Quality Over Quantity is Key to Mining Profits From Data

You shouldn’t be surprised if your car company knows that you’ve gained a little weight lately and then sends you an ad for Weight Watchers or reminds you of a donut shop on your route to work.

Choosing the diet or the donut is up to you.  Car makers can make money either way.  

As Big Tech has shown, there’s big money to be mined out of Big Data.  Google, for instance, hoovers up an estimated $180 billion a year by slicing, dicing, dissecting, combining and analyzing personal information we give up about ourselves as we blithely surf the Internet.  Online data led to an estimated $861 billion in eCommerce sales in 2020, a pandemic-motivated 21.3 percent increase from the previous year.    

Much of that data and the resulting economic activity is generated by the cars we drive.  With 20 times the computing power of the laptop you are likely using to read this, new models are outfitted with sensors, cameras, accelerometers and navigational telemetry, all designed to improve safety and driving efficiency.  And yes, car makers include weight-tracking sensors for the front seats, primarily to regulate the airbag systems. 

But these systems also generate mountains of data.  

The chip maker Intel estimates that a new hybrid sends 25 gigabytes of data an hour back to car makers.  The AAA Foundation estimates the average American drives 220 miles a week.  You can do the math. Add even more from cell phones connected to the car.

For car manufacturers, it adds up to terabytes of data that can translate into new revenue streams important in a competitive industry with declining global revenue facing big future investment costs. The consulting company McKinsey estimates that market to be worth $750 billion, more than three times the capitalization of Toyota.  Some car makers think that as much as 30 percent of future net profit could be generated from the data a car generates, making the database almost worth more than the car.

But just like panning gold out of a stream, capturing revenue from this gusher of data faces obstacles. There is Congressional concern over Big Tech’s business model.  Privacy laws in California and Europe prevent manufacturers from capturing personal information without permission.  To get that permission, manufacturers are leaning on old tried and true retail marketing methods that have worked since the days when motorists were rewarded with sets of drinking glasses for buying all their gas at the same station.   Facebook knew it was still true decades later; people give up their privacy for free stuff.

Motorists can choose from an increasing number of freebies, from discounts in the service department to navigation services and other conveniences.  Some manufacturers even upsell “freemium” options, such as remote ignition, bypassing dealers to increase their direct contact with car buyers.  

As drivers get freebies and discounts, manufacturers get permission to collect data and share it with dealers and other affiliates, such as their finance and loan operations. But it takes more than collecting data to make money from it.

In a McKinsey survey of 500 senior executives, 85 percent said it was important that IT departments be directed from the C-suite and that investing in analytical software and staff was a top priority to wring the most value out of data.  According to the survey, high performing companies are three times more likely to have CEO’s direct their analytic projects, underscoring their profit potential.

Internally, data can help automakers save money, such as by identifying needed changes in prototypes to cut production costs otherwise known as “reductive design”.  But the external retail market for vehicle data is vast and unlimited, from banks and other financial institutions, repair shops and everyday retail vendors.

The key to unlocking this potential is improving the quality of data.  Modern and agile  infrastructure can organize data across a large variety of sources and formats.  It will always outperform more rigid legacy systems that can’t keep up with the volume of data or that can’t perform the sophisticated analytical work needed to produce high quality databases.  

Privacy laws preclude identifying individual drivers personally, but artificial intelligence can mix and match data sets to segment individuals anonymously in cohorts, based on a wide range of socio-demographic factors, such as age, sex, sales history, geography or common interests.

The results are audiences targeted with the precision of a surgeon’s scalpel in databases that command premium advertising rates and high sale prices.  This segmentation also opens up opportunities for customized ads or messages that cater to specialized interests that produce deeper customer interest that advertisers reward with top dollar investments.

The market for high quality data is far broader than the auto industry.  It includes almost every segment of the economy from sports, trade groups, entertainment, dining, banking, and government agencies to retail vendors.  Banks, for example, might see recent service department charges as a lead for a new car loan.  A trade group might identify potential new members in a strong database based on their industry. And yes, a national chain might pay to know who drives past their donut shops on their everyday commute.   

However, collecting high quality data from the vehicles of today and tomorrow is much easier said than done. Modern connected cars can generate more than 25 gigs of data per hour whereas an autonomous car can generate up to 3600 gigs per hour. Multiply that across an OEM fleet of hundreds of thousands or millions of cars and that’s a staggering amount of data to make sense of! For years the industry norm has been to log all data using a ‘blackbox’ approach. However with automotive data management platform tools such as those offered by Sibros, automakers can specify highly granular data to be collected using very specific events, conditions and timeframes (down to milliseconds) they set. Taking this approach can help keep cellular and data storage costs at bay while arming the automaker with only relevant, high quality data that is actionable and valuable. 

The market for data is only limited by its quality.  Event and rules based data collection along with a definitive data management strategy play a vital role in producing the highest quality data that others are willing to pay top dollar for and that the auto manufacturers can turn into premium profits.