billion offline purchase transactions this year and have a goal of applying cognitive computing techniques to the data we’re accumulating to develop even more valuable applications.
This is just one example of Mobivity’s data science at work for the future. We imagine fascinating possibilities where we could accurately predict things like sandwich sales down to an individual store to optimize marketing and inventory processes generating millions of dollars of value for our customers.