what does a good service level agreement look like?

In previous post we discussed Getting best performance from your outsourced services, for this entry we thought it worth going into a bit more detail as to what a good Service Level Agreement is and how to evaluate one if it is presented to you by a vendor. 

Firstly, what do we mean by Service Level Agreement?  At its simplest, these are commitments your IT provider is making for continuity of your IT assets and addressing issues that might come up for their contracted service. 
Common elements to any service level agreements are:
 

• Scope and exclusions of what is being supported  
• Availability promises 
• Issue response and resolution based on gradings of impact to the business 
• Costs for availability of support 

Let’s break these down along with things to look for under each heading. 

Scope and exclusions of what is being supported  
There should be a clear description of what you are paying for and what aspects of that the Service Levels apply to.  Typically, you have already made your IT investment, this section describes what of those IT assets the vendor will be “looking after”.  An example here might be an agreement to support your computer servers, the scope might include keeping your servers up to date with the latest operating system patches (fixes and improvements to the software the runs the computer) but exclude patching the time sheeting software running on the server. 

You can see why it is important to be clear on the scope and exclusions.  If your operation system change affects the time sheeting software, you will find yourself with a timesheet vendor pointing at the server support vendor and vice versa. 

Our recommendation is to ensure you have a clear view of our IT assets – hardware and software and who is looking after what.  And what “looking after” means; is it monitoring it for issues, is it regular maintenance, is it break-fix.  These play into who will notify whom of issues when they do occur – we’ll over issues later in our discussion. 

Availability promises 
Keeping a system available to your staff can mean the difference between meeting your clients' expectations or letting them down.  Depending on your importance to your client the great the impact on them and thus on your reputation, ability to receive payment or deliver your services.   

A service level agreement will typically set parameters for how available your system is; is it available 24/7 365 days a year or is it okay if you can’t access if after hours or weekends, or are there key times of the year, week or day that it has to be running for your business to function.  We have other blogs about business continuity planning (something we can help with), they will need to kick in should your system become unavailable.  

There are two dimensions to availability: 

1. Unplanned downtime – when the system, unexpectedly, becomes unavailable to you  
2. Planned downtime – when you and the vendor agree to it being unavailable for some reason (usually planned upgrades and maintenance) 

The availability is usually expressed as a percentage with caveats, e.g. 98% availability, in business hours or 99.5% availability 365 days a year, excluding 10 hours per month for maintenance. 

Usually, the greater the availability the higher the cost as there is more redundancy built into the system and risk on the vendor.  You need to ensure the setting matches that systems importance to your business. 

Issue response and resolution based on gradings of impact to the business 
So, we know what the vendor is looking after and how available it will be, but things happen, sometimes the solution fails, sometime parts of the solution fails or you find an issue you’ve not encountered before but shouldn’t be there.   

How do you want your vendor to respond will depend on how big the issue is, or the priority.  Your POS system being down when you're in the middle of a customer rush would qualify as high-priority, whereas having to click a button twice to make an action happen is annoying but not going to stop you from functioning.   

Your agreement will show include a description of what qualifies an issue as high-priority or lower and who decides the priority. From there it will describe what they will do about it and how quickly.  There are some “tricks” or terms that’s sneak in here around response time: agreements state a response within 15 minutes for a priority 1 issue.  Often vendors will consider an automated reply as meeting that timeframe, it is worth getting clear on what a response means and resolution times.   

Resolution times can be a challenge for vendors to commit to as there is an element of the unknown for issues.  A tough one to balance, however, don’t let that stop you from setting expectations. 

Costs 
These will vary based on the configuration of the above elements, the vendors own resource costs and margins.  Usually this is split into at least two parts: 

1. Availability / support costs (A set fee per month for the ability raise issues, monitoring etc) 
2. Additional hours costs (sometimes on increasing scale depending between different staff capabilities) 

Consider what you will get in the first cost – does it cover x hours of effort to resolve issues, or simply the resolution off certain issues or just the maintenance and monitoring?  How will you authorise additional hours if they are necessary?  These questions need to be answered in the agreement and clarified with the vendor before signing. 
​
 The above is just a quick run over the things you need to consider when reviewing a service level agreement.  If you are considering one or looking to engage a vendor for such services feel free to contact us for support with your negotiations – we can help explain things to your or your needs to the vendor.