Maintenance made manageable: A guide to SLAs

Flynn Maloy is starting to spend a lot less time explaining to customers why system downtime is like an iceberg.

The worldwide marketing manager for technology services with Palo Alto, Calif.-based Hewlett Packard Co., had often used the larger submersed portion of an iceberg to represent the often-overlooked contributors to system failure, like processes and people. “These days it’s not about the device, it’s about the environment,” he said.

The fact that customers — and maintenance vendors alike — are taking this more holistic approach is one example of how the maintenance service market is changing.

ComputerWorld Canada talked to a variety of players in the market to help IT managers get a better understanding of how to improve the way they handle the ongoing needs of their IT infrastructure.

Parsing the agreement

Understanding maintenance agreements requires focus on key components, in particular a vendor’s definition of service level agreements (SLAs) or the amount of time within which the vendor commits to delivering support, said Patrick Zanella, product manager for data center support services with Westborough, Mass.-based third-party maintenance provider Akibia Inc.

He said a customer also needs to know how to submit queries to the vendor and how the call gets handled: by phone, e-mail or online.

“Those are critical questions to find out because if at two or three or four o’clock in the morning when you’ve got a hard system down and you’ve got people looking for help, the last thing our customers are looking for is a pager or a call back,” he said.

If dealing with a third-party maintainer (TPM) versus the original equipment manufacturer (OEM), he said customers should find out when technical and field staff will be ready to provide support after a new product is launched. As well, know which software versions are supported, he added. And in general for all vendors, he said know whether they provide the maintenance support themselves, or through a subcontractor.

The term of the agreement also varies and typically depends on the age of the hardware being maintained, said Zanella. It’s also a good idea, he said, to re-examine maintenance contracts to account for changes in infrastructure like a previously mission critical system assuming a more basic status, or vice versa. “Unless someone knows to go in and modify the service levels to make it less aggressive, they’re still paying for a higher service level agreement,” he said.

SMBs have similar needs

But overall, the specifics of a maintenance agreement boil down to the impact of failed systems to the business, or of client-facing systems to customers, said Benjamin Jacob, director of sales of professional services at Santa Clara, Calif.-based Sun Microsystems Canada Inc. “A personal desktop may not have as big an impact…as back-end servers that might be supporting thousands of e-mail users,” he said.

Furthermore, Jacob doesn’t see significant difference between the maintenance requirements of an enterprise versus a smaller business.

But small-to-medium-sized businesses have a “different culture” that influences maintenance demands, said Bob Verbeke, vice-president of product support for the technology support and maintenance practice with Blue Bell, Pennsylvania-based third party maintainer Unisys Corp. SMBs may need similar types of service levels during the business day, but he’s observed a lesser level of coverage required outside of that timeframe. Also, given the varying level of in-house technical expertise at SMBs, a different support model is often required, he said.

But company size aside, all businesses have the option of buying support from one or numerous vendors depending on their IT infrastructure and budget, said Mark Jacobson, director of sales and business development with Hightstown, New Jersey-based third-party maintainer QSGI Inc. Having a single point of contact for all maintenance issues can be an easy sell for IT admins, he said, but, the need to cut costs can often mean switching vendors for a particular group of equipment or platform.


Besides understanding the basics of an agreement, businesses have the choice of buying services from the OEM or TPM. Both have their virtues, said Ron Silliman, principal analyst for infrastructure support services at Stamford, Conn.-based research firm Gartner Inc.

Enterprise-type organizations with mission critical systems like data centres may prefer to seek maintenance from the OEM, and benefit from a price strategy that the OEM can afford to play. However, Silliman said OEMs have a tendency “to do things their way” because of their experience with what they believe makes equipment work best. TPMs, on the other hand, he added, “made a virtue of being flexible”.

Customers will often hear “scare stories” of TPMs being unable to source hardware parts and understand the equipment, but Silliman said TPMs have over time developed their own methodology for providing service — something that’s worth listening to. Furthermore, a customers’ due diligence should entail speaking to vendor references.

Market shifts

But agreement essentials aside, the maintenance market itself has morphed due to factors like the U.S. recession at the start of this millennium, he said, which made organizations want to actively control support costs. Before, customers had the same service level agreements across all equipment, mission critical or not.

Such trends continue to shape the maintenance market today, and in turn, the service options offered to the customer. According to QSGI’s Jacobson, the last 12-18 months has seen IT managers really pushing back on maintenance pricing. He figures that’s because maintenance is the sole area where businesses feel they can cut costs considering IT staff is already meager and software and hardware costs are pretty much fixed.

Sun Microsystems’ Jacob said although he hasn’t witnessed that cost pushback, he did note customers are increasingly demanding managed services on top of traditional hardware and software support, like preventative-type services that can free up the in-house staff’s time to do other things besides the “proverbial putting out fires.” In fact, Jacob said he’s seeing more enterprises interested in offloading their data centre entirely to OEMs. Along a similar vein, Hewlett Packard’s Maloy said CIOs who typically allocate at least 80 per cent of the IT budget to keep systems running are looking to change that with an “in-house resource shift” that brings on intermittent support staff for tasks that really don’t need full-time supervision.

Among other trends, Unisys’ Verbeke expects to see a shared maintenance approach between vendor and customer with the emergence of technologies that are more supportable in-house. Whereas maintenance logistics have traditionally been managed by the vendor, that paradigm will extend to the customer who “will be effecting those repairs” if they choose that option. The idea that customers might share the maintenance is reflected in a newly-emerging approach taken by companies with data centres who are looking to cut costs. Rick Gray, vice-president and general manager with Devon, Pennsylvania-based third-party maintainer DecisionOne Inc. has observed a “surprising increase in customer interest to move to that self-maintainer strategy” where some products allow in-house staff to easily swap devices themselves assuming they have the technically-competent people.

Also, the injection of bonuses and penalties into maintenance agreements have become commonplace, said Gray. That way, customers can get financial reimbursements if the vendor underperforms, and conversely, will provide financial bonuses if the vendor over delivers.

Another emerging interest among customers is the potential to migrate to open source technology, said Verbeke. “That’s presenting a challenge in dealing with support communities” with which necessary relationships have yet to be bridged compared to those built and solidified with more established entities, he said. “We really haven’t seen where we’ve been able to get commitments from them to provide the type of support that you can get on an OEM piece of software.”

Customer also have concerns around green IT, said Maloy, especially considering 45 per cent of the cost of running a data centre is the electricity.

Related Download
The Landscape of Self Service Analytics Sponsor: IBM
The Landscape of Self Service Analytics
Download this report to examine the current state of self-service analytics across all industries and company sizes, and view the technology decisions and analytical performance of organizations that reported high levels of self-service in their analytical use base.
Register Now