Scope creep can eat any development team alive. Understanding the cone of uncertainty helps to nip scope creep in the bud.

Initial scope on a project will land in the range of 1/4 to 4 times actual scope. Interestingly, you can actually guess where someone will scope a project based on their role in the company (Sales the highest, CEO’s the lowest, Dev just below accurate and Project managers about exactly right). Tidbits aside, this principal demonstrates that scope creep may result from simple poor scoping. Similarly, poor scoping may result in sticker shock to the client due to a wildly inflated price.
The Cone Of Uncertainty

A well developed plan dissolves the cone of uncertainty.

Effectively applying something like the Unified Process starts the journey. Adding the MoSCoW approach and regularly writing SOW’s will render the ability to accurately scope a project. Once a solid scope has been created, the entire team must understand it, and understand their role in limiting scope creep.

The greatest scope in the world falls to pieces if no one stays true to it. If time starts to slip away, either cost goes up, or quality goes down, and neither scenario works for me. Proper scope should include flex time and should be built to adjust on the fly as a project deploys in order to maintain cost, and quality as time varies.