Technology and changing customer expectations are transforming the automobile buying experience from purchase to service. Industry analyst and CXOTalk host, Michael Krigsman, talks with a top technology executive at America's largest auto retailer.

Adam Rasner is head of IT Infrastructure and Operations for AutoNation, which has served over ten million customers in over 340 locations across 15 states. AutoNation is part of the Fortune 500, with revenues over $19 billion.

Transcript

Michael Krigsman: The auto industry, buying cars, we have all bought a car, a used car or a new car. Like every other part of our world, the car retail industry is changing. Today, on Episode #273 of CxOTalk, we are speaking with AutoNation, which is the nation's largest retailer of cars, of automobiles.

Before we begin, I want to say a heartfelt thanks to Cohesity, which provides secondary storage, because they're underwriting today's episode, and they're making it possible for us to be here. I've been working with Cohesity; preparing for this episode. I have to say, they're just great to work with.

Without further ado, I want to introduce Adam Rasner, who is with AutoNation. Adam, how are you? Thank you for being here.

Adam Rasner: Hey, Michael. Doing great. Thanks. I'm excited to be here today and talk to you.

Michael Krigsman: Tell us about AutoNation and tell us about what you do there.

Adam Rasner: Sure. AutoNation is the country's largest automotive retailer. We're made up of a couple different lines of business. We have about 300 locations made up of new car stores, preowned car stores, AutoNation USA, as well as collision centers. We also operate multiple auction houses, and then we've just launched a new line of business called AutoNation Precision Parts. That's going to compete with the original equipment manufacturer of parts.

We're in 14 states, again about 300 locations. We're about $22 billion in revenue. The company is growing tremendously. We're about 27,000 associates.

Michael Krigsman: You are, what, I think 126 on the Fortune 500 right now.

Adam Rasner: That's correct.

Michael Krigsman: You're an extraordinarily large company, so your view of the car sales is, I suppose, more expansive than just about anybody.

Adam Rasner: Yeah, there are other automotive groups out there, but we're by far the largest, which has pluses, and it also has its challenges.  Yeah, we are the largest retailer in the country.

Michael Krigsman: Now, tell us about your role. What do you do for AutoNation?

Adam Rasner: Sure. I'm over the infrastructure and operations team. It's basically made up of systems engineering, so data center operations, network services, operations, so change management, incident management, enterprise monitoring. We also have the end user computing team, so we basically have the whole underlying infrastructure of all the systems that run AutoNation.

Michael Krigsman: You're responsible for these systems for hundreds of locations. I think it's about 350 or so.

Adam Rasner: Yeah, it's over 300 and about 27,000 associates.

Michael Krigsman: Now, Adam, let's begin by talking about the car buying experience and how all of that is changing. Tell us about the car buying experience.

Adam Rasner: Yeah. Anybody listening today, when they go out and they've bought a car, there are very few people that come back and say, "I had an amazing car buying experience." It's an industry that hasn't changed much in 50, 60 years. It's a four- to eight-hour ordeal at best, and it sometimes is painful. We think that there's a lot of technology that can be added into the process to make it easier, simpler.

We have a day in the future where we see an Amazon-like experience for buying a car. I think that's coming very, very soon.

Michael Krigsman: We're going to have to talk about that Amazon-like experience. But, when you say that car buying is difficult, what makes it hard, and what are the technology elements that must be put into place in order to reach that kind of Amazon experience you were describing?

Adam Rasner: Yeah. I think, today, the challenge is, even though consumers are way more educated than they were even 20 years ago when they go in to buy a car, they're able to go, search, and understand what the pricing should be for the new car or the pre-owned car that they're buying, there's still a lot of skepticism by the consumer about the dealerships and the sales process

Michael Krigsman: The auto industry, buying cars, we have all bought a car, a used car or a new car. Like every other part of our world, the car retail industry is changing. Today, on Episode #273 of CxOTalk, we are speaking with AutoNation, which is the nation's largest retailer of cars, of automobiles.

Before we begin, I want to say a heartfelt thanks to Cohesity, which provides secondary storage, because they're underwriting today's episode, and they're making it possible for us to be here. I've been working with Cohesity; preparing for this episode. I have to say, they're just great to work with.

Without further ado, I want to introduce Adam Rasner, who is with AutoNation. Adam, how are you? Thank you for being here.

Adam Rasner: Hey, Michael. Doing great. Thanks. I'm excited to be here today and talk to you.

Michael Krigsman: Tell us about AutoNation and tell us about what you do there.

Adam Rasner: Sure. AutoNation is the country's largest automotive retailer. We're made up of a couple different lines of business. We have about 300 locations made up of new car stores, preowned car stores, AutoNation USA, as well as collision centers. We also operate multiple auction houses, and then we've just launched a new line of business called AutoNation Precision Parts. That's going to compete with the original equipment manufacturer of parts.

We're in 14 states, again about 300 locations. We're about $22 billion in revenue. The company is growing tremendously. We're about 27,000 associates.

Michael Krigsman: You are, what, I think 126 on the Fortune 500 right now.

Adam Rasner: That's correct.

Michael Krigsman: You're an extraordinarily large company, so your view of the car sales is, I suppose, more expansive than just about anybody.

Adam Rasner: Yeah, there are other automotive groups out there, but we're by far the largest, which has pluses, and it also has its challenges.  Yeah, we are the largest retailer in the country.

Michael Krigsman: Now, tell us about your role. What do you do for AutoNation?

Adam Rasner: Sure. I'm over the infrastructure and operations team. It's basically made up of systems engineering, so data center operations, network services, operations, so change management, incident management, enterprise monitoring. We also have the end user computing team, so we basically have the whole underlying infrastructure of all the systems that run AutoNation.

Michael Krigsman: You're responsible for these systems for hundreds of locations. I think it's about 350 or so.

Adam Rasner: Yeah, it's over 300 and about 27,000 associates.

Michael Krigsman: Now, Adam, let's begin by talking about the car buying experience and how all of that is changing. Tell us about the car buying experience.

Adam Rasner: Yeah. Anybody listening today, when they go out and they've bought a car, there are very few people that come back and say, "I had an amazing car buying experience." It's an industry that hasn't changed much in 50, 60 years. It's a four- to eight-hour ordeal at best, and it sometimes is painful. We think that there's a lot of technology that can be added into the process to make it easier, simpler.

We have a day in the future where we see an Amazon-like experience for buying a car. I think that's coming very, very soon.

Michael Krigsman: We're going to have to talk about that Amazon-like experience. But, when you say that car buying is difficult, what makes it hard, and what are the technology elements that must be put into place in order to reach that kind of Amazon experience you were describing?

Adam Rasner: Yeah. I think, today, the challenge is, even though consumers are way more educated than they were even 20 years ago when they go in to buy a car, they're able to go, search, and understand what the pricing should be for the new car or the pre-owned car that they're buying, there's still a lot of skepticism by the consumer about the dealerships and the sales process, and it's difficult. We see a day in the future where you can go to a website. You see the car built the way you want [and] and all the competitors' pricing. You apply for your financing online. You add your extras, like your wheel protection, your extended warranties, and things like that.

The dealership, maybe you don't interact with the dealership at all. Maybe the dealership is a fulfillment center, and we deliver the car to your front door. You sign electronically, and it's a very quick experience. You have very little interaction on the sales side. I think that's coming.

There are some people that may take advantage of that full experience, do the entire transaction online, and the car rolls into their house. They sign an iPad, and they're done. There may be people that aren't comfortable today going all that way, and so they do some of the transaction or a lot of the transaction online, and they still go to the dealership for that last signing or maybe a test drive before they sign. There are baby steps or iterations to get to that full Amazon experience. I'm not sure the consumer is completely ready for the home delivery without dealing with a salesperson at all, but I think the millennials of the world are certainly game for trying something like that, so we're working towards it.

Michael Krigsman: Now, I want to first off say thank you to the almost 14,000 people who are watching right now and to remind everybody that we're speaking with Adam Rasner from AutoNation about how the car buying experience is changing. I want to say thank you to Cohesity for underwriting this episode of CxOTalk. Right now, there is a tweet chat taking place, and you can join in using the hashtag #CxOTalk.

Adam, you've just described this vision of an Amazon-like car buying experience. What will it take for the industry to actually get to that point?

Adam Rasner: There are things that are within our control: developing software and systems that can support this kind of all electronic experience. Then there are things that are regulatory that are government challenges. In some states, they require live ink on every document to buy a car. That's going to really challenge our vision. In other states, you can do some of the documents electronically and some you have to sign on paper.

There's also DMV elements and other things that are challenging. There's also a lot of third parties in the mix, so to be able to sell you extended warranties from their parties or things like that, all that integration has to be built. We're actively working on it, and it's coming, but it takes time.

Again, there are things that are within our ecosystem that we have total control over and there are things that are going to take more time, are going to require more effort. The vision is there. The path to get there is hard to define an exact time, but we're working towards it.

Michael Krigsman: It sounds like in order to achieve this very transformative vision, you need technology. There are data issues, and there are also process change issues. It sounds like it's all of these things coming together.

Adam Rasner: Yeah. One of the challenges is, our systems that we use today are systems that have been around for a long time. I have to find a way to keep the lights on and keep those systems running to continue to conduct business today while we innovate. That certainly is a challenge for us. We have to figure out how do we continue with the systems that we have today, as well as innovate, start brand new, and take the company in a different direction. That has its challenges, and we're finding our way slowly but surely. This has support of the executive leadership here, and so we're going to keep moving it forward.

Michael Krigsman: The car industry, like so many others, is undergoing this dramatic change. At your very, very large scale, how do you go about innovating and yet, at the same time, keep your operations running with 27,000 people [and] over 300 locations? How do you go about doing that?

Adam Rasner: Yeah. Well, that's certainly the million-dollar question. The strategy we've taken is we have teams that are kind of focused on the legacy applications that are continuing to support them. Even the legacy applications need enhancements and additional functionality before we're ready to cut over to the vision that's some ways away. We have teams focused on that, and then we're building new teams for the innovation to support and develop the new applications [and] new, underlying infrastructure.

That's the approach we've taken just because, again, you're right; our size and scale [makes] it's very hard. We can't just close the door on what we're doing today and go in a totally different direction. It's got to be more of a transitional, phased approach. That's the way we're handling it.

Michael Krigsman: Where do the consumer experiences or the consumer expectations come into play with all of this?

Adam Rasner: Well, I think today's consumer is kind of frustrated in today's car buying process. I go back to what I started with. You're hard-pressed to find somebody that really says, "Wow, I had a great car buying experience." They're out there, but they are few and far between.

I think today's consumer wants a much easier transaction. They want transparency. I think a lot of them want to do much more of it online. The whole ability to not really be able to negotiate when you're dealing with a Web portal that has a set price, it's a cultural thing that we're going to have to work to change. It's going to have to make the experience, this Amazon-like experience, be really seamless, and people are going to have to trust that the price that we're giving them is a great price. That may be showing them our competitors' pricing so that when they push the button to buy this car, whether it's online or they've done most of it online and they decide to go to the dealership for the last piece, they have a level of comfort that they got a great deal and they had a great experience. The consumer is demanding something different, and so we're trying to deliver that.

Michael Krigsman: How do you know? Maybe it seems like kind of an obvious question, but what are the signals from the market that tell you that the consumers are making these, have these other demands, these additional demands of you, expectations?

Adam Rasner: Yeah, I think a lot of it has to do with the traffic that we see to our website. The first interaction we usually have with anybody's leads through AutoNation.com or through microsites that we have for all the different OEM dealerships, that first contact is almost always online now. We hear regularly, "How come I can't do my financing end-to-end?" or, "How come I still have to sign 30 pieces of physical paper to do this transaction?" "Why does it take so long?" Those are all great questions, and I think they all can be solved with technology.

Michael Krigsman: How do you break it down? Do you look at it by process, by step in the lifecycle? How do you begin to break this all down?

Adam Rasner: I think what we try to do is, at a high level, dissect the different pieces that are part of a transaction. The trade-in of your old car, the buying of the new car, the financing piece is a big component, the add-on extras that you may or may not want to buy in a transaction, and how do we make those individual components all electronic? How do we make them smooth? How do we make it an easy, online experience? Then we build that into one website that you can conduct the end-to-end transaction. It's an oversimplified explanation, but it's a very complex problem that we're trying to come up with an easy, consumer-facing solution for. But, we will get there.

Michael Krigsman: You break it down into processes. For you, the processes are what? You have sales. You have service. How do you look at that journey?

Adam Rasner: We want to own the entire customer experience from the minute that they first interact with us and we get that lead from one of the websites, all the conversation that happens in between, to then they bought the car, [and] their service experience, which we haven't even talked about.

Today, there's not a lot of technology introduced in the service experience. We're working towards online scheduling. You drop your car off. Your rental is already pre-setup. You grab the keys, kind of like your rental car companies today. You look for your name on the board. You grab the key and go. You come back. We text you when your car is ready. You can pay for that service online. If there are things that come up in the diagnostics that you didn't know about, you can authorize it via text message.

It's, again, not just the initial sale. It's the service experience afterward. Then, when the time comes for you to sell that car again, we want to start all over again with you. It's a whole lifecycle of the consumer. Of course, AutoNation wants to own that experience end-to-end.

Michael Krigsman: That's how you break it up. In fact, you have all of these different businesses that are kind of tailored or oriented towards those different points in the lifecycle process.

Adam Rasner: Exactly. Exactly. The Precision Parts is the newest thing. We sold you the car. Hopefully, over your lifespan, we've sold you multiple cars at AutoNation. You brought it in for service. Now we can offer you a battery at half the price of the OEM battery. You, as the consumer, have the option to save some money and get a battery with the same warranty that the EOM offered. It just seems like the natural progression. We're continuing to look for more opportunities to be part of the lifecycle.

Michael Krigsman: From a data standpoint, I'm assuming that you must somehow share data or thinking about the sharing of data across different points in the lifecycle to support that consumer experience. Obviously, personalize the experience to support selling.

Adam Rasner: Yeah, exactly. It's pretty complex. But, if you go into an AutoNation Land Rover today and you buy a car from us, and then five years from now you go into AutoNation Mercedes, BMW, whatever it may be, we want to know that you were our customer at these other dealers. Here is who your preferred lender is. Here's how you finance it. We can try to bring what we learned in that first transaction forward to the new transaction.

We have elements of that today, but we don't have it all. It's something, again, that we're working on. There's a lot of data to be mined there, and there's huge value in it. Yeah, it's very important to us.

Michael Krigsman: There are so many different directions we can take this conversation, but you're responsible for technology infrastructure, and so we should talk about that. Tell us about your technology infrastructure.

Adam Rasner: Yeah, we're about 85% on-prem today and about 15% in the cloud. My challenge is, again, I've got to provide great infrastructure, reliable infrastructure to the existing legacy apps. Then I have to be nimble enough to be able to have a scalable infrastructure to be able to be the runway or the landing ground for all these new apps that are coming and the new innovations.

We are definitely building a lot of net new in the cloud because of the scalability and some of the cost savings. In the meantime, we continue to support a lot of on-prem infrastructure. The plan is, as the legacy stuff kind of gets sunset, as the new technology comes up, we'll probably be much more of a cloud infrastructure company and much less on-prem.

Today, we've got multiple data centers in several cities.

Michael Krigsman: Do you have any kind of timeframe for making that cloud transition, or is it just at this point kind of a general objective?

Adam Rasner: Yeah, so anything net new that comes into the environment, we always look. Hey, is this a great candidate for cloud? There are some economics that you have to look at. There are some technical constraints that you have to think about in cloud. A lot of things are making sense. When there is opportunity to either refactor an existing app or build net new, we always look to cloud first if the numbers both economically and technically make sense.

I see the scale changing very drastically over the next five years where today we're 85/15, 85% on-prem and maybe 15% cloud. That shifting to being--I don't know--if I had to guess, in the next five years it could be 50/50 or even more cloud than on-prem. It really depends on what happens with a lot of the legacy technology.

Michael Krigsman: Does the shift from cloud to on-premise demand a lot of retraining and new processes?  Are there people issues, or is it mostly a technology issue for you?

Adam Rasner: Mostly technology issues. Most of your audience out there knows when you go to move something from on-prem to cloud, a lot of times it's not just a lift and shift. Things have to be refactored, and that has cost. Our challenge is, especially since we're kind of going through a lot of technology innovation, do we refactor these old apps to make them "cloudable" where they can be moved to the cloud, or do we leave them as is and focus on the net new stuff for cloud? We're making those decisions on an app-by-app basis as it makes sense.

Michael Krigsman: One of the things that you and spoke about during our last conversation is the unique set of enterprise of apps inside the car retailing business. It's pretty different from many other industries. Maybe describe that to us, if you would.

Adam Rasner: Sure. I think that the cornerstone of the car buying experience is the dealer management system, which is actually software as a service. There are only two or three big players out there. We rely very heavily on these third-party providers for sourcing parts, the car purchasing experience. Because we're so large, we also have written a lot of custom integration into these applications. There are also things for the DMV, for service, so we control some of our destiny and other parts of it are sitting with other third-party software companies. That has its pros and cons.

Most of the dealership groups out there are much smaller than us, and so they don't have kind of these enterprise-wide features that we really strongly desire.  We've taken an out-of-the-box software as a solution and really done a lot of customization, which again has its pros and cons. It's expensive. It's hard to maintain. It's hard to work through updates and things like that. But, I think, in this new world, we really want to control more of our own destiny, so we're seeing what are those features that we can build more into our own platform rather than have it sitting with third parties.

Michael Krigsman: In your case, going down the road of custom apps to own your own destiny is the strategy that you've pursued.

Adam Rasner: Yeah, and it's a big deal because not only is it developing the technology [and] building the infrastructure. That's probably the easy part, but it's lots of process changes. Some of these dealer management systems have been around for a very long time. The sales team and the service department, they're trained on how to use these systems. When you start building your own software, that whole learning curve has to change. We've got 27,000 associates out there that have to be retrained, so it's a major undertaking. It's years and years in the making.

Michael Krigsman: Now, I want to talk about the relationship that you have with Cohesity. I again want to say that we're very grateful to Cohesity for underwriting CxOTalk. We depend on our underwriters to do this.

To being, tell us about what you're doing with Cohesity. What are you doing with Cohesity now?

Adam Rasner: Sure. We came to Cohesity to deal with the current challenge we were facing. Enterprise backup was a huge pain point for us. We were using a product that was kind of antiquated. It didn't scale well for the size of the organization and the amount of data that we needed to back up. We started a search basically first for a new enterprise backup solution. I put the companies that do this work in two buckets: kind of the legacy providers and then you have kind of the next-gen providers.

What we found is, the legacy providers, they didn't scale well, expensive, and they really hadn't done a whole lot with the platform in the last ten-plus years. We then turned to the next-gen solutions, and Cohesity was brought to us by one of our value-added resellers as a really good candidate.

We vetted a couple of players in that space. We did a great proof of concept with them. The beat out the other competitors in what I'm calling the next-gen category. As an added bonus, not only does it scale infinitely and easily. That was kind of a big win for us. We didn't want to replace appliances every couple of years as they hit their capacity. We also now are able to use the Cohesity as a tier II storage. Applications that didn't need our most expensive SSD storage on our very expensive SANS, we were able to use Cohesity as a target.

We got this added benefit that we didn't really anticipate up front. It was really going to be an enterprise backup solution. Now, we're able to pivot and use it for other things. A very successful implementation and very happy customers.

Michael Krigsman: When you make the distinction between a legacy provider and what you call a next-gen provider of technology, I'm interested. Is it just the technology alone, or are there elements of relationship? Are there elements of mindset and culture? How do you make that distinction?

Adam Rasner: I think the big thing about the next gen providers is their hyper-converged architecture. It just scales infinitely, easily. We had used hyper-converged computing in other use cases in the business, so we liked that model. That’s how we came to kind of key in on Cohesity. I'd say the technology is the big differentiator.

The legacy providers, again, just haven't innovated very much. When you hit the capacity of that piece of hardware, you're replacing it every year. That gets very expensive. You have to do data migrations. On the technology front, the next-gen thing was just in a different category on its own.

As far as relationships, really, you're putting your enterprise backup in the hands of a software company, a software and hardware company, so it's super critical that we made a good decision. When we need to go hit the button and restore, we have to push that button with complete confidence that the data is going to be there, it's going to restore quickly, and that the data is good. We took this decision very seriously, and it was through the proof of concept process that we gained that confidence.

Again, we did a POC with not only Cohesity, but the other players in that space. We also even brought in a couple, one or two, of the legacy providers just to do our due diligence. At the end of the day, everything we expected out of the next gen feature-wise was there. Then Cohesity just simply beat out the other competitor in that next-gen space on performance, on execution, and that sort of thing.

Michael Krigsman: What about the dimension of trust, because this type of product is completely mission critical? If you ever have a problem with your systems or your data and you can't get back to where you were, I mean the results for your business can be catastrophic. How do you evaluate trust and those kinds of business dimensions that go beyond the technology alone?

Adam Rasner: Sure. There are kind of the soft things you can do, which is, talk to the Gartners of the world. Do the customer references. Talk to your peers. Then there is actual proof in the POC, the proof of concept. This is where the big differentiator is.

With the competitor to Cohesity that we had tried in the POC, we ran into quite a few technical challenges. In fact, pushing the button to restore the data and the data not being there, that's a big problem. Even though there was a technical issue and got it resolved, it killed trust immediately. If things fail during the POC process, you've lost the trust to go forward. Even though things can be fixed, you're always nervous that it's not doing what it's supposed to do. That's where I think things went really right for Cohesity and didn't go right with the competitor that we brought in.

Michael Krigsman: When things fail, you're talking about technology failures of one type or another.

Adam Rasner: Yeah, things like saying the backup job completed and the data is copied. Then we goo to restore it and it's not there. That's a big problem. That's a core competency of a backup solution. When you're doing a POC and these companies are supposed to be putting their best foot forward, then you have an anomaly like that, whether it's their fault or not, the trust. That's back to your question of trust. I don't want to worry at night that if we had selected this other vendor and we push the button to do the restore, I don't want to worry that the data is not there, the data is corrupted, or something like that.

Michael Krigsman: At the end of the day, you have technology solution, and then you have the confidence that, when you need it, that technology solution is actually going to work.

Adam Rasner: 100%, yes. The sales guys, they do their job, and they do a great job. But, it's really about trying the solution in your environment, talking to your peers. I think Gartner is a good resource for getting some unbiased opinions. But, at the end of the day, the engineering team here and myself had to have a comfort that we were picking the right enterprise product. Backup, while it's not sexy and a lot of fun, it's critical. We took the decision very, very seriously, and we spent quite a bit of time doing our due diligence.

Michael Krigsman: How do you evaluate this type of--I was going to say--complex relationship because, again, I always look at things through the lens of the technology side and the business side? When you have mission-critical technology, there's always technology risk and major business risk. How do you evaluate the vendor in this type of situation?

Adam Rasner: Yeah, and again that's a great question because both of these next-gen solutions that are out there today are competing very heavily against each other. They're relatively new companies, and so that's a concern, you know, viability. It was really, again, we did our own kind of analysis of the company, their financials. Are they solid? We want to make sure they're not here today, gone tomorrow. We do our own kind of vetting before we cut a PO of this size to any company, especially when it's relatively new.

Then again, it's back to did we talk to Gartner? Did they validate what we believe is the roadmap for this company? I trust my peers, so I do ask for and solicit feedback from my peers. At the end of the day, that coupled with POC experience is how we come to make decisions.

Michael Krigsman: In terms of evaluating the success of this kind of relationship, are there metrics or KPIs? How do you go about making that evaluation?

Adam Rasner: Yeah. There are hard metrics. The solution we were using previously, we were dealing with something like 6,000 backup failures a month from hardware issues, software problems. Basically, almost 60%, 70% of an FTE, a full-time employee, babysitting backup jobs to now we're dealing with a few a month, and we actually can go in and have a good understanding of what's causing those and remediate.

From a statistics perspective, our backup percentage is much better. Again, I mentioned that we moved one of our key document management system storage to Cohesity. Didn't plan on that in the beginning. It just was an added bonus that here is an opportunity to save some money and move storage off high-cost SSDs. That was an added bonus.

Then there's just a comfort level. We had a lot of challenges with the old system. I don't want to wake up in the morning… if we had to do a restore, I need to have a level of comfort now.

I think the engineering team is really on the hook for execution when we need to do a restore. I think they have a level of comfort now that they didn't have with the old. Again, there are soft things that you measure and feelings that you get, and then there are kind of hard statistics. We're winning on both ends of that.

Michael Krigsman: It's funny because, as you said, backup is not the most sexy technology that's out there, or sexy technology domain, but the technology itself sounds like you love it, and it's sexy, geeky stuff.

Adam Rasner: Yeah. I can tell you. Our team never really was excited about backup, but when we finally made the decision, they said, "This is going to make our lives easier." The interface is easy to use. We were up and running very quickly. Again, they're the ones they call at 3:00 in the morning that an upgrade didn't go well and they need to revert back and restore something. Those guys have a new level of comfort that the data is going to be restored quickly, that it's not going to be corrupt, and so we're happy.

Michael Krigsman: Just in closing this segment of our discussion, I just have to say I'm still floored by the notion of 6,000 backup failures a month. How does that even happen?

Adam Rasner: Yeah. Without naming the vendor, we were on a hardware/software platform again that didn't scale to the size. We're over a petabyte of data, which is pretty significant. All of that data is being backed up.

We had a lot of hardware challenges. It was an appliance-based solution with a software overlay. We had a lot of hardware related failures, and then the challenge was we would go rerun the job. It would be successful. Then the next time it would fail.

We could never really get a good handle on why things were failing. We tried to work with the vendor. There came a point with the old vendor that basically said, "We don't have a customer of your size using this solution. This is not the solution for you." It's something we had in the environment for a long time, and so it was definitely time to make a change.

Yeah, we spent an inordinate amount of time, and that's high risk. If those jobs aren't rerun and they're not rerun successfully, we're at potential data loss if there's a failure on primary storage or something like that.

Michael Krigsman: Well, it's certainly been an education for me about data storage and data backup. In general, what advice do you have or organizations that are looking to replace mission-critical systems like you were just describing? What should they do? What should organizations making this kind of a change do to ensure that both of these dimensions, the technology dimension and that trust and confidence dimension, are satisfied by the vendor?

Adam Rasner: Sure. I think, really, it's all about due diligence and to spend the time. Do the POC. Not everybody loves doing it. It takes time. Our data centers are not local to our corporate office. We had to have people travel. The vendor had to have people travel.

Especially on a mission-critical app, make the time investment to try it out. It's one thing to sit in a sales presentation in your corporate office and have them kind of give you the dog and pony show. That's great, but the proof is in the pudding. That is installing it, seeing how it operates in your environment, and then getting a really good feel for how hard it is to deploy, what's the learning curve for your team going to be, those sorts of things.

Then there's the easier stuff, which is talking to your peers, reading the trade magazines, [and] doing customer references. I usually try to find customer references on my own rather than use the ones that the sales guys give me because, in 20 years of doing this, the sales guys never give me a bad reference for their product, so I try to find people on my own that are using it. If you do all of that I think, at the end of the road, you should be at a place where you're relatively comfortable to make a big decision.

Michael Krigsman: Dive into that a little bit about the finding of the references. I agree. The salesperson is never ever going to give you a reference that's a bad reference.

Adam Rasner: Yes. It's never happened. While those are great, it's really through networking. Again, we have a Gartner membership. You can go on there. You can find people that are users of the product. You can interact with them electronically.

Also, just from people in the industry. I've been in the industry a long time, and so I have a pretty good network of people to reach out and say, "Hey, is anybody using Cohesity? What's your experience? Was it everything the sales guy told you it was going to be? What were your challenges? What are the features that you don't have today that you'd like to see tomorrow?" Outside of the POC, that's really the best information you can get because those are the people that have already lived the path you're about to go on.

Michael Krigsman: Yeah. I think that's absolutely great advice because, if you can find those references through some path other than the vendor, they're going to be honest references. Not that the vendor is going to supply a dishonest reference. I don't mean to say that. But, you're going to get a clearer, more neutral reference, I think.

Adam Rasner: Absolutely. I've had references that I've found on my own that have gone both ways. Sometimes it's in the middle. Sometimes it's, "This is a horrible solution. This is not the best solution. It's done pretty well. Here's where my challenges are. Here's what I like about it. Here's what I don't." That's the best feedback you can get because then you go in open-eyed or wide-eyed to what you're about to get into.

Michael Krigsman: The reference checking becomes really almost advice session as well. "If you buy this product, if you go down this path, try this; don't do that," that kind of thing.

Adam Rasner: Yeah. "Here are the pitfalls or here is an expectation I had of this product that the product met or didn't meet," or, "Here is something that I thought was in there day one, but it's in the roadmap to the future." There's a lot of this kind of information that you're not going to get from the sales team. It's the real world where the rubber meets the road, and so I really like to operate that way. It helps me collect a lot of really good information.

Michael Krigsman: As we wind down towards the end of this show, and boy it's been a fast almost 45 minutes, what's next for AutoNation? Where are you taking it? You're going towards that Amazon-like experience. What are the steps along the way? What's your grand vision?

Adam Rasner: Yeah. I think the next year or two is to start building the applications and the underlying infrastructure to support that vision. A lot of that is already underway. Again, for the pieces of that experience that we can control in its entirety, I think we have a good roadmap and we're on the path to being able to deliver it. There are, again, things that I can't control that are regulatory and otherwise that we're going to work through. But, it's coming, and I think it's coming faster than people think. Yeah, be on the lookout.

Michael Krigsman: I read that AutoNation is now getting involved with Waymo, the driverless car company.

Adam Rasner: That's correct.

Michael Krigsman: I don't know. I don't want to put you on the spot. Can you say anything at all about that?

Adam Rasner: Yeah. Basically, we've signed on to be one of the exclusive servicing companies for these autonomous vehicles. Autonomous vehicles, in my personal opinion, we're some ways away from completely driverless cars, but it's coming. And so, AutoNation decided strategically, if consumers aren't going to buy cars and we're all going to be on a rideshare model or something like that, who is going to service these cars? Wisely, the leadership team made a strategic partnership with Waymo. We're going to be servicing those cars as they go into the real world. It is an exciting partnership for us.

Michael Krigsman: Yeah. I found it fascinating because here you are car sellers, but you have all of these other businesses grouped around the different processes like service. You're kind of hedging your bets because, even when there are driverless cars, those cars still need to be fixed.

Adam Rasner: Right.

Michael Krigsman: [Laughter]

Adam Rasner: If the driverless car and the ride-sharing thing take off like some of the projections show, we won't be selling to consumers anymore. We'll be selling to fleets of driverless car companies. We've got to make sure we pivot the business to be able to accommodate that.

An easy win for us was, well, who is going to service all these cars. We have a great footprint nationwide to be able to service these cars, so that's how that deal kind of came about.

Michael Krigsman: The margins on selling new cars is virtually nothing anyway.

Adam Rasner: Yeah. There is not a ton of money to be made anymore in selling cars, and that's why the company has kind of gone in different directions with the Precision Parts line, AutoNation USA, which is our used car only or preowned car only superstores. We've got a bunch of those. Then rather than sending off cars that we would normally send to a third-party auction, in the volume that we're doing, we said, "Let's start doing that ourselves," so we operate multiple auctions across the country.

Then collision centers have also been a big business for us. I think we're at around 80 locations right now. We're going to flex and bend based on the industry, and these are the moves that we're making. So far, they've been pretty good.

Michael Krigsman: As we close out, I just have to ask you about this notion of innovating while maintaining stability in the business because it seems like basically that's your life. [Laughter]

Adam Rasner: [Laughter] Yeah. Yeah, it is the single biggest challenge because everybody in IT leadership, every one of us are going to say we're resource constrained and the business has more work than we could ever possibly do. That's just for keeping the lights on.

Then you add all this innovation you want to do. It's a challenge.

Our strategy has been, again, to dedicate teams. We've kind of drawn a line in the sand. We said these teams are going to support legacy applications, the enhancements because, again, this is probably a year or two or more away. We've got to be able to support those legacy applications and continue to add new functionality to them. They can't just sit idle for two years while we build the next new thing.

Then we're kind of building new teams to work on just the net new. I guess we're fortunate that we're able to kind of structure the team that way, and we have enough capital and resources to be able to do that. That's our approach. That's the only way we see ourselves as being successful. If we try to comingle legacy and the innovation stuff, the focus usually winds up being on the legacy stuff, and we're not able to ever move these things forward. That's the strategy.

Michael Krigsman: All right. As we go out, what advice have you got for other organizations that are facing change in the way that you're facing and undergoing transformation? What advice do you have for going through it, sailing through it seamlessly? Let's put it that way.

Adam Rasner: Sure. A lot easier said than done. But, what I would tell you, and I'm looking at it from an infrastructure and operations point of you, you have to have really tight alignment with the business and understanding what they're trying to accomplish, what that's going to mean to you staffing wise, hardware wise, operation wise, and being very tightly aligned to the business objectives. Then there are the tough conversations like we were just talking about of, "Well, how do you want to staff this to be successful? How do you want to divide up the work to make sure that everything gets done?"

It's difficult. Usually a lot of times it comes down to the bottom line in money and properly resourcing. It's a challenge, but I think we're on the right path. I think we're really tightly aligned to what the business wants to do going forward.

Michael Krigsman: Well, given your enormous size as an organization and how geographically dispersed you are, it's almost amazing that you're able to do any innovation on the ground and still keep those operations flowing flawlessly. Thanks so much for sharing those lessons today.

Adam Rasner: Yeah. Thanks for having me, Michael. I really appreciate it.

Michael Krigsman: We have been speaking with Adam Rasner from AutoNation. I am Michael Krigsman, and you've been watching Episode #273 of CxOTalk. Again, I want to say a heartfelt thank you to Cohesity for supporting CxOTalk. I've been working with them. They're a great company. Check out Cohesity.com.

We will see you again next week. Actually, no, we'll see you again on Friday for another show. Thanks so much, everybody. Have a great day. Bye-bye.