I am very happy to undertake an investigation of a system or perhaps to assist in manipulating data while on your site.  These are always interesting challenges.

Most people however come to me wanting a new computer solution based around a database, or an update to an existing system.  

By its nature this sort of work is a project - "to produce a unique and predefined outcome or result at a pre-specified time using predetermined resources".  In other words some reasonably predictable outcome, in a reasonable time using reasonable resources.  A project also should minimise the risks to outcome, resource use and time.

I apply an appropriate level of project management to all software development.  This is often considered a considerable overhead, but Prince 2 has been developed to be completely scalable.  I believe that it is through mastery of the complete method that the work can be done with maximum benefit but minimal overhead.

For a good example of appropriate scaling, download the free software from http://www.projectinabox.org.uk/ and examine the "office move" project.  

Another good reason for using a proven methodology is that it becomes clear when someone is opting for "wishful thinking" instead of tackling the reality of the project.   I ret to develop this clarity through practising the methodology at every opportunity including all solutions I develop for internal use.  There is no project too small for project management.  In fact I find that a small project is more likely to stay small with appropriate Project Management.

There are often two main stages to a development project.  Clearly there is the coding, installation and handover.  At the outset though there is often no clear idea of exactly how the software should perform, so there is a need for a prior stage of requirements analysis.  Usually it is not even clear what the scope of the requirements are, so it is important to have a project initiation stage. 

The project is thus now in three stages and the price and duration of each cannot be known until the previous one has been completed.  Does this expose the customer to unnecessary risk and uncertainty?  Yes there is uncertainty, but it is an unavoidable reality of moving forward.  I aim to reduce this risk by having each stage well defined and pre-priced, and capable of being handed to another supplier, so you are never tied in to CDP.  The final payment for each stage is not due until you are satisfied that the work can indeed be handed over.

Denying the reality of uncertainty I find to be unhelpful.  A fixed price for the whole work serves only to tempt a developer to design a simple-to-build solution that maximises profits but doesn't maximise utility to the customer.

Reducing risk is something that the customer can influence though to quite a degree.  Take up references,  produce coherent briefs,  be involved throughout the project, ask questions about the method,  check the quality of design products, take into account the humanity of the supplier and his or her need to be motivated and make staff available for the analysis stage and testing stage.  Don't worry - I won't expect you to drop everything while the project is underway - I do want your appropriate input though.

One of my highest concerns is with the right level of partnership to allow you to achieve the results that you want.