A 2018 Rightscale survey reveals that 96% of respondents use cloud technology in some aspect of their business. But have managers kept up with their cloud vendors to ensure that service-level agreements (SLAs) for uptime, performance, problem resolutions, etc., are where they need to be?

If not, here are five best practices to apply to every SLA:

SEE: Research: Cloud vs. data center adoption rates, usage, and migration plans (Tech Pro Research)

1. Clear lines of demarcation

When a business enters into an SLA agreement, lines of demarcation should be drawn and agreed to by all participants as to who is responsible for what. For example, say a user notices that it takes an exceptionally long time to pull up an application that an outside service provider hosts and manages. This may mean a problem is with the application, the problem can be in the network, or it can be a hardware problem on the end user’s desktop.

The first responsibility in this case is to report the problem. This is the responsibility of the person who first notices the problem (in this case, the end user). Next, the issue must be reported to corporate IT and the service provider. Then, it is likely that a period of problem co-determination between IT and the service provider will occur.

If the problem rests in the corporate network (or any other internal corporate resource), the resolution responsibility belongs to corporate IT. If the problem concerns facilities and/or networks belonging to the service provider, responsibility for problem resolution lies there.

Taking ownership (and assuming responsibility) for communication is the most important aspect to any SLA management. As long as everyone stays in the communications loop and status is actively communicated, the political and personal problems that sometimes ignite during problem identification and resolution gets minimized.

2. Regular communication

Parties to any SLA agreement should meet regularly, whether or not there are problems with the system. Many companies and cloud vendors meet quarterly, where they review SLA performance and necessary tuneups.

3. SLA tuneups

SLA tuneups are extremely important because the more experience you gain with a particular cloud provider, the better sense that IT, end users and business partners gain on what’s really important in terms of cloud performance. This is why every SLA agreement within a contract with a cloud vendor should leave room for at least annual revisitations of SLAs and possible revisions to those SLAs based upon new information.

SEE: Cloud migration decision tool (Tech Pro Research)

4. Contract enforcement

When SLAs are identified and agreed to at the time a contract is formed with a cloud vendor, they should also include reward and even penalty clauses for either exceeding or failing to meet SLAs. In the case of outside service providers, these reward/penalty provisions almost always come in the form of money that is either earned or forfeited. based upon SLA performance. If the SLA holder is IT, reward and penalty provisions (especially in large enterprises with chargeback systems) can also be monetary.

5. Link back to business value

Both IT and business users need a way to link back SLA performance to the business values they deliver. For example, if split-second transaction times are an integral part of customer service, and the company sees a high abandon rate from online customers because of wait times, then this SLA issue directly impacts revenues. In this case, diligent SLA management is required–from the onset of the contract through service delivery, and the ultimate analysis of effectiveness (and value) of service quality.

Image: vaeenma, Getty Images/iStockphoto