Clients get a FalseSenseOfProgress
every time that I show them a beta/mock up of a web site. They think that if they see the pretty pictures that must mean the work is nearly done, when in reality, pretty pictures are all that currently exist.
They then are dissatisfied with progress from there out since to them the demo still looks the same.
The opposite can be true as well. Clients get a false sense of lack
of progress when I take a full-featured application framework and evolve it to meet their needs. They see the lack of trendy widgets or graphic-designer niceties as evidence the work has a long way to go, when really the thing is solid and ready-to-use with some tolerance for the process of fine-tuning.
Therefore, focus attention on customer-chosen features using the UserStory
and its priority. When the story calls for a GUI, always make what's under it real (See DeadlyGuiPrototypes
). When building on a framework, always make your work GUI-visible (See InvisibleFrameworkBuilding?
More generally perhaps, "fake the most consistent velocity at all times". That kinda summarises what we learnt from both kinds of problem in small step delivery for twelve years - and we had plenty of both.
But when I say "fake", I mean in small ways only. The process should be UserStory
driven as has been said. A lot of genuine smoothing comes from small step development driven that way. But we've not been against (or above) arranging things to give the maximum impression of consistency - more radical refactoring to pad out this delivery, one or two least likely exceptions not handled for that one. There is a layer of the process invisible to the customer, quite rightly, and we should use that to give them the maximum sense of security. This more or less is
professionalism as I understand it.
The commercial tensions are inevitably at their greatest at the start, when the big budget has not yet been committed and a "proof of concept" is genuinely valuable for users - but can too easily be sold as an "almost finished" system. In this context I was heartened to hear recently from my accountant of early stage venture capital being raised solely for "Proof of Concept" development. As long as people know substantially what they're paying for, great. They don't want to know more! -- RichardDrake
Joel Spolsky documents this in his article, "The Iceberg Secret, Revealed" [http://www.joelonsoftware.com/articles/fog0000000356.html