Network managers don't really know what kinds of service level agreements they can get from telecommunication service providers, and when they do know what they can get, they aren't terribly interested in more than a guarantee that the network will keep working, according to a study released this month.
Participants "really had a hard time" giving specifics about network uptime guarantees or SLA (service level agreement) features, when researchers from Sage Research in the US, pressed 15 network managers in two online focus groups. The results also reflect the views of Australian IT managers.
"Part of it is that SLAs have been around for quite a while, but only in the last few years has there been any emphasis on them," said Jared Huizenga, a senior analyst with Sage.
Many companies established their SLAs before service providers began touting guarantees as a marketing tool, and older SLAs aren't reviewed with frequency, he said. Having a service level agreement, and understanding what's in it, are two different things though, an analyst from researcher IDC said.
"These can be really protracted legal negotiations," said IDC senior analyst Paul Bugala. The executives making the agreements are often not the same people involved in the day-to-day operation of the network, leaving an information gap between the negotiators and the network managers.
For Steve Tucker, IT manager of Adelaide-based aircraft service operator National Jet Systems, his division has various SLAs for voice and data carrier services, and network, hardware and business application system support, and understands what the SLAs contain. However, he said general business managers within his company "don't understand the concept of an IT-style SLA," but the organisation can get by with this because he is "not convinced they have to [understand them], necessarily."
Cynicism deepens when network managers have no way to judge whether or not a service provider is upholding their end of the bargain, said IDC's Bugala. "There is not enough information about how the service provider is delivering the service," he said. "There's a curtain...what has to be clear to a customer is a guarantee about latency, or throughput."
Larger customers tend to have the in-house resources to monitor their agreements, but smaller organisations do not. And making service providers fulfil on the contract usually requires the customer to provide evidence of service failure.
In Tucker's view, there is "very little way" of managing network uptime guarantees in his company of 900 employees. National Jet Systems consolidates and prints its SLAs in table form, then refers to them in day to day operations, calling on its helpdesk system when the need arises, he said.
More important, getting service providers to abide by contracts is a constant challenge, Tucker said, as effectiveness "depends on how tangible the product being provided is."
"Things like bandwidth, availability and uptime' can be difficult to ascertain if well-sorted monitoring or reporting mechanisms are not in place."
Selena Chan, Shiseido Australia's manager of information systems, said companies, regardless of their size, struggled to assess whether or not their provider delivered adequate levels of service due to poor customer education by vendors on performance monitoring. "We don't have enough information or tools to help us evaluate situations," she said, adding that: "Changes in the [network management] area are so quick that IT managers may not be aware of all the new services or new technologies available to them."
Specific concerns Chan cited in relation to her company's SLA with its ISP were the technical incompetence and sluggishness of support staff, and the provider's refusal to accept responsibility for clients' connection problems. "If we've experienced those problems, we have to spend a lot of time finding out who the problem party was. A lot of money is wasted, and in the end it's proved that the provider was the problem."
When asked to rate the most important elements of an SLA, network managers in the study chose guaranteed availability and uptime as most important, over other seemingly important service issues like latency guarantees, packet loss guarantees, and notification time for security breaches.
"What people were saying is that as long as the network was up, it's OK," Huizenga said. "If the end user within the company isn't experiencing downtime ... then the organisation is going to say it doesn't matter, that maybe it's not such a big deal."