From Code Jockey to CIO| Alteryx
Yeah, so I started out as your typical computer science "code jockey" writing code in a cube. I initially worked on pension and annuity calculations for the Chicago Teamsters Union for about 7-8 years. After that, mobile devices started to take off, and I began working in iOS and Mac developing apps for the App Store. That was sort of like the 21st-century gold rush. It felt like everybody was making apps and trying to sell them on the App Store. Anyways, while building apps I had the realization that I really enjoyed working with and analyzing the data that the application collected. Right around that time in 2010/2011 data warehouses and data analytics was starting to take off. I don’t think many people had mastered the competency at this time - even the big players. I began working for Symphony Health and as new products like Alteryx and Tableau started to come to the forefront, we started integrating those tools into our processes. “Alteryx played a critical role in enabling many users to start to produce and consume analytics. And I personally became more visible in the organization, which propelled me from a data analyst to an analytics manager, and then eventually into this Chief Analytics / Chief Information Officer role.”
Yeah, it's interesting because the title is somewhat misleading. I have my more traditional CIO work – meeting with the Applications teams, Help Desk team, and Clinical Operations team, understanding project delivery timelines and project status, etc. From this standpoint, it’s very much an executive role. But the analytics role is much more functional; it’s more 50/50 executive guidance and hands-on functional work. Even as an executive, I’m still using Alteryx, I'm not just telling my analyst to build a workflow and upload it to the server. I'm doing the work; I write SQL queries all the time. A big part of this is just the reality of a midsized business, where I'm not detached from what's happening in the weeds. And then sometimes, you know, things break, and all heck breaks loose, and you have to roll up your sleeves and deal with those problems too. So yeah, those make for a fun day.
“You have to partner with your business counterparts to be successful and move the needle ─ it’s a balancing act.” I have become more engaged with the sales and marketing teams, for example. I don’t just deliver information to them; I consult with them and help them understand, for example, what the data tells us about the marketplace. Then I’ll help determine what actions we should take to address shifting trends in the market. The tools in healthcare historically haven’t been the best – CRM, EMR, etc. And as a result, many third-party analytics add-on tools struggle to deliver value. I’ve been asked to replace some of those add-on tools with Alteryx-driven analytic apps. The big question in the industry seems to be: “Where do you go for new business?” That’s where I can add the most value; I need to identify opportunities to find patient growth and bring those opportunities back to other business leaders. This has been a big change ─ understanding from a data perspective where we go to market using our data or where to open new service lines. Healthcare tends to lag in technology maturity ─ not just analytics (again, think EHRs, for example). Now, we are starting to come into our own. Remember that healthcare is complicated ─ we are overburdened with complexity, which can always impact innovation. But we are increasing our data analytics maturity, which underlines how important analytics is to our mission today.
I'd say the one thing that comes to mind is this: Every time we get a new request to deliver an analytic product, it’s difficult to nail down what the business leader wants. A good analogy for me is that it feels like I am throwing darts at a dartboard, blindfolded, while the dartboard is also swinging on a pendulum. The business leader has this sort of moving target, and they're also relaying to me what they want, but I can't see what they see – I can only interpret what they describe to me. So playing out that analogy, I’m firing darts, just trying to come close to what they want, and what follows is this highly iterative process – the back and forth between what the business leader initially asked and then what you're able to deliver from an analytic standpoint. Much of this is exploratory, so they may only begin with a hypothesis. Then you serve them what they requested – and this is typically just the beginning of the process. That leader comes back to you, and you iterate on this process repeatedly. And the reality is that many analysts grow tired when faced with a high degree of repetitive work, rework, uncertainty, and lack of clarity. It’s common to hear an analyst say, “Well, I created that report. Isn't that good enough for you?” And I have to remind them – that business leader initially thinks problem A is the issue when it's really problem B that they are interested in exploring further. You're sort of like Sherlock Holmes; you have to help them find the actual underlying cause. “To me, this insight generation process is the biggest bottleneck and where tools like Alteryx Designer and Auto Insights come in – because we can rapidly read, recreate and adjust information to get to that endpoint.” Those traditional bottlenecks of data storage and speed are gone. You know, those were problems ten years ago now. It's getting to the answer faster. That's the biggest bottleneck.
Exactly. It's funny that you ask that question because I had a conversation earlier today with one of my retiring analysts. I asked, “can you propose somebody that might be a good fit to backfill your role?” And they said, “Well, I have someone highly skilled, but their attitude and temperament might not be a good fit.” I said, “That's because you know exactly what the issue is. An individual can be highly skilled, but they need that soft skill of working with people in the business.” And I think there's sometimes a misunderstanding that an analyst isn’t a “people person,” but they are – the good ones have to be.
Auto Insights is very much like a shortcut. It automatically performs much of what I think an analyst would ask or do in that back-and-forth to create any type of insight. I'm glad we're sticking to the word insight and not reports and dashboards because I'm trying to separate the two. Generally, we'll create lots of reports to get one insight, but Auto Insights gives us a direct look at exactly what is causing a particular metric to move. From there, we can hone in on the root causes and supplement them with more data. From a productivity standpoint, you're looking at what would normally take, on average, 40-80 hours at minimum to develop an “Insight.” With Auto Insights, the time to insight occurs in minutes or hours – it’s less than one day of work. So, five days of work compressed into one day – that's what I've seen. Sometimes it's even more productivity gained than that. Regarding decision-making – Auto Insights enables us to take an insight and immediately start making decisions. For example, we can see clinical decisions that are made on admits and discharges that really move the needle for our business - and that's huge. It's much faster than where we've been before.
Yes, it's faster decisions that are made with more clarity. There's less uncertainty around the decision. In the past, our business stakeholders would say something like, “the data is telling me X, and I think the analyst did this the right way. So, I'm 80% sure this is the right decision.” Contrast that with Auto Insights, where we have a lot more trust in the analysis, there is absolute confidence in the business, and the software-generated insight tells us exactly what's going on. And I think that's huge. You know, trust is a huge issue in analytics. If they don't trust what we're building, they don't come to us for help. They'll use canned reports or built-in stuff, or they'll pull things into Excel themselves and do it on their own because they don't trust what we build. Learn more about Alteryx Auto Insights here.