...for the public sector.
Spend less time specifying and more time making stuff.
The big boys will blow any budget you have with expensive consultants who produce little. And then try to get the cash for development (or configuration as they will want to call it) from change control because they know they underestimated. And putting another layer between you and the coalface can only cost money.
Contract the team yourself. Write proper use cases and an FRS that specifies forms, processes and data at the field level. Get a estimates from the person who is going to be the technical lead and don't try to horsetrade to get the price down. He has nothing to gain by inflating it and you both look good if it comes in early and under budget.
Stop changing your mind. Changes to your signed off FRS have untold consequences.
Get coders making stuff as soon as there is an FRS. If you feel you need a detailed design write it as the coding is happening. It's the only way it will reflect reality.
If you have an existing apps platform make sure it can do 100% of what you want it to. If you do the 80:20 rule plan on the 20 to take 90% of the time not 20%. Working around COTS and trying to bend them to you will is pain for everyone.
If the coders think there is an application or tool that will help them get it for them.
If the coders say they are struggling with an application in the platform give them the freedom to look at a different way.
Have Boardroom Pilots every week. Have the Tech Lead justify everything you see against the FRS. Tick off use cases to measure progress.
Listen to the people who are making it. If they offer different ways to do things to make implementation easier and quicker consider them. The only reason not to do things a different way is if they don't meet the requirements. In short let the technical people decide the technical issues.
Most of these projects are a bit of web stuff, some workflow and some data. It's not hard.
My current contract ends at the end of the month...