Red Bull’s Eye in the (Cyberspace) Sky

AT&T recently announced they will provide full time Internet-of-Things (IoT) monitoring of up to one million Red Bull coolers in retail locations worldwide.  If successful, this will be a powerful example of what Big Data can do for retail brands.

It reminded me of Skybox Imaging’s remote sensing of the Ras Tanura oil storage facility in Saudi Arabia a few years ago. Skybox engineers and data scientists had determined they could determine the changes in stored oil at the depot through satellite imagery by calculating the length of the shadow created from the rise and fall of each tank’s flexible roof.  Imagine having that data and correlating it to near-term future prices of oil!

At that time, oil consumption was outpacing production.  As oil prices move, so moves the price of retail gasoline about six weeks later.  While some might characterize that kind of advanced notice on future demand priceless, others put a true number on the value:  Google purchased SkyBox Imaging for $500MIL a few months later.

Red Bull coolers are scattered across the world – yet both technologies are intended to (at their core) monitor varying volumes of a liquid.  Rather than one satellite overhead, AT&T is tying together millions of terrestrial sensors that, unlike those oil tanks, are not concentrated in a handful of permanent locations.

Big Data Opportunity!

AT&T could soon show Red Bull where their best performing coolers are located.  Now, you might think simple retail sales would do that – those with the highest sales are the best performing, right?  But why are those locations selling better, and where could other coolers be better positioned?  How could lesser performing cooler locations be improved without the expense of relocating them?

AT&T has enormous data sets on user and neighborhood demographics from their cellular business.  By combining those data layers with information from the refrigerator sensors, an enormous number of possibilities will open up.  Red Bull can use localized data to figure out the best promotions for specific areas.  They can determine the best advertising venues to reach individual markets.

For instance, to reach young women 18-24 in Cleveland it might be better to advertise via Twitter.  To reach teenage male athletes in Poughkeepsie it could be better to use sports podcasts.  Retirees in Florida might respond better to satellite radio ads, while busy moms in California will line up after a specific Facebook ad is used. 

Red Bull won’t have specific purchasing data – the individual retailers jealously guard that information.  But Red Bull will be able to cost effectively measure the efficacy of their marketing campaigns at a very local level all over the world!

Instead of a technology stack, what AT&T is offering is a data stack.  Skybox Imagins’s satellites had a variety of data technologies stacked together, from antennas to transmitters to high resolution cameras.  AT&T’s project is similar in concept but all in the digital domain – a stack of information layers that must be analyzed in very specific ways.

What other data layers would help these analyses?  How about holiday schedules?  Weather activity around each location.  Even gasoline prices could be a data layer, either encouraging or stifling traffic through gas station convenience stores.  Red Bull will also want time of day data – are they competing with Starbucks in the morning and don’t know it – as well as traffic patterns around these coolers.  This is easily and inexpensively acquired, (Waymo, I’m talking about you!).

The services provided by Skybox and AT&T provide predictive analysis on demand before the same data is publicly available (if it ever is).  How does Red Bull capitalize on this new bit of demand insight?  How does their supply chain improve (faster, cheaper, more selectively) from this data? 

Now, take it a step further.  Who might want some of Red Bulls data layers for other (non competing) analyses the company could license as a revenue source?  This further enhances the value of the data they’ve collected, improving its value and offsetting some of their costs.  In allowing others to cross-use the data there may be a number of hidden demands that can be teased out these local levels.

The impact of measuring cost effectiveness in promotional campaigns in near real time should not be underestimated.  A/B testing of social media vs radio, geolocating specific demand that is otherwise buried in the databases of retailers is a sizable advantage.  Amazon was very good at this, bringing A/B testing to online retailing. Barnes and Noble was still doing such research by hand in hard copy!  What took B&N weeks to figure out, Amazon knew in seconds. 

Richard Sherlund, the managing director of Software Investment Banking for Barclays, addressed an important aspect of this at The Wall Street Journal’s CIO Network Conference in February earlier this year.  He told attendees that companies were collecting an enormous amount of data, but they aren’t leveraging it.

AT&T is bringing this function to Red Bull, and the results will be worth watching.  What remains to be seen is how they will use it.  Can will Red Bull use this data to react to changing conditions faster than their competitors?