EP3 – Geospatial Data: The Rocket Fuel of GIS
In this episode
In this episode of On Point with Korem, I sat down with Dean Stoecker, Founder, former CEO and now executive chairman of the board of Alteryx. Dean founded the company in 1997 first as Spatial Re-Engineering Consultants or SRC and changed its name to Alteryx in 2010. Dean spoke about how his platform evolved and how he coined the phrase “Data is the rocket fuel of GIS, the elixir of life”. We started our conversation on the evolution of the technology and growth of geospatial as a part of analytics. Dean shared that “if we could nail geospatial and make it easy, all the other analytics as part of the analytic continuum would be ‘drop dead’ easy…”. Enjoy this discussion with one of the premier CEOs in the field of digital transformation!
Joe Francica: In this episode of On Point with Korem, I sat down with Dean Stoecker. He’s the founder, former CEO and now executive chairman of the board of Alteryx. Dean founded the company in 1997 as spatial re-engineering consultants, or SRC, and later changed the name in 2010 to Alteryx. Dean spoke about how his platform has evolved and how he coined the phrase: “data is the rocket fuel of GIS, the elixir of life”. We started our conversation about the evolution of the technology and the growth of geospatial as part of analytics. He shared that if we could nail geospatial and make it easy, all other analytics as part of the analytic continuum would be “drop dead easy”! Enjoy this conversation with one of the premier CEOs in the field of digital transformation. Stay tuned!
JF: Well Dean thanks again. I really appreciate you doing this, and I think in the interest of full disclosure, I have to say that Korem is a evaluated reseller of Alteryx and happy to be a partner, and the other caveat is, I think we’ve only known each other 30 years and that probably puts us in the category of somewhere in the Medicare range but I won’t even go that far.
Dean Stoecker: We shared that excitement over geospatial, so there you go!
JF: Well, that’s where maybe we start, because I have to ask this question because as I do presentations from time to time, I will never forget this about your statement to me about how data is the rocket fuel of GIS, and I sometimes even quote you on that instead of taking full advantage of that unique phrase!
DS: Well, so the true part of it though is you missed the last part of the statement. It’s: “data is the rocket fuel of GIS, the elixir of life”! But it’s true!
JF: So now we’re in this era of data and everything you said essentially came true. I mean, it’s less software oriented and much more data oriented it seems.
DS: I think you had to have the plumbing in place before the data became valuable. In my career, all the companies I worked for were data companies and I struggled because the ultimate value in content is when it becomes ubiquitous. In order to make it ubiquitous, you have to wrap it with elegant software that’s easy to use so that everybody can engage in it. Then you have to provide a layer of analytics on top of it, so that you can make the data dance to get bigger outcomes to drive the business. The challenge is when you’re a data company and you know instinctively that you have to have software and analytics as part of your offering, you end up investing, at least all the companies I worked for, Donnelly, Dunn and Bradstreet, and Strategic Mapping, they invested a third in data, a third in software, and third in analytics and low and behold they became mediocre at everything. That was really the impetus to starting Alteryx, is that we said we’re going to put all of our effort into software that’s data agnostic. Big, little, structured, unstructured, spatial, non-spatial, coming from every source. If you do that, then you can actually deliver on the promise of ubiquity for maximizing the value of content. I would say that we’ve accomplished that.
JF: Some of the early days, I remember going to your office in Orange, I’m getting kind of one of those early demos of the Alteryx platform, and it seemed to me, and you tell me whether you had the vision whether geospatial was going to be a central part of it, or whether it was just going to be a component of the application?
DS: For those that who are listening to the podcast who don’t remember, in the early days, the name of the company was Spatial Re-engineering and our goal was to actually fix a lot of the issues where people were analyzing data the wrong way and not applying more than geospatial into some particular use case. It wasn’t until I think 13 years later where we renamed the company Alteryx. We still have that heritage, I mean why x is related to the to the coordinate system? We just figured that geospatial is a weird science and a lot of people didn’t get it back in 1997. It was sort of secluded in the area of GIS expertise and if you didn’t use Esri or MapInfo or there’s probably some other systems that I’ve forgotten about, but if you didn’t use those systems, then nobody engaged in geospatial. I don’t think it wasn’t until maybe Google Maps came along, when they started making geospatial a little bit easier and the focus was not as much on the data or not as much on the map, as it was the data. I mean when you drive in your car and you get directions, you don’t look at the map necessarily to get your direction. You listen to the voice that tells you when to turn. We figured out if we could if we could nail geospatial and make it easy, all the other analytics, as part of the analytic continuum, would be dropped dead easy. From descriptive analytics to diagnostic analytics, to predictive modeling, even to automated machine learning, which we now have. Geospatial is one of a bunch of different data sets. I’ve always believed that it has the impetus to create extra value in any analytic outcome. Everything that happens in business happens somewhere. Think about where your crop is grown, where it’s picked where the truck is on the way to the plant, where it is on the shelf. I remember having a conversation with Jujo and in Denver at one GIS conference and we were geeking out over: “one day, everything in a grocery store would be real time economics. It would be all automated. You would walk in and you would have RFID on your cart and it would track you through the store and you would know who’s passing by based upon their loyalty card. It was a mix of content that was spatial and content that wasn’t spatial and you could adjust couponing as the switcher for coffee comes down the aisle. Then you would just walk out the store without even talking to a checker. Why? Because all the products would be RFID and your phone would tell you. This is being done today by Google.
JF: Right, the cashier-less system, right? The example, as I remember, was the bowling team would walk into the grocery store and the price of beer would go up!
DS: Well I think we talked about that too, real time economics.
JF: Going back to that issue of whether it’s as difficult today to educate people to what the value of geospatial is. In those early days, it was not too hard to get people like General Motors or Walmart to talk about what they’re doing. Now, they’re just so close-lipped about what their advantages are of using geospatial technology. I don’t know whether it’s harder to educate people or whether people just won’t talk about how much of a competitive advantage it is.
DS: I think there are some. Walmart’s always been tough to get them to speak about what they’re doing with Alteryx, for example. Then again, Alteryx is used in almost every single department at Walmart. While we started in the geospatial area of real estate analytics, network optimization, we’re now in every functional area of Walmart, from ethics and compliance to HR, to supply chain, store, the community. All the things that leverage data including geospatial. I think that when you demystify geospatial, then it becomes more approachable and if you give them tools that are easy to use, ours is a drag and drop, click and run platform that is code free for the person who may know complex VLOOKUP’s in Excel, but it’s also code friendly for those that want to write R, Python or Scala or any other environment that they want to leverage. When you demystify it, it becomes more approachable. We see people doing crazy things with Alteryx where they’re leveraging geospatial and it’s usually well beyond simple things like just geocoding practices to find out where your customers are. I can remember one of our partners in Austin, Texas, in digital marketing, built a Google AdWords optimization system for optimizing your media spend. They took the whole country into one mile grid cells, because the only thing you can place on Google is a coordinate, a radius and a dollar amount on a keyword; a little more sophisticated than that today, but ultimately, these guys figured out how to automate your digital media spend on Google and it was all geospatially driven. I think hiding geospatial making it, I think you said that Jack Dangermond called it out as “spatial is special”, I think that actually sequesters it in the GIS department. When you liberate it and you make it accessible to everybody, I think it becomes more powerful. It gets used more often and people have bigger and better outcomes.
JF: So, going back to that year when those conversations were being had about “spatial is special”, Jack certainly takes one position, I remember you and the other Jack, Jack Pellicci of Oracle at the time, kind of said: “no it’s just another data type”. Is that still true? You think that, like you said, there is some advantages that geospatial brings that other data types wouldn’t bring and maybe the other, more to the point, would you rather look at reams and reams of spreadsheets and tables or would you rather look at a few maps?
DS: Well the map is the metaphor for the conclusion. I suppose there are some use cases where the map becomes exploratory and you need to then go back and do more analytics, but I think that’s probably more public sector based, maybe oil and gas-based utilities. I’m not sure. I just think that, well in fact, I remember the one who was most vocal about it at that conference was Scott Neely, from Sun Microsystems. And Pellicci was here too, but I think spatial is another data type and I personally believe it can add value to almost any use case. I can remember a great use case around predicting which banks were going to get robbed. This was a cash pickup service, and it wasn’t just about where the bank was in the strip center, it was whether it was an in cap or a standalone, or in the middle of the center, but it was about how many minutes does it take to get to the nearest on-ramp on a freeway. Is it a one-way road or a two-way road? What is the ingress in egress having cashed in the bank and how much was almost irrelevant. It was more about the geospatial analytics of how quickly a car could get away. I think that there are use cases where geospatial is not relevant. I can find ways to insert geospatial in almost everything. I mean great use cases around medical malpractice insurance underwriting. We’ve seen crazy things about that using geospatial. I think it’s special, but I don’t think we should call it special, because I think that minimizes its use.
JF: That’s a fair point, and just for point of reference for people who are listening, we’re talking about a conference that was back in 2004 at the Wharton School, if I remember. It was the first one where I know what we tried to do was marry the people who were in the BI sector with the people who were in the geospatial sector. At that time, my observation was Peoplesoft, Siebel and everybody else who Oracle didn’t acquire over the course of the several years, they just didn’t get it at the time, and now of course every BI tool that’s out there has some element of spatial analytics.
DS: Yeah, I think it’s just part of it is evolution and I think that when customers see value in a tool like MapInfo or Esri they actually do want to democratize. They’re trying to figure out how do you get access to the 15 trillion dollars of value that’s locked up in corporate databases. If only three percent, I think that was the last number I read, was three percent of corporate data gets used at all, boy we ought to be figuring out ways to make spatial part of every use case possible. So I still think it’s a fascinating field, I still collect globes for example. I there’s just something about geography that is pretty exciting.
JF: Just to focus back on the Alteryx platform for a second, where does geospatial fit now and do you see that more capabilities need to be added, or have we kind of maxed out functionality either in the Alteryx platform or anything that you see there as far as geospatial platforms?
DS: No, I think that people have to remember that our platform is a little bit different, in that we provide, I think there’s 270 what we call building blocks, tools – some people refer to them as, and you can use them in almost any combination. So 270 factorial gives you billions of options on how you could string together tools, both tools that clean data, organize data, model data spatially, prepare or enhance data, building predictive models, and you can do almost anything. We’ve seen use cases where people take the capabilities that we have in the platform and extend it by adding new tooling, new macros. In fact, one of the most popular tools in Alteryx is macros where people automate these building blocks into a business process that allows them to reuse it and I’d say that the cost of data, to a certain extent, prohibits people from engaging in more of the use of geospatial. We’ve got good partnerships with the data vendors, but it’s still pretty pricey and I think once it becomes more and more commoditized, once we make it super easy to access through APIs, only then I think people will begin to do more things. There are probably some things we would add, but again, we’re adding new capabilities every release. We just rolled out 21.3 and it has computer vision, so we built tools, called Tensorflow, that allow you to detect what’s in an image, and those can be very helpful for use cases in retail, for example, where you want to know who’s walking down the aisles or who’s got their hand in the safe, and so, we’re adding all kinds of capabilities, but we never suggest that we’re not going to pursue more geospatial. I think the key is leveraging geospatial in more use cases.
JF: I’ve seen a couple of webinars and maybe you could explain a little bit more about analytic process automation and maybe where geospatial fits into that. I know it’s supposed to unify analytics and data science, but maybe get your more thoughtful definition.
DS: Well, there’s I think a couple of parts to it. One is just clarity in the marketplace. Analytics, between geospatial, BI and analytics, it’s been very confusing over the last decade where people have point solutions or they do one functional area of analytics. Very fragmented space. Everyone’s, especially today with machine learning and AI, there’s a zillion people in the space and nobody had defined it and we decided to define the category that we play in. Much like Workday did with human capital management or what Salesforce did with sales automation. There’s been lots of examples like this, but nobody had done it in the analytics field, and so what we’ve been seeing over the last almost 20 years, is that there’s been this sort of convergence into three sort of pillars of what’s necessary to get value out of corporate content or, for that matter, any data now that’s available in the in the cloud or on the ground. It’s the convergence of data, along with analytics, along with business process automation, so we came up with analytic process automation as the category, but most importantly around APA isn’t the technology. The most important piece of this is around humans, and if anyone’s seen some of our conferences, they’ll know that I’m not a believer in singularity, I’m actually a believer that if you amplify human intelligence, singularity will never come to be, so at the heart of APA is upskilling the workforce. Teaching them the data is an asset, showing them how you can get value out of it, and how you can take disparate data sources; structured, unstructured, spatial, non-spatial and integrate them into a cohesive analytic pipeline that drives value for businesses. Over my 24 years at Alteryx, I’ve just seen hundreds of use cases where people who had never engaged in analytics before, are now building incredible ROIs around commonplace things. So much so that I’ve had two customers over the last few years telling me they wanted to name their baby Alteryx! I said, “well, if you do, I will pay his or her way to college and one of the gentlemen in Saudi Arabia, he screamed back at me and said: “good! She’s going to Harvard!” APA is an important step, because it sort of sets the stage for what is an analytics platform and what is not an analytics platform. I would say that the GISs are not an analytics platform, although you can do geospatial analytics, very hard to do other things. Tableau as it is a descriptive analytics platform, so it doesn’t cover the entire continuum, so that’s a critical element. We went at this market to be horizontal. We said: “we don’t want to bucket ourselves into a vertical or a particular functional area, because that minimizes your addressable market” and that’s hard, because now you got to cover off on a lot of ground. You got to know every industry, you got to know every use case, and anyone who wants to go see thousands of use cases go up to community.alteryx.com and you’ll see crazy stuff that people have built. We didn’t even know that they were going to build them. We just built the capabilities and the interoperability of these building blocks that allow you to create these analytic process pipelines.
JF: I think I remember early on, and I don’t want to take too much more your time, but there was maybe some challenge with positioning of the Alteryx platform. I remember even you saying it I think early on, but now it seems of kind of found its footing in this area of data manipulation, data integration, ETL, spatial ETL. Do you see it that way too or do you see this as more than that?
DS: I think early on, people have to remember that when we started, we were a services company because we didn’t have any products yet, we built bespoke cloud solutions for companies. We’ve made a lot of money doing that, but the whole purpose of that, was to pay – because I didn’t raise money for 14 years – the whole purpose of that was to pay for the process of building components that would end up in the platform. Between 1997 and 2006, we built a bunch of engines with their own APIs, and it wasn’t until 2006 where we took all those components and put them into the Alteryx platform. We kind of knew it would take a while, because our intent was to make it a self-service platform that anybody could use and sell service hadn’t emerged. I mean, it’s a good thing we focused on geospatial. We were selling high value use cases on the platform for fifty thousand dollars a seat, because we knew where geospatial had such an important impact. Closing a big box retailer to open one and having made a mistake to close it may cost you 100 million dollars. We knew it would take some time and it wasn’t until, honestly, 2000 – see we changed the name in 2010, I started raising money in 2011. As much as I didn’t believe in raising money, we raised 163 million over three rounds in four years and then went public – but it wasn’t until 2014 where people got it. I think it was Click had come along and people said: “oh you can do mapping and Click”, but it was still it was not self-service. It was bought by IT, managed by IT, paid for by IT. Then we knew that there would be something more self-service, and that was Tableau. Tableau came along and we did a bunch of a/b testing on Tableau customers with our platform, and we’ve figured out the sweet spot in the end of 2013, 2014. We lowered the price to five thousand dollars a seat and we went from nine logos a quarter, to 250 new customers a quarter instantly.
JF: Wow, fantastic! All right, so one last question. Your quads are retired. What’s next?
DS: Well I’m on another board of a cool little SaaS company that does genome sequencing. The reason I joined them is they look just like Alteryx. It’s the same plan and expand process. It’s a SaaS service going after bioinformaticians and they’re trying to democratize genome sequencing. Covid, as horrible as this has all been, has been a just an incredible tailwind for these guys, so I’m doing some of that. I’m mentoring some CEOs who are first-time CEOs who don’t know that this is going to be a long effing journey for them. I’m also writing in my book. My book is called “Masterpiece: the emotional journey to creating anything great”, and trust me, it has the dark swamps of despair and everything in between. I’ll probably get on boards at some point after the book is out and I don’t think I’ll start a business. My wife and I are very active in using my balance sheet now to give back and we have a foundation called “I Rise” and it’s all about helping. It’s back to APA, it’s upskilling the workforce and I can remember when I was graduating high school, my parents said: “you know, balancing your checkbook is the most important thing you’re ever going to know” and today, if you don’t have a data science and analytics skill, your career is going to be fairly limited, so we’re going to give thousands of nanodegree scholarships, because I think academia is messed up. I think we’ve got to instill in kids that learning is a lifelong requirement. When I went to four years at CU, you get out and all the technology has changed already, so we’re going to just help people get up skilled to make sure that as they take care of the rest of us, that they have the skills to do so.
JF: Great, Dean! Can’t thank you enough. Been a fantastic catching up and good to see you as usual.
DS: You too, and one of these days, we’ll get together for a beer, and we’ll keep talking about what the next leg of geospatial is all about.
JF: Let’s do it, let’s do it for sure. All right thanks Dean, take care, bye-bye!
JF : Thanks again for joining us on another On Point with Korem. If you liked today’s podcast, please leave a comment in the comment box where this podcast is posted, which could be Apple Podcast, Google Podcast, Spotify or YouTube. I hope you’ll join us next time for another On Point with Korem, where we’ll get on point.