The sudden evolution of the phone from a single-function device for making calls to a multifunctional pocket computer is one that’s already changed how we conduct business and live our lives a great deal. Now more than ever before, however, our behavior is also influencing this evolution to create products and services that respond to our unique needs. As the computing power of our handheld windows to the world continues to increase, so to will our ability to instruct them to do what we need them to do. This is where business leaders like Predict.io co-founder Silvan Rath see new opportunities for phones to streamline an array of daily tasks for consumers and provide better insights for businesses. Here are some of his thoughts about phone sensors, the retail market, and what lies ahead in an era of ever more personalized computers.
What was your motivation to start Predict.io? Have things unfolded as you expected?
The market has a way of breaking things that don’t change. So we did. We initially started with a SaaS model that helps cities and transport vendors to understand mobility behavior. But then retailers, QSRs and other verticals started picking up on the value that sits in location data. Hence, we started offering what they need. What they needed was an integration-free means to target customers of their competition.
Explain the advantages of ‘hyperlocal targeting’ when it comes to retail.
Complete our SAP x Data Natives CDO Club survey now, and help us to help you
Any business with brick & mortar stores desperately needs to understand the offline behavior of consumers. We enable clients to not only get information about what users do when they are online, but about all of their favorite physical stores. There is tremendous value in being able to target your competitor’s clientele.
What is one disadvantage?
Our technology is the adtech equivalent of a laser beam. You can shine very brightly into extremely dense segments. On the downside, this doesn’t give you the reach of a torch – which would be the adtech equivalent of classical segmentation. Currently, the entire market is looking to find methods that convert as well as find methods for retargeting. Targeting a competitor’s visitors can deliver the desired results for many smart retailers.
Where do you see areas of untapped potential when it comes to utilizing smartphones as tools?
The device manufacturers are currently investing heavily in on-device hardware. One example includes specialized chipsets that can run Machine Learning processes very efficiently. This is the gateway to the very near future of hyper-personalizationion. We will see a wave of companies going extinct because they misssed the chance to invest in personalization. I was born in 1980. And even I have no tolerance for untargeted ads – be it online or by post. Imagine the latest generation growing up with voice assistants which will understand all your preferences.
What do you ultimately hope to provide for an end user when you develop these new sensor-based tools?
There are many benefits of personalization beyond targeted advertising. Why do I need to buy a train ticket at the train station? Your phone already knows where you went. Why would I need to build a list of bookmarks of my favorite restaurants? My phone already knows where I regularly go. Why do I need to switch on the light every time I enter a room? The room knows I am there. The list goes on and on…
How does work on hospitality tools differ from developing a tool for banking and insurance?
The underlying challenge is the same. You need to deeply understand your customer or prospect. You need to be non-invasive but helpful. The data points that matter to each industry, however, differ widely. A restaurant would be very interested in what other places you frequent, whereas a bank would like to know if your credit card and your phone are located in the same place in order to help prevent fraud. Throughout all of these processes, we also find it important to regulate what we monitor. We are not pulling up insurance data for risk analysis. We also don’t pursue details like sexual orientation, race and other deeply personal data that could be inferred from location information.We feel it is not fair game to use location data for such purposes. The new European GDPR does provide boundaries in this area, but it doesn’t regulate everything. That’s why we also self-regulate.
As more and more devices and digital tools do and decide things for us automatically, what types of actions do you believe people will want to continue to do, even if they theoretically could be automated?
Let’s talk about cars for a second. I don’t want to drive in traffic. But I do want to drive on a scenic road. Also, in home automation, current types automation are far from fool proof. Personally, I think there should always be a choice. The same way I can choose to cook or go out for dinner. Automation, after all, is a service.