Not many people know what a Service Level Agreement actually is but as the name implies, it formally defines a service provided by a supplier along with a detailed description of various aspects like responsibilities, quality, and scope that go into providing this service. An SLA also serves as a contract between the service provider and the customer related to on-going enhancements, inventory restocking, and on-going payments.

Now you may wonder why preparing and having a Service Level Agreement is even necessary in the first place, so here is why. An SLA works to protect both parties involved in the contract, the service provider and the customer where future work is concerned. Service Level Agreements are thought for service based businesses and are applicable to a variety of businesses involving custom built software systems, vehicles, appliances and products stocked on shelves. But what is even more interesting to note is that SLAs are also used by attorneys, doctors and insurance companies.

That is not all; SLAs have become more popular nowadays as more and more telecommunication companies and internet service providers have started preparing Service Level Agreements in more nonprofessional terms. For instance, SLAs offered by internet service providers or telecommunication companies may focus on specifics like MTTR (Mean Time to Recover or Repair) or MTBF (Mean Time between Failures) and different rates of data.

Understanding and Managing Service Level Agreements

© | Ralf Kleemann

In this article, we will help you understand what an SLA actually is and how it works. You will also learn 1) the different types of Service Level Agreements that are applicable to different types of businesses, 2) the key elements of a service level agreement, 3) SLA management 101, 4) process of SLA, and 5) some examples.


Service Level Agreements can be defined at different levels and thus build three major types of SLAs explained below:

Customer Based SLA: A customer based SLA can be best defined as a contract or agreement with an individual group of customers which contains details regarding all the services they use. An example of this type of SLA is an agreement between an IT services provider and the finance department of a multinational company for services provided like procurement system, payroll system, billing, and finance system.

Service Based SLA: Service based SLA is a contract between customers using various services offered by the service provider. For instance, a car service station may offer routine car maintenance services of a certain level along with the universal charging process. A telecommunication company, on the other hand, may offer a routine service to its customers while offering a certain part of maintenance as part of the offer with universal charges.

Multilevel SLA: The Multilevel SLA is divided into three parts – corporate level, customer level, and service level that address three different types of customers for the same services being offered by the company, within the same SLA.

  • In the corporate level SLA, all service level management issues are addressed in detail appropriate to all customers using the service. These issues remain the same over time and are hardly subject to change which is why it does not need to be updated religiously from time to time.
  • A customer level SLA, on the other hand, focuses on all the SLM (service level management) issues from a particular group of customers regardless of the type of services being used.
  • The service level SLA covers various SLM issues related to the specific services being offered by the organization while keeping in mind a specific group of customers.


Just like any other document, a Service Level Agreement is also created in line with some important key elements that are discussed in detail below.

1) Description of Services in Detail

The first key element of the SLA is a detailed description of each and every service provided by the organization along with the amount of time the service is available for use. Service availability can easily be measured by the time slot. For instance, a website that makes millions of dollars an hour would require a 99.99% percent uptime during e-commerce hours.

Defining the right metrics is just as important because it helps the organization attain a reasonable performance level. You can set a proper baseline as far as the metrics are concerned from time to time by revisiting and readjusting the settings at a future date.

In short, the main objective of giving a detailed description in the SLA is to identify the scope of all the services in terms of the processes, functions, activities, and projects.

2) Service Standards

Service standards refer to the technical quality in outsourced development. Measurement of technical quality is usually done by using commercial analysis tools which give a clear cut result related to coding defects and program size.

Performance identification of results and outputs expected by the customers from the arrangements also come under service standards and form an important part of the SLA.

3) Duration

It is important that you clearly mention when the agreement begins and ends. The start date of your agreement will enable you to track the IT performance starting from that date and onwards.

You must also take notice on when the agreement ends since you will have to negotiate equipment leases and maintenance contracts on behalf of your customers when you enter in a long-term agreement with other service providers. For instance, if you offer a lease for all the equipment provided to a customer valid for 18 months, you may be able to keep your costs low. But this has a downside too; the customer will no longer be obliged to make payments if the lease ends within 12 months and you will have to resort to another option to fund the equipment lease.

4) Roles and Responsibilities

Managing the expectations of your customers in IT-based service businesses is not easy, and this may turn into a major problem if the roles and responsibilities of both parties are not clearified in the agreement. It is important for you to understand that both, the customers, and the organizations have duties and responsibilities to each other that must be defined properly.

Most responsibilities that you will chart out for your customers will vary according to the type of Corporate IT strategy your organization uses. For example, if you want to keep the costs running low, you may stress the importance of self-service or direct access instead of providing high-tech support 24/7.

Under the roles and responsibilities section, it is also critical that you chart down all the basic requirements that the customers must fulfill; particular PC software purchases and installations. You must also specify that customers must report issues to the organization as soon as they encounter the problem and not days later. This will keep your performance running high, and the customers satisfied with the level of service they receive.

5) The Customer Representative

A customer representative is the person who represents all of your customers when it comes to discussing and negotiating the delivery of IT services. Also, the customer representative is responsible for communicating all the information drafted in the service-level agreement to the customer they represent.

When selecting a customer representative, remember that it is important to pick the right individual who belongs to the mid-level of the organization such as a director in place of a senior officer or vice president because he will have more sound knowledge regarding the business unit’s strategy as compared to anyone else and will be able to understand the concerns his employees may have regarding the IT services provided.

6) Service Level Managers

The Service Level Manager is the one who shoulders the responsibility of managing the IT service level. He or she will have to report to the IT department of the organization as well as the IT customer. The Service Level Manager is also responsible for maintaining, negotiating and reporting against the SLA with the customers.

He or she may also have to meet up with the customers from time to time to discuss the performance or any other service concerns if they arise at any point in time.

7) Evaluation Criteria

In order to determine how well the IT department is doing as far as the performance is concerned, it is important to have some evaluation criteria. Think twice when you select evaluation metrics with your customer representative because you will have to deliver the exact level of service as mentioned in the SLA. Make sure you have all the right tools needed to track what your customer wants before agreeing on the measurement.

One of the mutual benefits of having an SLA is that you and your customer both, will have a fair say when it comes to determining the kind of service they are receiving. Also, having the right objective measurements will eliminate all guesswork and save time and energy too. It is also important for you to understand that your customers may not be satisfied with the level of service provided to them even if you meet all the objectives defined in the SLA.

In order to keep your staff from feeling de-motivated in such a situation, it is important to select the right metrics that are meaningful to the IT service level from your customer’s point of view instead of the organization. You may also have to add specific measurement metrics in line with the unique environment and scope of the business. Take a look at some of the important measures for evaluating the performance of customer service.

  • Abandonment rate refers to the percentage of calls dropped or abandoned while waiting to be answered.
  • The average time it usually takes for a customer’s call to be answered by the service desk is called ASA (Average Speed to Answer). The measurement time is usually measured in seconds.
  • Time Service Factor also known as TSF is the percentage calculation of calls answered within any given timeframe.
  • A percentage of incoming calls that are resolved without a need for any follow-up calls is called First Call Resolution (FCR).
  • The calculation of the time taken to complete a specific task is called Turnaround Time (TAT).

Similar metrics are used to measure hosting performance:

  • A common metric often used for measuring the overall performance of data services like virtual private servers, dedicated servers and shared hosting is Uptime.
  • Mean Time to Recover (MTTR) is used for calculating the average time it takes to recover after a service outage.
  • Defect rate is the measure of errors in major deliverables. Missed deadlines, incomplete backup, and restoration, coded errors and rework are all measured by calculating the defect rate.
  • In outsourced application development, technical quality refers to the measurement of technical quality in terms of coding defects and program size.
  • Application and network security breaches can cost you big-time in these hyper-regulated times. You can measure controllable security metrics like patching and antivirus updates to prevent the chances of a mishap.


An organization cannot have clear goals without having a service level definition and management. Here are the 6 important steps involved in building and supporting a service level model.

Step 1. Analyze Technical Goals and Constraints

The ideal method to come up with realistic and sustainable technical goals is to think of the various requirements and technical goals that could improve the level of service over time. Technical goals usually include throughput, delay, jitter, availability levels, scalability, response time, new application introductions, security, new feature introductions, and cost.

Once all the technical goals are penned down, it is time to look at the constraints and risks involved in achieving the technical goal. Risks and constraints can be categorized mainly in three categories – life cycle practices, current traffic load or application behavior, and network technology, configuration, and resiliency.

Step 2. Determine the Availability Budget

The expected theoretical availability of the network between two defined points is called an availability budget. You can determine the overall availability budget by multiplying the availability for these areas:

  • Hardware availability
  • Software availability
  • Environmental and Power Availability
  • Link or Carrier Failure
  • Network Design
  • User Error and Process

Step 3. Create Application Profiles

Creating an application profile makes it easier for the networking organization to understand and define service level requirements and the network services provided by the organization overall. It also serves as a documented baseline for the network service support and aligns network service goals with the business requirements.

Step 4. Define Availability and Performance Standards

The service expectations for an organization are set by keeping in line with availability and performance standards. These can be determined for different areas of a specific application or network. Now there are several ways you can measure the performance of the service in terms of throughput, overall scalability, and jitter, round-trip delay, and bandwidth commitments.

The organization must further define each of the service standards so that IT groups and customers can have a clear understanding of the level of service being provided to them.

Step 5. Define Network Service

The last step towards basic service level management is defining network service. Network service defines all the proactive and reactive network processes and network management capabilities that can be implemented to achieve service level goals. The network service definition is included in the last document of the agreement called an operations support plan.

The operations support plan must include these three key aspects:

  • Full description of the reactive and proactive processes used to achieve service level goals
  • Management of the service process
  • Measurement of service level goals and service processes

Step 6. Collect Metrics and Monitoring the Service Level Definition

A service level definition on its own is completely useless if the organization fails to collect the metrics and monitor the performance of service being provided to the customers. When preparing a service level agreement, ensure that a proper definition is given as to how the service level will be measured and reported.


Step 1. Meet Prerequisites for SLAs

Experts believe that there are a few prerequisites that the organization must meet in order to come up with a spot on the service level agreement. Here are a few prerequisites you may want to consider.

  • The organization must have a service oriented culture
  • All IT activities must be driven by customer/business initiatives
  • The organization must fully commit to the SLA process and contract

Step 2. Determine the Parties Involved In the SLA

You must choose wisely when it comes to choosing that parties must be involved in the SLA so that it serves as a platform for coming up with new service level goals. Some of the possible goals are:

  • Create and meet reactive support business objectives
  • Provide the top level availability by clearly defining proactive SLAs
  • Promote or sell a service

Step 3. Determine Service Elements

Typical service level agreements usually have many different components including support level, how it will be measured along with overall budget concerns and escalation path for SLA reconciliation. Having proactive service definitions and reactive goals are a must for high availability environments. Some of them are listed below.

  • Onsite support business hours and procedures for off-support timings
  • Priority definitions like the type of problem, maximum time for solving the problem and escalation processes
  • Products and services that are to be supported and ranked in order of business criticality
  • Support level issues pertaining to geographic and business units
  • Goals for improving help desk service
  • Network error detection and service response
  • Funding the implemented SLA
  • Procedures for conflict resolution
  • Measurement of network availability and reporting

Step 4. Define the SLA Required For Each Group

Primary support service level agreements must include functional group representation, critical business units, server operations, networking operations and application support groups. Such groups are usually on business needs as well as the roles they play in the support process.

Step 5. Negotiate the SLA

The last step of creating an SLA is final negotiation and sign off. SLA negotiation is further split up into the following steps.

  • Review of the draft
  • Negotiation of the contents
  • Edit and revise the document
  • Obtain the final approval

Step 6. Measure and Monitor SLA Conformance

To ensure long-term consistency and results, measuring SLA conformance and reporting results becomes an important aspect that one must pay due attention to. It is recommended that all major components of the SLA must be measurable, and a measurement method should be put in place prior to implementing the SLA. Problems can then be solved by holding meetings with the customers quarterly or as and when problems arise.


  • Backbone internet providers usually create and state their service level agreement on their official website.
  • A WSLA or Web Service Level Agreement is usually put in place so as to monitor web services. Authors also get a chance to specify their performance metrics associated with a web service application.
  • Cloud computing services also provide a service based agreement instead of a customer based agreement. It contains important details such as measuring, monitoring and reporting related to the cloud’s performance.

You may not have known much about service level agreements before but now that you have read this article, we hope you are able to understand what an SLA actually is, how it works and best SLA practices that are followed by top IT companies in the world.

Comments are closed.