State Auto Insurance is a $2B personal and commercial lines insurer based in Columbus, Ohio. Like most players in the industry, State Auto is embracing digital transformation in order to stay competitive and relevant to its customers. A 28-year veteran of the industry, Greg Tacchetti, State Auto's SVP/Chief Information and Strategy Officer, was brought on to rewrite the corporate strategy and realign the entire organization around digital.
Thanks for taking some time to talk about IT leadership and the changing culture at State Auto. Can you tell me a little bit about what’s going on at the company?
I started at State Auto leading IT organization two-and-a-half years ago. I am also responsible for strategy for the corporation but this is really my first stop running IT.
At State Auto, the argument I make with the board is that all the ingredients are in the soup for a lot of disruption today. And while you probably could say that's true over the last 20 or 30 years, I think the velocity has gone up quite a bit.
What I refer to as 'the ingredients'; it's the rise of the public cloud giving almost anybody, on demand, a huge compute capacity. It's gaming chips with massive, parallel processing that allow huge data sets both to be stored and then crunched. And then it's tools like smartphones and the ubiquitous nature of sensors that are everywhere creating all this data that, particularly in my industry, are going to make risk a lot more transparent.
In our world, the more transparent the risk is, the better I can price for it and pass those savings on to the consumers. You see this in insurance with “InsureTech” and people who have all kinds of different ideas creating new products or new forms of distribution. It's almost mind-numbing trying to keep up with it.
What were the IT challenges you faced when you first arrived at State Auto?
Coming into this role at a company that had severely underinvested in IT for about 10 or 15 years, we had to do a massive back office uplift in order to even get to the place where you could start doing things like a feature-rich app that customers take as table-stakes today. You can't do any of those things if you don't have a good underlying data architecture to serve up the data and make it all function well.
To put all that together, I needed to get a leadership team in place. That's where I spent a lot of my time initially and I still do.
One of my favorite meetings is a biweekly leadership forum. It's with some of my younger managers and also people who don't have direct managerial responsibility but are in roles of influence, like recall technical leaders. You know, they may not have any direct reports but they are influencing young developers all across the organization.
It's 30 of us in a room, with no agenda. Somebody will say, "Hey, I've got a challenge. Can I get some advice?" And it's just a really nice way for groups to share things that they are working on and challenges they are having with development or coordination across the department or, even more broadly, across the company.
This sounds like a scrum in Agile, right?
Yeah, but just on leadership topics. One of the things I like about it is that it gives me an opportunity to hear the challenges my teams are dealing with and to impart, in a more casual way across my whole organization, different leadership advice around things that I'd like people to be thinking about. Technology departments don't focus on this area enough— just the communication and how we lead our teams.
What is more important to achieving the outcomes the business wants: leadership of technology or people?
It cannot be about the tech. You have to understand how all these pieces fit together but the role that I play in [corporate] strategy and leading the broader digital transformation at the company is to try to paint that roadmap of where we're trying to get to and then that helps my team make decisions.
And allocate resources?
The one thing that I've seen, whether it was GEICO or any other major insurance companies I’ve worked with, is they get so big that you end up having a really fractured IT environment—the claims group is running the claims projects and the policy group is running the policy projects and you end up having these data silos all over your company that you're not leveraging to really try to get to a good experience for your consumers.
That's really what I'm able to do here in a company our size: make sure that we're being really smart about our investments and that we have the right endgame in mind.
Speaking of making the right investments, how are you funding the more forward-looking ideas?
Innovation is part of the culture that has developed here over the last two years. There isn't a specific leader or org structure and therefore no specific budget. Innovation comes from all around State Auto. Budgetary control ultimately flows to the senior leadership team, who meets monthly to evaluate proofs-of-concept (POCs) and approve future funding.
We also have a venture capital fund, State Auto Labs, where we look at a lot of start-ups for potential investment. But this is typically outside of our “normal” internal innovation.
When you talk about strategy, are you referring just to the technology strategy to achieve predetermined business goals or are you "at the table" when it comes to developing it?
Oh, no. I mean that I developed and wrote strategy for the whole corporation. IT has to fit into that but it's for the company.
So you wrote the business strategy and then took that to your team or did you have an idea that you wanted to use cloud in one area and go in-house in another? Did you have an idea of the architecture or did you just let the teams tell you what they thought was best?
With respect specifically to cloud, I had used Amazon previously so I was familiar with it. And given the state of the state and how aggressively we needed to move to build the back office I mentioned earlier, I looked at the development cycle and the number of overlapping code threads that we were going to have with this aggressive schedule. When I talked to the I&O team, I asked one question, "How long does it take you to stand up a new environment?" They said, "Forty-five days."
I made the decision right then that we had to move to Amazon. We just didn't have time to buy enough servers and rack them and stack them and get them configured. And I would have probably ended up having 50%, 60% excess capacity on servers until we grew into it in a couple of years, just to get through all of the testing.
It was more of a unique situation. I know a lot of my competitors locally will spend nine months evaluating and come out of a situation and say, "We're going to have a hybrid cloud strategy." We were more existential, I guess, in that I didn't really have much choice.
It sounds like some major changes needed to happen.
That whole department needed to get turned upside down. Now, the I&O side is very well-versed in Amazon. They're doing a bunch of work with GCP [Google Cloud Platform]. And we're moving into infrastructure-as-a-service.
We've brought in a lot of tools to automate the environment builds. When I was in my startup [Greg was the owner-founder of an insurance-related company based in Seattle], I could build an environment in about 20 minutes. They've got it down here to about 30 minutes for a much more complex environment. That’s the path to DevOps that this team has consumed very quickly and I couldn't tell you half the stuff that's in their head but I could paint the vision of continuous integration and help them do the research and figure out who to partner with to make that happen.
Your job was to set the goal posts and find the people to help you if you didn't already have them. I'm sure you also had to make some hard decisions about personnel and who was going to be able to help you facilitate that vision, right?
You know, I think we had 120 people just in I&O and I have 15 now. So we outsourced a lot of the arms and legs to a third-party provider and then we just kept the best to manage that offshore-onshore resource as they got up to speed on some of the things like we've been talking about. The cyber team had 25 people. I kept one.
Step one for me was to get the right leaders in place and make sure they understood the mission that we were trying to accomplish. And then a big part of what we talked about after that was attracting and retaining top talent. I need a talented group that has the right kind of customer-centric focus to be doing the things that a lot of other modern e-commerce shops do: looking at where buttons are in a page and doing A/B testing.
Because if you're even a day behind the times, you're going to lose clients.
That's exactly right.
So what changes did you make to facilitate that need?
We did two things. One, we really flattened the organization on the managerial side. There's really only three layers, including me. And then we also went through all of the jobs in the department. You know, people who were very good technically wanted to make more money. So they sought a promotion into management and that really is not where their skillsets or their passions lay.
When we rebuilt the job families, we put in a career path that had more upside for individual contributor tech leads. A number of people who had been in jobs they didn't love opted to say "I'd rather just be a technical lead. That's what I really enjoy, technology, and I'll let other people do the coaching piece of it." That allowed us to have fewer managers and was a big part of how we were able to get things stabilized and moving quickly in another direction.
What other changes have you made to your team to help State Auto embrace digital innovation?
We have good developers, but I don't think we have enough good design people. It just hasn't been part of this company's delivery mechanism and we've got to solve for that because now that the data architecture is fixed and the back office is fixed, I'm going to be putting a lot more things in front of consumers.
There is zero tolerance [from customers] for screwing up. I mean if you download an app and it doesn't do what you want it to do, you're never going back to that app. You'll delete it off your phone and say to yourself, "That company doesn't get it," or, "amateur hour," or whatever.
Okay. I see where you're going with that. But to your point, this type of digital integration—from IT all the way out to the customer—requires a new set of skills. It requires different thinking. It’s really about people, right? Is that what leadership is ultimately about?
Well, it's a circle of people, processes, and technology but I start with people because if they aren't onboard—hearts and minds—it just gets one heck of a lot harder. For people to work well together, your processes have to be well understood and communicated and you can't have tool overload.
We've eliminated probably two-thirds of the tools that were here when I got here and added in some newer ones. Having a vision for how all the technology fits together, whether it's at the very technical, “wonky” level with enterprise architecture or me painting a picture of a capability set and the assets we're going to invest in from the big picture perspective, is critical.
You try to treat people as best as you can but at the end of this, I want hearts and minds. One of the things we talk about a lot here is what top talent looks like. And the root of this for me goes back to the three years I did with my startup when I was able to leverage relationships I had built over the years and put together a really talented team that managed a global operation and was able to move mountains.
Twenty people did more work than most of the big companies I've dealt with because they were good communicators, they were good strategic thinkers, they were very consumer-centric, and they went out and found problems and made them go away. And the lesson for me is that when you get the right group of people, there's not a whole lot you need to do. You know, they will figure it out themselves.
Bringing that startup mentality to a bigger company is never going to be easy but that's part of what I've been trying to drive here. This constant focus on attracting and retaining top talent is something that I talk about with my leadership team all the time and I challenge them to go out and find it.
One of the big challenges for most IT organizations is communicating value to senior leadership. How do you do this?
We are in the business of serving our associates, agents, policyholders, and claimants so we have built an integrated team where business people are embedded into the development process. Everyone understands that unless we work as one team to deliver technology solutions, we all will not be successful. Everyone understands the importance of getting great solutions to our customers and the embedded business team members handle the prioritization. Also, they are very involved in the exploration/innovation process of finding new solutions to problems.
The only real examples where we don't deeply collaborate are with very technical work, like whether or not to pursue a cloud strategy or technical architectural decisions.
Those are really great insights, Greg. Before we wrap things up, I was wondering what books you've been reading or recommend to help other CIOs do what you are doing?
My two favorites from the last year include Michael Lewis' book, The Undoing Project, where he talks about two Israeli psychologists, Danny Kahneman and Amos Tversky and their psychological research on human biases. That to me was a fascinating read. I mean, he's one of my favorite authors but it really opened my eyes to unconscious things that are, whether experiential or environmental, underlying your gut reaction to things.
It's helped me work with my leadership team here because I can challenge them and they can challenge me. "Why are you saying that? Are you saying that because that's what worked at GEICO or because the three of us have worked together for 25 years?" Because maybe that's not the right relative lens to be looking at this strategic problem through.
And then I'm reading Simon Sinek's book right now, Leaders Eat Last, and Daniel Kahneman's book, Thinking, Fast and Slow, which has a little more detail on the research Michael Lewis talks about in The Undoing Project.
About the interviewer: Allen Bernard is a technology journalist, editor and business writer, as well as the contributing editor for the TBM Council's book, "Technology Business Management: The Four Value Conversations CIOs Must Have With Their Businesses."
»Continue reading: We have to do WHAT? An interview on culture with CIO Gerry Maritz.