Technology teams are failing to meet around a quarter of service-level agreements made with other business departments in their organisation.
Although four out of five (81 percent) of businesses adopt formal service-level agreements (SLAs), these are only met 74 percent of the time, according to research carried out by Forrester Consulting.
This SLA failure can result in poor performance for business applications within different business units.
More than half (57 percent) of respondents said they've experienced increased costs as a result of poor application performance, while 48 percent said they've lost revenue for the same reason.
Other effects of SLA failure cited by respondents include a negative effect on external customer satisfaction and a slowing of production.
According to Forrester, the failure of these SLAs demonstrates that many business units have unrealistic expectations of the technology team, showing a 'disconnect' between IT and business.
Forrester said one of the reasons for this mismatch is the use of metrics which focus on the technology rather than the business objectives.
The research showed 41 percent of nearly 400 respondents only have a basic insight into service levels, while SLA information often fails to get sent on to executives regularly.
Jean-Pierre Garbani, principal analyst with Forrester Research, said the end user is the ultimate judge of IT and business alignment, so IT departments must check they are delivering on expectations in terms of accuracy, availability, performance and usability.
The research was commissioned by Compuware.






Talkback
Having to get back into the market is difficult as most of the modern IT teams/ departments seem to be based around the flawed Microsoft development concept. If MS can't do it why would any smaller company succeed using their models?
All my previous employers have found SLAs superfluous and so they have never been adopted saving vast specification and tracking effort. However without the strong foundation, SLA's are necessary as a measure. This just proves the whole setup needs improving and should not be used as justification for out-sourcing, where one finds that the external SLAs are not as stringent as the internal ones, with better SLA performance as what was once an internal problem causing an SLA failure is now an enhancement request awaiting resource funding and hence outside the SLA parameters.