SDLC - Implement the Solution

This stage "rolls out" the solution to the people who will be using it. It is necessary to:

• train users - who needs training on what topics? How, where and when will it be done? Read up on Training here.
• install the system and bring it online, using direct, phased, parallel and/or pilot changeover schemes.
• use the solution for real, and monitor its performance

There are various implementation strategies you can choose from depending on the circumstances...

System Implementation ("Rollout" or "Cutover") Methods

Changing over from the old to the new system can be anything from trivial to traumatic, depending on the size of the visible differences between the two systems.

There are four main styles of changing from system to system: the first 2 are complete strategies; the final 2 are optional accessories.

Remember - if you choose pilot and/or parallel, you still need to use direct or phased implementation to finish the job!

Direct conversion:

The old system is stopped and the new system replaces it immediately. It is a quick transition, which may be unnerving if the changes are large. If used, it should not be done in a peak period where transition "hiccups" will upset the organisation.

Phased conversion:

If the new system has several components, they can be introduced one at a time. For example, an organisation introducing modern communications may first introduce the internet, then email, then videoconferencing, then electronic ordering. This takes more time, but each small step is less traumatic than one huge one, and staff can become accustomed to one change before facing the next. Also, problems with any step can be detected and fixed before the next step is taken. Phased conversion is impossible if the system is "one piece", for example when introducing new hard disks you can't add a little piece of a hard disk at a time: it has to be "all or nothing."

Pilot scheme:

This is not an alternative to the first 2 methods: it is an optional extra. If the organisation has several branches or departments where the new system will be implemented, it may decide to trial the new system in one location first and see how it goes. Any faults and problems will be limited to that one location and will not cripple the whole organisation. Pilot conversion may be impossible if the introduction of the new system in the test location makes it incompatible with other unchanged locations. (e.g. if every branch used MS Word and the pilot branch converted to Brand X Word Processor, it might not be able to share documents with the other branches or head office.)

Parallel operation:

This is not an alternative to the first 2 methods: it is an optional extra. If it is possible to keep the old system in place and running while the new system is installed, you gain certain benefits.

• You can directly compare the effectiveness and efficiency of the new and old systems
• If the new system fails, the old system is still there chugging away as normal so no harm is done.

Parallel conversion is only possible if the old and new systems are completely independent. It is impossible to achieve if the new system is built on the old system.

Human, technical and procedural requirements to be considered when implementing changes to an information system

• HUMAN
  • staff training
  • acceptance by users
  • health and safety of users
  • effects on job responsibilities - retraining may be required
  • redundancy
• TECHNICAL
  • ongoing technical support
  • compatibility with existing equipment
  • compatibility with corporate partners
• PROCEDURAL
  • adaptation or creation of workplace procedures
  • disruption to normal operations
  • changes to org charts and data flow diagrams