A lot of money is spent on something, and you expect it to work well. It is reasonable to assume (at least for a while) that if you bought a new car from a dealership, it would drive problem-free. The dealership will likely issue you a warranty that outlines the types of repairs and maintenance that will be covered in the first few years of your ownership.
Service level agreements (SLAs) are warranties between a software seller and a customer (instead of a vehicle warranty). An SLA defines what service is provided by a service provider, as well as the expectation regarding issues and downtime.
It’s one of those tasks that’s uber-important yet uber-outside your scope as an early-stage founder. It has to be done, and it falls to you since you have the most product expertise and customer visibility. Although, this isn’t a case of throwing spaghetti against a wall and seeing if it sticks, then iterating. Contracts are legal documents. In the world of contracts, you can’t move fast and break things.
Because of this, we explored what makes a good SLA-so you won’t have to. Slack, Google, and Amazon have all provided outstanding examples for setting up service level agreements, and we’ve developed a model that can be used by any company.
This service level agreement template can be used for your own agreements. Defining a set of components around service performance should be done in simple language in an SLA. An SLA can act as a DIY instructions manual and problem-solver for your customers. Your customers should be told four things
- What they can expect around functionality from your service
- How the service will be monitored
- A plan of action in case something goes wrong
Steps for remediation
Describe the level of service functionality your customers can expect
The top of the page should contain a definition of the type of service. In this section, you’ll be able to explain to your customers what type of service you’re providing. In addition, it will provide an estimated performance level, which is usually expressed as a percentage (for example, 99.5% uptime).
Monitor your service with metrics
Describe how you will monitor performance and where the customer can access the monitoring data. Depending on how many third-party apps your customers use, it may be difficult for them to pinpoint where the issue is if one of them goes down. As a result of the outage of Amazon Web Services in 2020, thousands of websites were unavailable. It will help your customers see when there is a problem if you give them a status page for your app.
Offer an escalation for resolving issues
Explain what customers should do if they encounter any problems. Do you have an outage on your app? If they want more information, provide a contact form, an email address, a phone number for customer service, or a Slack channel. As soon as the problem is resolved, you’ll be able to inform them before the status bar on your monitoring site changes. It’s essential to communicate well with your customers in order to keep them happy.
Provide remediation if an issue arises
As much as your engineers may be called rock stars or wizards these days, they are still just humans. Sometimes, technical problems happen, and they can take a little bit of time to resolve. In your SLA, be sure to give a response-time window, where the customer can expect issues to be looked into and resolved. This might sound impossible, but do your best. You can also spell out clear, actionable ways customers can escalate if a problem persists outside of the agreed-upon window, such as applying for refunds for lost time.