Her academic background is equally as serious with an undergraduate degree from Yale, a Master’s degree in Engineering from Stanford, an MBA from UC-Berkeley, and is currently pursuing a law degree from the University of Santa Clara.
We sat with Joan briefly to learn about her people-oriented, collaborative leadership style that has left an immense impact within the companies she has led, not only in driving business success but in developing employees to grow and thrive.
M.R. Rangaswami: What lessons have you learned from integrating a relatively small startup like Fyusion with a multi-billion-dollar enterprise such as Cox?
Joan Wrabetz: We are very lucky to have been acquired by a company that has acquired many previous companies and is very good at the process. They have been especially careful to try to prevent big company processes from slowing us down, given our very lean organization. Cox has many brands under its umbrella and those brands are valued as independent brands. As a result, on both a process and marketing level, we have been encouraged to maintain some independence.
The learning has been heavily on our side as many of our employees have never worked for another company, let alone a big company like Cox. We have learned that it is important to be specific in asking for help or input from the bigger company because if you are vague and general, there are literally hundreds of people who might offer that help and it can be difficult and confusing to sift through all of that. Also, we are learning that in large organizations, there is less uniformity and consistency of objectives, and communication between people and teams doesn’t always happen quickly. Of course, we have these problems even in our small company, but we usually catch it more quickly.
M.R.: When implementing a novel, cutting-edge technology like computer vision within an industry with well-established processes such as automotive, what are the keys to driving adoption?
Joan: There are two keys to adoption: Firstly, implementing the technology within existing processes except in areas where the benefit of the new technology would be lost without process change. Second, extensive training implemented from the customer’s point of view.
For example, at Fyusion we enable 360-degree interactive images of cars for a wide range of applications in automotive, and these images can be captured with most common smartphones. We also extract a gallery of traditional 2D images from the 360-degree images. Initially, our customers wanted us to just replicate their existing process of capturing photos. It took some time for them to recognize that the new technology-enabled the capture of photos as a side-effect of capturing one 360 image through walking around the car.
This solution is advantageous to our customers, but it also means changing long-standing processes. Helping customers adjust to this change has involved both reframing how they think about imaging and being willing to constantly adjust our customer training programs and product workflow based on customer feedback.
It has also made us adept at focusing on the real value we provide, and wrapping into existing processes where it’s not core to our value proposition.
M.R.: What do you think will differentiate successful computer vision applications from those that fail?
Joan: Computer vision is only a part of a solution that solves a real-world problem. The applications that are successful will be those that apply the technology to address a real-world problem in a disruptive way to lower costs or achieve better outcomes. For example, if we were simply accurately reproducing a car in the digital world, we would be improving the customer’s ability to buy and sell cars online. But, we do more than that—we add information that the user would not easily be able to get if they are just physically looking at a car, information that we have “learned” through computer vision technology. The digital experience is actually richer than the physical experience and outcomes are not just possible online, but they are possible faster, with better information and with more trust. It is this type of change that differentiates successful applications from those that fail.