Service development projects

The most successful projects are those that deliver the service that returns the best value to the business.

Comments

Many IT departments perform software development projects with a focus on delivering software with required functionality by a certain time and for the benefit of business. Unfortunately, this focus for such projects is wrong.

Software delivery is for the purpose of delivering service capability; to deliver a new service to a business’ customers. Therefore, the focus of the project must be on the service capability being delivered.

Service lifecycle

Like projects, a service will progress through its lifecycle. The most successful projects are those that deliver the service that returns the best value to the business.

Source: ITIL Software Asset Management

A new service (or a significant enhancement in an existing service) will go through the concept, development and deployment stages. A service development project will collect the requirements, design the solution, build and procure components required for that service and deploy that service – make it available to be employed by the business.

Up until the service is implemented, a service development project is a cost to the business bottom line. And the return on investment (ROI) only begins when the business operates the service. The longer the service runs in the operations phase, the greater the return for the business.

Do not be afraid to cancel a project that has deviated irreconcilably from the plan

During the development stage, a project must focus on delivering a service that operates to deliver the best long-term value for the business. Does the service deliver the functions requested (utility) and does the service deliver the non-functional requirements of (warranty)? The key element that is often overlooked here is the warranty component. Ask yourself:

Has the service been designed for high availability, easy and low cost maintenance?
Is it compliant with security?
Has it been designed to maximise the value returned to the business?

A project must go through regular review gates during its life cycle. The gates should assess the project for its ability to deliver on its promises. A review of the current scope, costs and timeframe is essential to gauge the viability of the project. Do not be afraid to cancel a project that has deviated irreconcilably from the plan. Applying earned value (EV) measures is a good method for assessing this.

A development project may take one year but the service will most likely be in operation for 10 years.

However, the health of a project during reviews must also be assessed for the viability to deliver an end product – the service – with its state measure of utility and warranty. When assessing the viability, it’s important to compare the service being constructed back to the original business case. Have decisions been made that will mean the end product will or will not return the on-going value required to make a project valuable, to make it deliver the return on investment promised in the business case? A development project may take one year but the service will most likely be in operation for 10 years.

A project delivered on time, under budget and on scope, may still not be a successful one if the resulting service is not cost justified. The on-going cost of operating that service must not only be lower than the revenue returned by the service, it must also be sufficiently lower that it returns a positive return on the project investment, in an acceptable time frame.

Not just one delivery

There will be regular optimisation reviews during the operations of the new service. The business world has evolved since the delivery, or since the last optimisation. The service is likely to require modifications to ensure it continues to deliver the maximum return — either in lowering the costs of providing the service or increasing the revenue the service brings in or both.

Has the service development project delivered a service that is easy to update? Is it easy to implement subsequent releases to the service? There is likely to be many new releases over the life of the service so can they be implemented without any down-time to the service? How easy is it to retrain both business and operations staff? How easy is it to update supporting documentation and supporting tools?

In short, the big question is whether the project planned and being managed will deliver software, or deliver a valued service?

Gary Percival is an IT service manager with more than 30 years' experience in IT. He has worked in operations, application development and support, and well as project management and service management.