Having strong exit clauses: These have become imperative in all contracts. Not only are they important because of the relative financial weakness of many players, but strong exit clauses will also remain important during the next several years as managed services mature and organizations suffer through poor performance. Performance-based clauses should be sought (e.g., vendor not meeting service levels during X period of time, failing to perform X service by a specified date, or failing to support X technology by a specified time).
Additionally, company-health-based exit clauses should be sought. Although it is not normal to see exit clauses based on bankruptcy (though they do appear in limited cases), it is common to see change-in-control clauses (enabling exit after a transition period).
Automatic colleting and reporting of service credits: It is not acceptable for the vendor to force the customer to track when a service credit may be due and file a request. This practice is now outdated. A cap should be expected on the amount of the fees that can be credited back--usually between 33 percent and 50 percent of quarterly fees, but we now see some innovative vendors offering 100 percent credits (no cap). We do not expect this practice to continue due to profitability concerns, but we do expect to see excessive accrual of service credits triggering exit clauses or additional services.
Clearly defining chronic and severe problems: This should identify how often a problem can occur during X period of time before being treated as chronic (e.g., a similar alarm 3x times in four hours). Also included here is accounting for the length of time that must pass after a "fix" is in place before an alarm can be considered new and unrelated to the previous problem. Chronic problems should trigger a higher level of investigation and attention from the service provider. Also, a post-mortem meeting should take place after a severe or chronic problem is resolved to learn from what happened, review the reaction, and work to prevent it in the future.
Measuring the speed of new infrastructure deployment or procurement: This includes gaining a commitment from the vendor regarding how fast a new server, database, etc. will be deployed or even procured (depending on the relationship).
Disclosing all underlying SLAs supporting a given contract:
Service providers engage with other service providers to fulfill contracts. A list should be obtained of which "hidden" service providers are used and what SLAs are in place between those vendors.
Expecting and tracking detailed performance data: Although this includes availability, it also goes beyond it and covers response time and availability of not just the application, but of the underlying infrastructure as well.
