Not all managed services are created equal. New research (available for free until Aug. 29; registration required) from Aberdeen Group highlights some of the features that distinguish stellar network services from the rest of the bunch.

The research firm surveyed 143 organizations in April and May in an effort to glean best practices for using managed network services. To gauge the most successful engagements, it focused on three performance criteria: application response-time improvements; how quickly issues with network performance are repaired; and application availability gains.

For starters, Aberdeen looked at why enterprises are adopting managed network services. The top three drivers are the need to reduce operational expenses (cited by 44% of survey respondents); a lack of internal resources for managing network performance (42%); and the need to reduce network management complexity (35%).

When it comes to realizing those goals, the most successful companies share common operational and technical abilities. Operationally, some important factors include the ability to track trouble tickets in real time and perform rules-based root cause analysis. Successful users of managed network services also typically have staff dedicated to working with their providers, Aberdeen says.

On the technology front, Aberdeen found it’s important for customers to have access to tools that provide: an analysis of packet flow data; remote management of network assets; automated comparison of network performance to SLAs; an integrated portal for visibility into network services; and WAN acceleration.

One common stumbling block for companies using managed network services is related to how success is measured. There’s a disconnect between the goals that organizations are trying to achieve and the SLAs that they are currently using, says research analyst Bojan Simic, who authored the report.

Companies associate a successful engagement with improved application performance - but 60% of companies surveyed have not deployed SLAs for application performance. SLAs today are most commonly related to network downtime (used by 71% of respondents); mean time to repair (61%); responsiveness to escalation event (58%); and time to implement (46%).

Ann Bednarz is associate news editor at Network World.