Online Extra: What a What-If? Agreement Looks Like

Many things can go wrong with Web service. Here's a sample of one provider's contract, which tries to cover all bases

Thinking about a service provider to host your company's Web site or software? Reputable companies will ask you to sign a service-level agreement. An SLA outlines the responsibility of your business and its service provider should your Web site or software become unavailable or begin functioning poorly. It should explain, for example, what sort of remuneration your business will receive if your customers can't log on to your site or your employees have difficulty executing day-to-day online routines.

Below is a sample SLA offered by Digital Fuel Technologies Inc., a Brisbane (Calif.) company that tracks SLA compliance for a fee. Typically, SLAs are far more detailed than the one offered below. As such, consider it a foundation for building a more comprehensive agreement.


This document defines the scope of the Application Service offering to be delivered by the Service Provider to the Client, service level criteria, procedures for modifying or terminating the service, and Client remedies in the case that service level criteria are not adequately met.

Description of Services

The Client's Application Service consists of:

* Application <<define application>> hosting

* Application data storage up to XX <<insert appropriate measure>>

* Application feature one includes <<insert description>>. Restrictions include...

* Etc.

Description of Service Environment

Service Provider platform (include 3rd party hosting providers, ISPs, operating environment, database environment, disaster recovery procedures, etc.)

Description of User Environment

* Number of users

* Geographic location(s) of users

* Technical prerequisites (i.e., browser type, operating environment, hardware requirements, network connectivity, etc.) for users to access the Application Service

Measurement Conventions

This SLA uses the following conventions to measure times and percents. <<Insert time formats, time zones, business hours, schedule of operations, etc.>>

Monitoring and Management

The following systems and services will be used to monitor the Application Service. These systems and services will provide the data by which the Service Provider's compliance with Service Level Objectives may be evaluated. <<Insert a list of monitoring tools and other systems required for tracking SLOs.>>

Service Level Objectives (SLOs)

Service Availability

Defined as the percentage of time that the Application is available and key <<insert list>> functionality is accessible during specified operating hours.

Notification of Outages and Service Interruption Events

Defined as XX time between an outage is detected and the Customer is notified. Includes a definition of outage.

Backups and Restores

Defined as the maximum amount of time from backup initiation through backup completion.

Security/Application Access

Defined as maximum number of simultaneous Application logins from defined geographical regions.

Security/Virus Checks

Defined to include to metrics and the interval at which virus checks are performed within the Client's hosted environment and the interval at which the Service Provider updates Virus Check software.

Request for Service Modification

Additional Application Resources Available

Defined as the time from when a customer requests an additional number of Application Service users until those users are actually added to the customer's Application Service and verified as "available."

Additional Data Storage Resources Available

Defined as the amount of time from when a customer requests the addition of a predefined amount of Application data storage space to their environment until the time when the additional storage becomes available.

Dispute Resolution and Mediation

Both parties agree that, in the case of dispute, XX <<describe entity-third party or mediation board>> will provide mediation services. Both parties agree that the decision of this entity will be binding.


If Client is unable to access the Application because the Service Provider failed to provide services for more than XX consecutive minutes, the provider will credit Client's account the credit of Application service charges for XX days of service in any XX calendar month.


If, over any XX time period, XX outages cause the client to seek XX or more days of credit, then client may terminate the applicable Client Service Order for cause and without penalty by notifying the service provider within XX days following the designated time period. Such termination will be effective XX days after receipt of such notice by the Service Provider.

Before it's here, it's on the Bloomberg Terminal.