E-Commerce

“See the moments of decision”

Real time in e-commerce: How human understanding can be brought to online commerce as explained by Hendryk Hosemann, CEO of the startup Brytes.
Author: Heinz-Jürgen Köhler
Photos: Brytes


Brytes was founded by Hendryk Hosemann and four co­founders in Dortmund in 2017. The start­up is currently at the end of its pilot phase and plans to enter the market in 2019. Brytes currently has 11 employees, including software specialists, behavioral economists and data scientists.

What service does Brytes offer and what customers do you attract? 

We offer online stores what is called psychographic online personalization. There are three steps: 1) We identify the visitor to an online store and track how they navigate the page; 2) We attempt to gain an understanding of the user: “What kind of person is this at this moment? What are they looking for? What do they want?” 3) We use this understanding and respond in real time by making use of, say, messages. Overall, it is a matter of better understanding the visitor in that specific situation.
We make websites more empathetic and bring human understanding to online communication. It is suitable for any company that places value on the customer experience. This definitely pertains not only to high­end consumer goods, but also mass markets. It also has a purpose in grocery shopping, since that is about habit and a feeling of familiarity.

What parameters do you use to analyze a user’s current mental state? 

We use over 200 parameters to understand what drives the visitor’s decisions and purchases. These include scrolling and mouse movements – Are they moving in one direction? How fast are they? In what order are they looking at offers? Where are they pausing? Are they going from the overview page or from other recommendations? We talk of a decision journey, which is basically a broken­down customer journey with which you can see the moments of decision.

To what extent do you access other information the user leaves behind online? 

Analyzing user behavior from earlier visits to our client’s online store or on the Internet in general is irrelevant to us. Our system reacts in real time to a visitor’s current behavior. Even if they were to come repeatedly or were to have just come from a competitor’s store, the system would not respond differently. And even if we could technically do that, we do not save the customer’s information.

How do you define real time and what do you do to accom plish it?

When we talk about real time, we are talking about milliseconds. Faster is not technically possible: We load as quickly as the page loads. Delay­free website performance is important to our customers and we adapt ourselves to that. Without a delay subjectively perceived by the user, we analyze their behavior and respond to it. This, of course, makes demands on infrastructure and server capacity. We also keep a content delivery network ready for our clients to help cushion peak loads.