The Rise of Real-Time Analytics

Yes, that’s it, another Three Letter Acronym (or TLA for short) – Real-Time Analytics. It’s been proven that it’s beneficial to bid in real-time, creating an auction style marketplace for ad inventory. But why not analyze and react to the analytics variables received, in real-time as well? Mobile users have an array of meta data associated with location, plus time, sentiment, etc. If I were an advertiser, I’d be most interested in valuing my impressions appropriately, based on things like audience characteristics, location and behavior. While, today, most RTB platforms do an excellent job of bidding on impressions based off of pre-defined parameters, I’m excited for the day when the programmatic space, especially in mobile, is using big data, not unlike PlaceIQ’s, to analyze buying decisions combining intelligence with auction to create smart auctions. We often hear about analytics driving real-time decisions, but truly dig deep and you’ll find that is just not the case. Usually, data’s used in a predictive nature. Moving toward real-time could be potentially game changing and add value to inventory, while increasing budgets toward programmatic sources.

Imagine placing ads for a business traveler target for an airline and you want to be able to optimize the campaign toward airports. Fortunately, real time analysis of the actual engagement for the flight deal the brand is pushing, shows that the engagement happens more at white collar offices, late in the week, during the afternoon when travel’s being booked. Conversely, exposure may behigh in the airport, but the acrobatics necessary to get through checkout, head to the plane, pull out tickets (paper or digital) and scurry to write that last email before the closing of the cabin door and an overzealous flight attendant thwart your efforts, are not conducive to click-through. Then again, maybe the goal is branding and the work moments represent far fewer impressions, so the low click-through is of little concern. The only way to optimize toward audience elements on the fly is to integrate data and associate the proper characteristics to algorithmic scrutiny, creating Real-Time Analytics (RTA).

RTA could also be used to inform creative elements, like completing an A/B split for two different types of “Click Now” buttons or some other call to action. Instead of stopping a campaign to traffick the creative that works the best, an ad server could just programmatically optimize toward it with real time analytics. Maybe some markets or audience types click on the less desirable creative option, but the analysis says to keep it running in certain markets or toward specific audience types. This is an efficiency not yet available and could have real impact on the value of campaign dollars, by eliminating waste and increasing the return on creative development necessary to design multiple creatives.
In this short post, it’s tough to game plan every scenario that RTA can impact, but it’s fun to think about how programmatic buyers could add big data analysis as an input into the overall campaign strategy when buying programmatically. Imagine having the smartest inventory, at low auction prices, at scale. This could be the newest paradigm. What would you want to measure with real-time campaign analytics?