Accurate Estimation
Accurate Estimation: The Key to Successful Project Planning & Scheduling
Many organizations do not have the experience or skills necessary to estimate project tasks well and as a result are unable to manage their project's time. This will ultimately lead to project failure.
Based upon extensive experience managing projects large and small, VP has developed an estimation model that has proven to be reliable and accurate.
The VP Estimation Model™
The basis of the model is the insight that effort estimates are proportional to the complexity of the work to be accomplished. Here is how it would be applied to develop a Rough Order of Magnitude estimate during the planning phase of the project.
- Identify all the units of work in the project, e.g. use cases
- Identify what drives the complexity of the work
- Break each complexity driver into four levels of effort
- Define each level of effort as objectively as possible
- Go through each unit of work, assigning a level of difficulty score based on the definitions in (4)
- Calibrate the difficulty scores to actual durations using historical data, where available
When no historical data is available, calibration can be performed by running the model on an already completed set of tasks from a similar project. If this doesn't exist then a session can be conducted to reach a consensus of the calibration values.
Case Study
Large Global Healthcare Provider
This operator of dialysis clinics embarked upon a sweeping technology strategy that involved the procurement, customization, and implementation of a suite of clinical and revenue mangement health care services from a well-known solution provider.
Valenti Partners was responsible for managing the integration of the data services project which during the planning phase, required the development of a Rough Order of Magnitude (ROM) estimate.
Using the VP Estimation Model, the complexity drivers of each use case were identified and then assigned diffculty scores. Since the client had no previous data against which to calibrate the scores with actual durations, VP held brainstorming sessions with the development team.
Based on the actual delivery schedule, the ROM estimates proved to be quite accurate, highly unsusal for a high-level estimate. This reflected the soundness of the model and the process by which the developers assumed ownership for the results.
The model was used successfully duing subsequent phases, now using historical data from the initial use of the model.