Almost Famous: Pat Hanrahan of Tableau
A feature wherein All Things Digital looks at up-and-coming and innovative start-ups you should know about.
This week: We dropped by the Gates Computer Science building at Stanford University for an interview with Pat Hanrahan, professor of computer science and electrical engineering, as well as chief technology officer at Tableau, a business intelligence start-up with Ph.D level chops in data visualization.
Who: Pat Hanrahan
What: Chief Technology Officer
Why: Last Thursday, Tableau launched a public version of the data visualization product it sells to the likes of Microsoft (MSFT), eBay (EBAY) and Google (GOOG). Tableau Public is a free service aimed at journalists, bloggers and academics who want to create original, data-driven graphics similar to those from major news outlets.
Where: tableausoftware.com (Web site); @tableau (Twitter); Seattle and San Francisco (analog places)
Who else: Tableau competes directly with huge enterprise software companies like Oracle (ORCL), IBM (IBM) and SAP (SAP). Tableau Public, on the other hand, signals its entrance into a new market where the field is wide open.
Five Stats You Won’t Find in His Facebook Profile
Worst Job Ever: I’ve been pretty lucky. I’ve had mostly good jobs. I guess the worst was when I worked in a paper mill in college. I’d be on fire duty, which meant standing around with a hose and doing nothing. That said, if you go a week in a paper mill without a fire, you are doing well. All that dust accumulates and practically becomes explosive.
Geek Crush: Francis Crick, the molecular biologist. I got my Ph.D in biophysics, and he was one of the only physicists ever to be successful in biology. He also brought theory to biology at a time when it was unheard of, and I thought that was a really big thing. This was back in the late 1970s when it was basically impossible to be a theoretical biologist. I’m a big fan of the mixing of theory and practice. He kind of brought the two of those together.
Gadget of the Moment: You know, I’m a little bit of a gadget guy, but I’m more of a maker type. I like electronics, mechanics, chemistry–lots of things. My favorite recent project was building a cat wheel. It’s like a hamster wheel, but giant, four feet in diameter. I’ve got a Bengal cat. He’s very energetic.
Secret Fame: Pat has two technical Oscars for his founding work on the RenderMan software at Pixar.
Secret Shame: He can’t sing or dance to save his life.
Bio in 140 Characters
Pat grew up in Green Bay. Wisconsin made him a Ph.D chess champion. A self-taught programmer, now he’s a CS professor and entrepreneur.
The Five Questions
You say Tableau is in business intelligence, but what do you really do?
Well, Tableau’s center is really about answering questions with data. A lot of data visualization research is really about making pretty pictures, but we worked with psychologists and graphic designers to understand how people deal with visual data and process it. Let’s say you could answer a question by making a picture that shows the answer. If you want to know what the maximum selling product is, you make a picture where maximum stands out. If you want to know spatial distribution, you make a map. We create pictures that answer questions, but we do it for businesses that want to know things about their own metrics. It has been termed visual analysis–sort of doing a Q&A with data and images.
Who is using it well?
I’ve been really surprised by how many businesses use the sorts of metrics that work well with Tableau. We sell to category managers at eBay, for instance. Google uses us a lot for managing its data centers. We are not really vertical at all. Tableau is useful for anyone who has data.
A really interesting example is our relationship with Xbox. They record all the game play and then offer data through us to their game developers so that the developers can see what the actual game play experience is like. When are people dying? Are players spending time where the developers think they should? Stuff like that.
It is really everything. Some churches use us to keep track of who is donating what on Sundays. Most of our users are the Excel user; maybe they have data, but not a way to visualize it. It’s amazing to me how quantitative so many people are.
So how does Tableau Public differ from your enterprise product?
Well, the market we’re going after right now is individual content producers who might want to put data online. The New York Times (NYT) is often held up as an example of these good graphics, but an individual blogger doesn’t have a huge graphics department.
We offer the service for free, with some limits on number of views, and if the graphics take off, then maybe we’ve earned a paying customer. Also, on the free version, the data is public. It’s good for us because we get exposure, and it’s good for others because they get free access to the technology.
We aren’t immediately concerned about making money with Tableau Public. We already have a robust business selling to other businesses, so we sort of came to the freemium model backwards of most start-ups.
Can you guys really compete with the likes of IBM, SAP and Oracle?
Well, one big reason we get our customers is the whole visual analysis thing that is at the core of what we do. It’s unique to us. We’re also really well known for being easy to use and easy to deploy. A lot of times, what happens in enterprise software, you get these monolithic, giant systems that can be clunky and painful to add new features to. This can be true especially in the analysis arena.
The Dallas Cowboys are a good example. The sales manager there would go to his data guy and say, “I want to know how many jerseys I sold yesterday.” And they’d start giving all these technical answers about the data cube not being connected to the servers and so on. He was sold on us because he could plug in a complex spreadsheet, and we could tell him that answer in a very concrete way in a reasonable amount of time. It all goes back to having that Q&A with your data.
You are a professor of computer science and electrical engineering; you must have a pretty amazing early technology memory that turned you on to the sciences.
For me, it was just science in general, just being a nerd and a scientist. I remember when I bought my first chemistry set from a company now called Elemental Scientific. I remember that I was about eight or so, and most of the research I did was just so I would know what to buy. I saved up all my money and went to the store with my grandmother and came out with this giant box of retorts and flasks and all kinds of stuff. I had a great time the rest of the summer just doing reactions.
The other big thing with me and science was chess. I was the Wisconsin state chess champion in high school, and that is what taught me to really study things. I’ve always been more interested in ideas than technology I guess.