Jock Mackinlay : Interview ”The Early Days and What’s Ahead”

THE EARLY DAYS

 Lee Feinberg (LEE)
I'm here with Jock Mackinlay, Tableau's first Technical Fellow, and we're going to explore what it was like  in the early days of the company.  Since Tableau's going through a major change right now, I figured it would be good to take a step back.  Jock, thanks so much for taking some time to chat with me.

 Jock Mackinlay (JOCK)
Delighted to be here, Lee. 

LEE
Let's just jump right in.  When you joined Tableau in 2004, how many people were at the company?

JOCK
I'm employee number six. 

LEE
Wow, what was it like back then?  What were some of the things you were talking about, how to even plan out the product at such an early stage?

JOCK
We had a really fabulous technology that Chris Stolte had brought out from Stanford, with Christian Chabot and Pat Hanrahan.  The first year before I arrived, they had done a core of work on getting it ready for productionWe were focused on the individual and self-service analytics. 

The first project that I did, I needed to learn what it took to do production coding.  I sat at the elbow of Chris Stolte, and we added a hierarchical filter widget to the technology, filtering being one of the most important things you need to do when you're doing self-service analytics.

LEE
So, right from the get-go, we had the famous filter function in the software.

JOCK
Yeah.

THE EARLY PRODUCT

LEE
Thinking back, I know it's been a long time, is there a feature that comes to mind that you wish, had made it into the product way back then, and why didn't it?

JOCK
It's very difficult to focus on any individual particular feature because it's been such a long time, 15 years.  But the reality of it is, there's a bunch of things that we haven't gotten to yet because, whether the team is six people, or Tableau Development is now about a thousand people, those are limited resources.  And, so, you can build up a long list of things, and then what you need to do is try and prioritize the top ones and work on those.  So, there were things on the priority list that we haven't gotten to yet. 

I was just thinking about it, there's things in filtering we still need to do because we've been doing other really interesting things.  I'll give you an example of one that's probably fairly far in our distant future. I came from the Xerox Palo Alto Research Center and had done research work with a great natural language processing team.  I believe that someday, Tableau will do significant work in the area of text analytics.

LEE
It's amazing to hear that.  So, you're employee number six and you guys are starting to put the product together  Did you ever dream that Tableau would become what it has today and have such an impact on the world and on people?

JOCK
Obviously, we had aspirations from the very beginning.  And we also knew that the basic technology was really easy to use.  But what we didn't know at the time was how many people in an organization would be interested in using the technology.  It was a risk for me personally to join this little, itty-bitty startup, but it was super exciting. 

What I know now, 15 years later looking back is, it wasn't so much about the use of the technology.  It was about what really drives people join organizations because they're interested in those organizations.  And it is the data that is relevant to the organization.

A person's interest in the organization gets them interested in the data, and then they go looking for software that's easy to use because they want to take advantage of the computational powers of the data.  And it turned out 15 years later, that is really a massive number of people that, because they're interested in their organizations, become interested in using data.

Want to be a data rockstar like Jock?!


Join our community to get the latest blogposts and know about events.

BIG MOMENTS FOR JOCK

LEE
Is there one thing or an event that happened that made you finally realize that?

JOCK
That was like an unfolding flower.  There were a significant number of events.  We're sitting here right now at the twelfth Tableau Conference.  The first Tableau Conference was a significant event.  That definitely gave a sense of it.  About 160 customers in the Edgewater Hotel in Seattle, all together in a room.  We all could fit in their biggest room. The enthusiasm was really, really, significant. 

LEE
That makes sense.  I'm sure it was very exciting for everybody at that first conference.  

We got a little bit of you in the past.  I want to jump ahead to today because you recently became Tableau's first Technical Fellow.  So, congratulations on that.

JOCK
Thank you very much.

LEE
Can you tell us what that means to you personally and also what it means to your work at Tableau?

JOCK
For me personally, it was huge.  The change to Technical Fellow was just about a year ago.  I had been a manager growing a team which had started many years before.  I started, originally, hiring the first designers at Tableau because as we grew bigger, we could get specialists; we also started an Industrial Research Group which was really fun given my background. 

But, that had grown and grown, and I had been spending most of my time being a manager.  And, then, essentially, walked through a door about a year ago where I had been for about six or seven years having 10 direct reports, a large team to worry about.  And then all of a sudden, my schedule opened up.  It was just fabulous. I'm still in the transition.  I had been pretty burned out.  So, the first part was just to decompress. 

The Industrial Research Group is still focusing three or four years out, and they moved under Andrew Beers who is our CTO.  I'm trying to focus a little closer than that because, as I said before, we have a 1000-person engineering team.  Lots of people have come to Tableau recently.  They don't have nearly as much experience with our customers as I do, and I'm trying to give them technical advice for product direction.  

We're trying to push the innovation down broadly into the team, but you have to have a basic understanding of why you're doing something.  And then, the engineering team can figure out how to do it.  So, that's what I've been doing.   And, yes, it's a big team; there's lots going on.  So, I spend my time listening to people , reading their documents, and then giving them advice.

LEE
It almost sounds like it's a hybrid of what you were doing before without the managerial aspects plus your roots in research.

JOCK
Yeah.  So, I tell stories now.  And I have more time to tell stories.  It ends up having more impact.

THE FUTURE

LEE
That's awesome.  I'm really glad you were able to get de-stressed.  So, to close this out, what are one or two of the big problems you would like to see solved over the next couple of years?

JOCK
I'm going to paint a slightly broader arc in the next couple of years.  When we first started, we were focusing on an individual person doing their self-service analytics.  This has now expanded to the entire organizations using data effectively. 

What's super interesting is, organizations tend to be very collaborative places, and there are lots of different people in the organization with different skillsets and interest areas.  Given that, and driven by our mission, our software also needs to expand and support all of those different types of people.

Among the obvious upper end of it would be data scientists who are trained in statistics, machine learning, things like that.  Interestingly enough, though, many data scientists need to collaborate with business people to be able to take their expertise and apply it to that particular organization that they're in.  Tableau can support that by making data visual.  It actually supports that collaboration between a data scientists and other business stakeholders. 

At the other end of the spectrum is people just trying to get their job done in their organization.  But data is now relevant to every single job in the organization.  So, we need to support those individuals just consuming data.   And, at any given moment, somebody could see something in the data relevant to their job that would drive their desire to do self-service analytics.  So, we need to support the entire system.  And, of course, organizations are many different types, small to large.  It's an extremely rich space.

LEE
It makes sense.  I remember in the early days working with Tableau and with the IT teams that they would say things like, "We own the data."  And I would say, "The company owns the data."  It reflects back on what you just said that it goes from very serious data science type work all the way down to everybody who's just, as you said, trying to get their job done.

JOCK
There's also, of course, companies collaborating with each other.  If they bring their data together, part of the collaboration is to integrate the data together.

LEE
I think that would be something really interesting to see happening.

JOCK
All companies are going to be moving into the cloud.  The cloud makes it easier for them to bring their data together and support their collaboration.

LEE
Definitely, collaboration is always a theme at TC, as well, so I think that's a great note to close on.  Jock, thanks so much, again, for taking time early in the morning.  I'm looking forward to Devs on Stage.

JOCK
Yes, I'm looking forward to Devs on Stage, today, as well, and my talk later today.  Thank you, Lee, for spending some time with me.

Want to be a data rockstar like Jock?!


Join our community to get the latest blogposts and know about events.

Leave a Reply 0 comments

Leave a Reply: