IT Cost Drivers
Tips to Reduce the IT Costs for Programs / Projects
The following is a list of a number of things which can be cost drivers for IT projects:
- The number of systems touched
- Quality of Requirements Documentation
- Requirements Capture and Solution Design process
- Document review process
- Extensive use of vendor resources
- Fear culture causing risk averse behaviour
- Engaging with different business models
- Approach to package implementations
- No re-use of Testing IP (scripts etc)
The number of systems touched
-
Makes Design more complex and expensive
-
Makes Development more difficult and expensive
-
Makes organisation rely heavily on experts
-
Makes Testing more complex and expensive
Solution
-
Implement strategy to reduce number of systems – NOTE – this is beyond the control of individual projects or programs but may be addressed as a strategic driver for the organisation
Quality of Requirements Documentation
- Text based requirement too easily misinterpreted (by users and developers)
- Not based on data dictionary – ambiguous
- Result is requirements are not locked down until far too late with major scope management issues late in projects when burn-rate is highest
Solution
- Screen and Report mock-ups included in requirements
- Use of Requirements Management Tools eg., Telelogic DOORS
Requirements Capture and Solution Design process
- Common approach is to produce a 2 pass approach – a signed-off High Level document followed by a Detailed document
- Sometimes only a High Level document is produced
Solution
- Need detailed Requirements Capture to be very unambiguous and succinct
- Need detailed Requirements to be documented as quickly as possible
- Need to map Solution Design to the Requirements document at individual requirement and component levels to ensure optimal design and development effort
- Use of Prototyping to capture requirements and solution – this allows users to get an early view of proposed solution and maximises their engagement
Document review process
-
Common approach is to have a sign-off process which involves multiple reviews / iterations
Solution
- Work-shopping is recommended wherever possible as it:
- allows reviewers to interact and share knowledge
- minimises iterations
- reduces time and cost of review process
Extensive use of vendor resources
- Usually, vendor resources are more expensive than internal resources
- Vendors retain a significant part of IP making it difficult to negotiate contracts
- There is a significant administrative overhead with using vendors because of the contract management issues
SolutioN
- In-source
- internal resources are cheaper
- this also retains IP within the company so that future development is also cheaper
- Negotiate better deals by proposing larger deals for each project if necessary outsourcing the whole project to vendors
- Negotiate larger organisation-wide outsource model – this is not something that can be managed by the project but can be addressed as an organisation strategy
Fear Culture Causes Risk Adverse Behavior
-
Fear of failure and the adverse reaction to failure causes people to choose safe options which may not be best for the organisation in the long-term
-
Projects which should be stopped are sometimes continued because of the repercussions associated with stopping them
Solution
-
Change the culture – there is much to be learnt from mistakes – ensure the learnings are banked
-
Reward people who propose that non-viable projects be stopped in the same way successful projects are rewarded
Engaging with different business models
- There is a cost in having multiple engagement models with various business units
- There should be a generic business engagement for IT / project resources to ensure that there is one set of processes
- Multiple engagement models make training and re-deployment of staff very difficult
Solution
-
Develop a standard engagement model with all business units – this is not something that an individual project can implement but is an IT-wide strategic change
Approach to package solutions
- The common approach is to start with a ‘blue-sky’ or wish-list requirements
- This approach means there is an expectation management issue when package is reviewed
- Usually many modifications are proposed
- Changing packages is a very substantial decision with often very serious ramifications to not only the project concerned but can cause ongoing issues with all future upgrades and interfaces to the package
Solution
- Use a gap-analysis approach whereby the package is reviewed and a gap analysis is documented
- The potential gaps then need to be investigated to determine if work-arounds can be found by changing or adapting current business process
- Packages should only be modified as a last resort
No re-use of Testing IP (scripts etc)
- The development of test scripts and test environments on a project by project basis
- This approach means there is no leverage on previously developed testing scripts
- There is a significant cost and lead time in setting up test environments
Solution
- Develop a test centre approach able to utilise previously developed knowledge and test processes, scripts and environments
The following is a list of the possible outcomes of proposed changes:
- Reduced re-work caused by scope change
- Reduced costs for Testing
- Reduced costs for Development
- Reduced project delivery times
- Improved perception of IT delivery
- Reduced expectation management and scope management issues