Manage Learn to apply best practices and optimize your operations.

Best practices for data center outsourcing

Some benefits of data center outsourcing include reduced costs and lighter IT workload, but only if all sides are communicating properly.

Data center outsourcing allows an organization to offload some of its operations to third-party service providers. It's often just the right solution when there's a shortage of space, power or cooling. It's also a helpful way to shift work away from a small IT staff that needs to concentrate on more important projects.

Play now:
Download for later:

Data center outsourcing with Pete Sclafani

  • Internet Explorer: Right Click > Save Target As
  • Firefox: Right Click > Save Link As

But a data center outsourcing initiative can also cause problems due to unmet expectations, improper scope, poor vendor communication and other factors.

Hello, I'm Tom Walat, site editor with to talk about data center outsourcing issues. I'm here with Pete Sclafani, Chief Information Officer (CIO) with 6connect, a network automation provider located in Redwood City, California. Pete thanks for joining us.

Pete Sclafani: Tom, thanks for having me.

Walat: Data center outsourcing project problems often start early, when expectations for the project are being discussed. How should businesses go about defining and confirming expectations for outsourcing projects?

Sclafani: That's a great question. One of the biggest issues on this front is not having the right people in the room during this initial discovery process. It's so easy to focus on those subject areas that you're most comfortable with, but I've seen this get bungled in the past so it's obviously something that you need to look at when looking at any sort of build out or outsourcing project.

The best example I have is during one client's data center build out. They had consulted with only the systems and networking teams but didn't involve their facilities team. The result was a data center footprint that could only support 50% of the load for which the data center was built. The solution was simple; it was just a matter of confirming the load and heat signatures of the data center gear and talking with the facilities team to even out the load in the data center. But unfortunately it all came down to basic communication issues, and if they'd been done earlier in the process (it) would have saved the company downtime, revenue and a lot of stress.

The second requirement that I cannot stress enough is to have an internal project champion along with the designated subject matter experts that can help track down the knowledge and resources for the project to be successful. Any company has a history and it's part of the personnel that work there. So involving them in the scoping stage is crucial not only assuring the project addresses the actual problems but also so you don't repeat mistakes from previous attempts.

Walat: With that in mind what are some strategies for managing and communicating with an data center outsourcing vendor?

Sclafani: I joke about it with my team that vendor management is less about checklists and Gantt charts and more about exchanging meaningful information. By using a communication method that is lightweight and gets the right people, you'd be surprised how meeting times and frequencies can be modified so that everyone is productive and no one is being blocked. I think when it comes down to the strategies around managing communication, I would start with defining the vendor types you're going to be dealing with – again it's about finding the right mix of communication vectors for the team at hand.

You don't want to assume anything, and it's helpful to have a good overview of the tools that you will need to be successful. The balance that I see is really between active and passive communication tools. Used together they can be very beneficial to project teams and especially vendor management. I generally like having a project portal that is accessible to internal teams as well as vendors. This way everyone is on the same page and has access to the correct information – if it encompasses a task and ticketing system, even better. And of course some basic milestone tools with reminders are a great way to publicly remind the team of the project status as well as your management. This solution is one part of the puzzle but it's a good baseline to at least have an environment to capture all that data.

Tom: What are some warning signs of impending problems with a data center outsourcing provider? What kind of bad behavior can kill an outsourcing relationship and when should you make the decision that it's time to jump ship?

Sclafani: That's a tough one and unfortunately it's going to depend on the vendor, the outsourcing relationship and how you're introduced. My two biggest warning signs are a vendor that does not communicate and a vendor that can't manage its time well. If a vendor is not communicating in a timely manner, that is not a good sign and chances are they're not communicating effectively to their team either. That means that you have to closely manage that vendor to ensure that your expectations are clear and that they can plan on meeting your internal milestones. Bad communication means that you end up relying on heroics to get a project over the finish line. While this does usually get the project done, a more process-oriented approach with vendor transparency is much less stressful.

Time management and communication issues have multiple symptoms. But the most obvious signs are missed milestones and no clear post-mortems for project issues. Those are obvious signs that whatever methods they're using are not working and the last thing you want is the vendor learning on your dime, unless that is an understood condition.

Walat: So what happens when the relationship collapses? How should you go about switching vendors and minimize the short-term disruptions and long-term impacts on your business?

Sclafani: Ideally you're going to attempt to salvage the relationship and the efforts that you've already applied. But you're right. If it's just not working out, you've got the unenviable position of transitioning the outputs of whatever you've done to date to another vendor and ramping them up or simply starting fresh with another vendor and pretending nothing happened. Now you're behind schedule.

Unfortunately this is largely a function of deadlines and information quality. If the vendor issues centered on communication and competence, the project outputs to date may not be usable or trustworthy. If that is the case, transitioning to another vendor may not even be feasible. If it does require you to start fresh – just as a caveat – the new vendor that you would go with will definitely have an opinion on this, because they will be dealing with your deadlines and that data and they'll have their contract on the line as well.

But no matter which way you go, there are always going to be delays. Your best bet is to create what I call "backup points" for various milestones to ensure that your management team is aware of them. The good side is that it requires complete understanding of the project so if you have the right people in the room this is not as daunting a task as it may seem.

Walat: Finally, Pete, when is it time to simply pull the plug on data center outsourcing and bring those operations back in-house? What's the best way to approach that challenge?

Sclafani: I think there are two answers here that you have to take into consideration. One is just purely business and the other has a financial component. Either way it requires having some metrics in place –which you would need for any outsourcing project – because, again, metrics are really what's going to help you drive what the purpose behind that project is. So bringing in an outsourced operation back in house "just because" is usually not a good idea because it usually means those metrics are either subjective or non-existent – typically a bad thing.

On the business side, if your metrics are around expenses or customer satisfaction, it's easy to do the math and identify the break point where it makes sense to move that service. Sometimes it may be cost-effective to outsource for a given time period. Let's say you outsource a call center for nine months to give your revenues and cash flow a chance to ramp up. Other times it may be cost-effective to outsource on a permanent basis. You may decide – if it's a commoditized product or service like email hosting – that it's completely fine to outsource it completely. You might say you're going to have that hosted by Microsoft and just wash your hands of it.

On the financial side you also have to look at the "whys" behind the data center outsourcing in the first place. From a financial perspective, maintaining a service expense may be more beneficial for the balance sheet than adding head count, so get to know the CFO and ensure that their feedback is taken into consideration as well for outsourcing projects.

Walat: Thanks Pete! That's all the time we have for today. I'm Tom Walat and you can learn more about data center outsourcing at Thanks for joining us.

Dig Deeper on Colocation, hosting and outsourcing management

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.