How KAR Auction Services is Driving the Road to ITOM Maturity

KAR Auction Services provides innovative, technology-driven remarketing solutions for the wholesale used vehicle industry. The company connects buyers and sellers in 110 countries, handling more than 4.4 million unit sales valued at over $40 billion. Headquartered in Carmel, Indiana, KAR has approximately 17,400 employees across the United States, Canada, Mexico and the United Kingdom.

Recently, we spoke with Brian Osborne, KAR’s IT Shared Services Manager, about the company’s ServiceNow ITOM rollout – and the critical factors that made this a success.

Brian, let’s start at the beginning. Why is ITOM is so important for KAR?

Ultimately, it’s about empowering our business. Great technology helps maintain our market leadership – and that technology has to keep on working 24×7. That means we need a single system of record for our entire IT environment – total visibility of our IT infrastructure and the business services it delivers. Service visibility is critical – unless you understand the relationships between infrastructure and services, it’s impossible to know the business impact of changes and incidents. We want IT to be proactive when issues arise – that means being the first to know and the first to report – so our business has confidence in the availability of the services we deliver.

So, it sounds like the CMDB is critical. Is that where you started?

Yes, we did. Unless you have an accurate, up-to-date CMDB, you’re flying blind. We used ServiceNow Discovery to automatically populate our CMDB – giving us that single source of truth we needed.

However, it’s important not to boil the ocean. Every CI (Configuration Item) in the CMDB needs to be managed. If you just turn on discovery and let all the data flow in, you’ll end up with 2 million CIs – and not know what to do with them. We took it step by step, bringing in a technology at a time – making sure we standardized our CIs and knocked down any issues we found.

Any other advice on making your CMDB a success?

Ownership is key. Don’t do everything in one central team. Engage the infrastructure and platform CI owners. After all, they’re the ones who benefit. For example, each of our owners is responsible for certifying their discovered CI data – and ServiceNow automates this by sending out certification tasks every week. This keeps our CMDB accurate – and lets us confirm that discovered changes were planned.

Once you had your CMDB in shape, what did you do next?

We added these CIs to our existing ITIL processes – incident, change and problem. For instance, when someone raises a change to patch 20 servers, they add all the CIs to the change request. This means that we can now track the full CI history – which is incredibly valuable. For example, when we have an incident, we now know if there was a recent change to an affected CI.

One other point. Don’t look for perfection. Encourage teams to add CI data, but don’t make it mandatory. We found that our existing processes and workflows needed to mature to consume CMDB data, and even partial data primes the pump. And, you can get there quickly. By setting up training sessions, we went from 17% adoption to 76% adoption in just five weeks.

We’ve been talking about infrastructure so far. What about business services?

That brings us to the next phase – event management and service mapping.

By deploying event management, we created a single source of truth for all the events in our environment. And, again, we took a step-by-step approach, rather than trying to boil the ocean. For instance, we made sure that when we turned on an alert, we also had a defined alert response procedure so that we knew exactly what to do when that alert was raised.

It was the same story with service mapping. Rather than tackling everything at once, we picked one application for each of our standard technology stacks – for example, one app running on JBoss, one for WebSphere, one for IIS, and so on. Overall, we started with 10 business services – which meant that we could deploy more quickly. And, we worked closely with our application development teams throughout the mapping process – having the right SMEs in place really saves time.

 What did it take to map these services?

Service mapping takes planning and good execution – but it’s worth it. The good news is that we could easily configure ServiceNow to map applications that were built to “out of the box” standards. The bad news? We found out that we had a lot of these non-standard applications. However, this was a blessing in disguise – we’ll now be able to work with our application development and infrastructure teams to improve how our services are architected.

What’s next?

We’re continuing to mature our IT Operations Management processes. Service availability management is a top priority, as is automating alert responses. That’s going to drive further improvements in our service quality. And, we’re focusing on service delivery as well. For example, we’re starting to automate infrastructure provisioning, and we’re creating a business service catalog.

Bottom line – we’re excited about the future.