in any project, you can get some pretty reliable estimates from this method alone. In fact, in my experience as both a programmer and the CEO of a software company, I have found it to be incredibly accurate and useful.
A second way to triangulate this project estimate is to ask experts in the area for their opinions–hopefully they will be better at project estimation than my coworker and I were that first time. A third way is to select an appropriate metric for estimation. For example, one could use line of code counts or function points in estimating the length and scope of software projects. For architecture projects, you might use number of pages in the drawings or square feet planned as similar analogies. Every project has some gross measure of its size that is available at the outset and can be used to plan the overall project in addition to this method I've described of tracking time against the earliest phases.
So back to the story. We really blew it on estimating and bidding on that first project for Tivoli, but when the next one came around, we had data on the portion of the overall project that the requirements phase had taken up. This allowed us to use Grady’s ratio to predict overall project size, and we found that on this second project, we came up with a very accurate project estimate. This worked very well for all of the subsequent fixed-cost consulting work we did for Tivoli.
Partially due to the strength of the solution and how well it ran on IBM's AIX operating system, Tivoli was able to eventually sell their company to IBM for 743 million dollars in 1995.
For a consultancy that is doing fixed-cost projects, this concept of using the standard ratio of requirements phase to overall project length is a very powerful project estimation technique. It can eliminate erroneous bidding and its resulting costs, which is a major concern for such companies.
Overbidding on a consulting job means that you won't get the work in the first place, because the potential customer will give it to your competitor at a cheaper price. Underbidding, however, means you will win the deal and lose money. Neither situation is acceptable for businesses today, and yet most consultancies do a poor job in this area. One way to make more precise bids is to use a key performance indicator, which is a tool used to measure progress towards a strategic business goal. For example, the number you want to minimize in this situation is defined by the formula [(E-A)/E], where:
E = estimated hours to complete the project
A = actual hours spent to complete the project
It is important to keep this KPI value as close to zero as possible, which indicates that you are bidding on projects more accurately.
Just tracking this number is a great first step towards better bidding, and you can get the necessary data to calculate it from any timesheet system, including a paper one. Automated timesheet systems, however, are generally even more effective in this area because they often have reports to calculate the KPI figure for you.
Improving adherence to your estimate can be difficult for some companies until they understand the ratio concept described above. An example of this is illustrated in the following diagram, which shows how the formula can work for any business. Your company's magic number may not be 6-8% like Grady's, but once you determine your own ratio for specification to total project length, you can use it again and