Todd Jones: My name is Todd Jones. I’m the chief analytics officer here at WebbMason analytics. We are a professional services firm helping our clients accelerate their analytic evolution. So I think my journey started about 10 years ago. Uh, I graduated from Princeton with a degree in operations research and financial engineering. So I could have basically taken f two paths. One, I could have went into the financial space or the second path I could have taken was going into the analytics space and I, and I chose the, the analytics space. I joined a very early company called Spry. When I joined. It was about four months old and primarily started off doing a lot of DOD contracting specific to analytics and data. And we eventually built that company to a pretty nice size. We expanded past the DOD space, got into commercial, started consulting with some large, uh, pharmaceutical companies, transportation companies, and really built that company up and then sold that in 2015.
Curtis: When you fill that is Webb Mason, the company that then bought Spry?
Todd: Correct. So Spry was again, another professional services firm specializing in data and analytics. WebbMason historically has been a marketing a firm and so they specialize in all aspects of marketing. And as you can imagine, analytics is definitely a big area of focus for them and their clients. And so they brought us in and about 20% of our revenue comes from marketing related activities through WebbMason and then 80% of our revenue still comes working with it and analytic groups outside of the WebbMason portfolio.
Curtis: Interesting. Okay. So there was some crossover there, but not as much as you might expect.
Todd: Yeah, definitely some crossover without a doubt. So that was definitely beneficial. But you know, as, as I’m sure you can imagine with any acquisition, you learn a lot. And so we’re in a great spot right now, and we’re able to generate very healthy stream of business independently, but then also find those synergies with WebbMason as it relates to the marketing activities.
Curtis: Sure. That’s awesome. So when you got started at Spry, ah what, what was your role? What did, what did that look like?
Todd: Yeah, so when I got started, most of my role at that time was consulting. So I was working directly with our stakeholders who at the time were within the Department of Defense. So I split my time between Crystal City, Virginia and the Pentagon. And really what we were trying to do was help them build a solution that gave them a enterprise view across the four military groups, specifically related to human resources. So if you think about it, when we, you know, when we fought world war two, you had, you know, one division, the Marines and the navy out in the Pacific and then you had the army in Europe and they, for the most part fought separate campaigns.
And then we started to get into Iraq and Afghanistan and all of a sudden all of these individuals started to really come together. And so you might look at a city block and you have the air force there, army there, you know, navy seals in the area. And so all of these groups now have to work very closely together. And one of the things that the DOD was trying to accomplish at that time was to start to get a better view of people across the different military branches. So, for example, rather if I need a particular skillset within a particular city block, can I get that skillset from the navy? Can I get that skillset from the army? Maybe the Marine Corps has that skillset. And so they needed a very, they needed a large enterprise view so that they could very easily and quickly start to develop these blended teams. And so that was definitely a combination of technology solutions as well as analytics solutions. And so we were consulting with individuals within the Pentagon to help them build that technology solution.
Curtis: That’s really interesting. And what were the biggest challenges there? Was it more sort of the historical data cleanup? Was it the algorithms? Was it presenting the data in a, in a usable manner? What, what were the biggest things that you ran into?
Todd: Pretty much all the above. So you can imagine just the sheer complexity when trying to do this. So our project was actually started out of, of pretty large DOD project where they tried to implement one ERP system for all of the military. And you can imagine, you know, that didn’t necessarily, that was not very successful. And the primary reason in there is because to adopt some of the processes that came with this ERP system. You’re talking about, you know, Congress changing laws, right? In terms of how you pay a person in the military. There’s a law that says exactly how you pay them. And so if you have maybe a better way to do it because you have this new ERP system now, you actually have to go change the law. And so there was a lot, it was a very interesting project for me early on in my career because it really allowed me to understand the relationships and dependencies between technology and business stakeholders. And so, you know, we, we kind of worked through that. We help them build solutions that did not necessarily rely on this single ERP, but it was definitely a pretty interesting time to be working with the government, and it was definitely a pretty interesting project to be working with them on.
Curtis: Now you mentioned that this is an interesting point and helping researchers do research better. What kind of solution does that entail?
Todd: Yeah. So one of the, I would say most common challenges that we see regardless of what industry we’re working in, is simply access to data. And so you can imagine if you have a bunch of scientists and they have spreadsheets and access databases and relational databases and you know flat files, their biggest challenge was having a group, you know, in one corner doing research that a group in another corner might really benefit from. And so for them the biggest challenge was around access to data. And I think for us that’s been probably the most beneficial use case when you look at big data and cloud computing is it’s really allowed us to start to solve some of the traditional data access issues that a lot of our clients face that really have prohibited them from adopting more data-driven insights in the past because they simply did not have an easy way to bring the data together, look at it holistically, and run their algorithms or their reports on top of that data.
Curtis: Right. This concept sort of, you know, the concept of maybe metadata on top of the data, like knowledge management, being able to search and find things. Is it sort of the data lake concept? What, what are we talking about here?
Todd: Yeah, so we’ve, we’ve used different approaches in the past. So when we first started off, the technology we were using at that time was federated technologies, and it was actually based on semantics and it’s a definitely a cool concept. You know, you leave the data and the databases, but at runtime when you’re ready to do your analysis, there’s a single view across all your databases and you can federate this data. The challenge that we had is the performance was just awful, right? If you’re trying to move, you know, data across the wire in one database is sitting in California and another one sitting in Virginia and you’re trying to do this join, you can imagine, you know, the performance issues you’re going to have. And so for us, that’s when we really started to look at the big data technologies. First starting off with Hadoop and then transitioning into more of the pure cloud solutions like S3 and EMR.
We started to look at these solutions as cost effective ways to bring data together. Duplicate data, which is, you know, often a bad word when talking to IT folks could actually duplicate data into a data lake that allows you to at least start blending the data together. And an often times we found that, you know, groups were blending the data together today. They were just doing it in a very manually intensive way. And so we just tried to help them by giving them an environment where they could at least automate some of the, some of the work that they’re doing today and then provide more time for them to look at meaningful ways to actually use the information that they were creating.
Curtis: Yeah, that’s interesting. So one of the issues that comes with that of course is, is how do you help people understand what’s in the data lake? What does the data mean? All these kinds of things, like the data dictionary concept. How do you guys approach that?
Todd: Yeah, so that’s been a very interesting space. I would say especially in the last two years. I think our history, having started with thinking about managing data, metadata in different forms, you know, for example, ontologies, has definitely opened our eyes to different approaches. For us, the number one thing has always been to start with lineage. So if a business user is looking at a report and they can at least understand the queries that were executed to generate this report, that’s the first start. So it doesn’t necessarily give you all of the GUIs that you need or the easy ways to do search, but at least provides a starting point for allowing the business to understand how these reports are, are brought together. And then from there we have definitely been piloting and working with a lot of the governance technologies out there. You know, your Informatica is your culebras to understand really what their solutions offer and then bring that to the data lake itself. But for us, the, the core of most of the governance work that we do really starts with having a way to at least capture the workflows that data engineers are running every time they try to put together a report.
Curtis: I’m curious, I haven’t talked to a lot of people that have done, and this is hearkening back to what you said initially, you know, working with the government. Is there a big difference that you have seen since you’ve now worked with both, you know, government versus commercial, and I know there’s a lot of nuance within those two groups, but I’m curious if you’ve seen differences, similarities between, between that work?
Todd: Um, you know, there’s definitely similarities and differences. I would say commercial is often an easier environment to work within if you’re a small, medium size business. Sometimes with the government contracting space, you definitely have a little bit more overhead in terms of managing contracts. But if you look at the challenges that they’re face, you know, data access, data quality, not having the right analytic skills, the technologies they’re using, the maturity there, there are a lot of similarities between large, you know, fortune 2000 organizations and large government organizations.
Curtis: It may not take an act of Congress to change some policies in commercial, but there’s still significant barriers.
Todd: Sometimes it seems like it would take an act of Congress.
Curtis: That’s right. We’re coming up on time here. So maybe a, just a final question for you would be, you know, talking to people that are maybe starting this journey or thinking about getting into data and wanting to get, you know, get into a role where they can make a difference. What’s your, what’s your advice from your experience that you’ve seen, beyond what you’ve already told us? What, what, what do you think your best uh, advice?
Todd: Um, so my, I think my biggest advice for a lot of people is that there’s definitely a technical side to data and analytics. But I think what we’re missing right now as an industry are some of those soft skills. So I think a very popular topic, you know, on, on podcasts and at conferences is activation. Why aren’t analytic capabilities being activated? Why aren’t we making better decisions with data? And I think a lot of that is based on the soft skills within an organization. So I think it’s definitely important to focus on your technical skill sets. If you want to go, you know, get AWS certification or you want to become a Google certified data engineer. I think those are important because it builds your expertise. But thinking that you’re just going to stop there I think is a, is a poor decision. And I think looking for opportunities to help you build out your communication skills or round out your presentation skills are very important for really activating analytics.
And I think that especially next year, this year, already next year in, in, in the near future, I do think more focus is gonna turn on activation than anytime in the recent past. And I think the s the individuals that can really show that, that they have experience in the capability of not just working with a technical or the analytics team to develop a solution, but working with business stakeholders to really activate and change the way they make a decision. I think that’s going to be one of the most sought after skills that all companies are going to go after.
Attributions
Music
“Loopster” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 3.0 License