Data is not the new oil, but it could keep customers loyal

Every business is after that sweet, sweet data. After all, data is the new oil. Businesses need to connect their products so they can mine data from them and then store the data in so-called data lakes in the cloud. These are actual sentences that have come out of speakers’ mouths at the conferences I attend. Or sentences written in the marketing emails that plague my inbox.

But the business case for collecting a ton of data and storing it is actually pretty limited. And most companies are on their way to finding that out. The problem is that data’s more like the new crude. It’s a potential energy/revenue source, but it also requires a lot of work and intelligence to turn it into something people can use.

And frankly, I think it’s more like sunlight in that you can get data from just about anywhere. Most data isn’t all that special. So, why should a business collect data? And what should they do with it? Jon Lindén, the CEO of Swedish startup Ekkono Solutions, offered up a pretty smart vision for how companies should think about data. And because it’s IoT, the solution involves machine learning at the edge.

Lindén thinks that any company collecting user data should build machine learning into their products that have the connectivity, and that the learning should help deliver better services over time as the product understands the environment in which it’s used or the demands the user makes on it. This can be as simple as a robotic vacuum generating a map of your home so it can avoid getting stuck under the same cabinet, or as complex as a coffee machine that learns when you need caffeine and starts brewing in anticipation of your post-lunch slump.

And this type of learning doesn’t require that a company grab and keep your personal data forever. Depending on its computing power and battery life, this type of learning could all happen locally on the device. Where the company could win, in Lindén’s opinion, is when you decide to, say, get a new robotic vacuum. Instead of losing the thousands of hours your robot has spent learning your floors, you might want to stay with the same brand and transfer the old vacuum’s knowledge to your new one.

This provides a value that could keep a customer happy with the brand as well as a bit of a moat to keep competition at bay. Although if a new vacuum came out that was much cheaper or much better a consumer would be free to dump the learning and start from scratch. Very few companies are thinking like this today. Instead, that robotic vacuum maker wants to sell its mapping data to someone else, which makes little sense for any consumer who’s rightly wondering why they spent $500 on a device that’s collecting their data so the company they bought it from can turn around and resell it. This isn’t a crazy example. Last summer, Roomba maker iRobot’s CEO set off alarm bells when he discussed sharing home mapping data created by Roombas.

Perhaps this value-focused vision comes from Lindén because he is from Sweden, which has stricter privacy rules and now, the General Data Protection Regulations. Or perhaps it’s because Ekkono, his company, sells machine learning software for connected devices. But I like it. It maps closer to the idea I have that connectivity and intelligence can offer really compelling user experiences while protecting user privacy. So much of data collection today is done without a reason other than a vague sense of “we can use this” or “someone will want this.”