Here at Guidewire Live, we like to emphasize how easy we make it for customers to join: You install a simple extract program, and we take care of the rest. And when you log in to our apps, we want you to see meaning, not “big data”. Who would wish it otherwise? Yet the peculiar side effect of our approach is that it hides from you all the work we do behind the scenes to ensure that you can trust the data underneath the apps.
In future posts I intend to pull back the curtain and demystify some of that work. And we know very well that no matter what explanations we give, we ultimately have to earn your trust over the long haul. If you are creeping along the freeway every day while your traffic app mocks you with a bright green line, you are going to quit using it! But I would like to start with a more personal story: why I care about your data as much as you do.
My name is Ryan Park Grant, and I'm the guy who promises that you can trust the numbers you see on Guidewire Live. That’s a big (and humbling) responsibility, and I don’t mean to imply that I do it all myself. We now have more than a dozen people applying their considerable technical expertise to that mission. But I was here when the benchmarking “program” consisted of Eugene handing me four simple customer extract files and saying, “See what you can find.” (Then standing over my shoulder and asking, “Is it done yet?”) And when we grew our development team to take that original program to the next level, I was the one explaining the intent behind each number that we show—and checking that we got it right.
I recognize the audacity of my claim to care as much as you do about something so central to your company’s success, but I sincerely mean it. By temperament, I am an inquisitive person who likes to analyze and understand how things work, to find stories within a seeming jumble numbers. By training, I am an economist and statistician with a master’s degree from the London School of Economics. And by profession, I have spent most of my career in a field that is most unforgiving of mistakes—expert testimony in lawsuits. In one of the worst moments of my professional life, I squirmed in my seat as my colleague on the witness stand responded inaccurately to cross-examination on one minor detail that I had not communicated to him beforehand. When the opposing attorney cleverly highlighted the error for the arbitrators, my colleague’s credibility—and with it our client’s hopes for a multimillion dollar judgment—went down in flames. When you’re playing with real money, there is no such thing as “close enough.”
That was a vivid negative experience, but I have many more positive memories of uncovering fascinating and valuable insights from data as diverse as Labor Department surveys, power plant operations, and franchise sales. Thus when Eugene shared his benchmarking brainchild with me, the most compelling aspect was the chance to break new ground—to use a fantastic data source to gain insight into issues that had never been explored across multiple companies with such granularity and accuracy. I want to know how new regulations influence indemnity and subrogation. I want to know if we can predict litigation with any accuracy. I want to know how industry performance is changing over time—and how to visualize it intuitively. I expect you want to know these things too, and that my curiosity is not idle but will serve you and help your business to thrive.
However, as I quickly learned in our first round of benchmarking, even fantastic data is not free of problems. And ever since, I have been working with our customers—no doubt including many of you who are reading this post—to understand what those problems are and how to correct them: how to recognize corrupt converted data, how to standardize coverages or recovery types for comparison across insurers, how to assign claim-level expenses across multiple exposures, and so on. Because for you and me to get the answers we want, I need to solve these problems. It’s an ongoing job that will never be done, because we at Guidewire Live will always strive to improve. My point for now is not that my work is perfect, but that I do indeed care about the quality of your data. And if perchance I don’t care as much as you do… well, then that says more about your dedication than it does about mine.