The Issue:

Under-resourcing is the #1 killer of projects.  Research shows that promising and attempting to do too much work in too little time results in extreme inefficiencies in the use of staff.  When project inevitably slip as they approach the original commitment date, the temptation is to "throw bodies" at the problem.  This enormously increases burn rate and has a huge negative effect on quality, but it doesn't usually help with schedule.

According to a 2003 study by Computer Weekly in the UK, only 16% of IT projects achieve their goals.  Numerous studies in the US confirm this approximately 20% success rate.  According to Simula Research Lab, only one in seven companies consider their estimation process to be "adequate".

The Solution:

Corvus has focused on project estimation as a specialty for over eight years.

When we look at estimation problems within companies, we see them in the context of the overall environment--companies may fail to meet their commitments because of technical issues or organizational issues.  They may be producing accurate estimates, but not using them correctly.  They may perform consistently and predictably in one lifecycle phase, but not in another.  It may be that the organization simply needs some tune up of its existing processes or it may simply be that a combination of factors makes the development process so unpredictable that no estimation process can work until the development environment is stabilized.

With our holistic view, we look at all aspects of the business practice that affect an organization's ability to make and keep reasonable commitments that achieve their business goals.  We do this by:

  Analyze the existing problems in estimation across the organization
Knowing the interplay of the business pressures and practices, technical development capabilities, and interpersonal and managerial approaches, allows companies to solve the problem, not just the symptoms that are often manifest as overrunning projects..
  Develop detailed calibrations of development capability
Using both proprietary estimation tools and capabilities we have built ourselves, we can perform detailed quantitative calibrations of just how the organization actually works.  These are calibrations are similar to medical diagnostics and help the organization both understand just what its productivity levels are, where the productivity may be compromised and, just as importantly, what the natural variance in performance is.  In establishing this level of quantitative analysis, companies find themselves already operating, at least in part, as a CMMI L4 organization.
  Establish predictive models
Using the calibrations, Corvus helps companies build predictive models that indicate, given specific situations, just how they are likely to perform.  This not only stabilizes the estimation process, it also allows our clients to quantitatively calculate just how much risk they are carrying.
  What-if? and close-in decision support of business practice
With calibrated quantitative models, companies can use this capability as a "business simulator" that allows prediction of likely results of decisions before executing the decisions on the project.  The value of this capability is literally enormous.

 

Home      About Corvus       Contact Us       Articles/Resources       Clients          Affiliations  

Copyright 2009 Corvus International Inc.  All Rights Reserved

 

Estimate Your Projects, Manage Your Portfolio...

Optimize Decision-Making and Communication

 
Case Study

Consulting Company

Corvus was contacted by the General Manager of a medium-sized software consulting company.  While the company has a good reputation and has excellent staff, they consistently overran on their project budgets and schedule commitments.  

Analysis: After conducting many interviews across and up and down the organization, we identified the key issues:
(a) A focus on "bid-to-win" commitments where the actual capability of the organization was ignored in an attempt to win business at any cost.
(b) Ad hoc processes to collect data and create estimate.
(c) No separation between the technical activity of estimation and the business activity of making commitments based on the estimates.
(d)  Arbitrary overruling of the estimates by senior executives based on (a).

Approach: Starting with a business process re-engineering approach focusing on the RFP response activity, we redesigned the data collection and estimation process.  The processes were written to conform to CMMI and PMBOK standards.  A proprietary estimation tool was installed and calibrated against the work products and existing estimates reassessed.  In addition to the calibration, we also implemented way of calculating quantitative risk.

Results: Estimates for existing projects were brought much closer in line with observed progress.  A major project was completely re-estimated resulting in much more efficient use of resources.  A benchmark of quality results from an offshore programming house resulting in a re-assessment of the contract.  Baseline RFP responses came much closer to actual costs resulting in a near elimination of cost-price imbalance.