The biggest challenge with respect to the classical estimate accuracy tunnel lies within the Front-End Loading where early/final capital investment decisions are made usually in a thick cloud of unknowns. Current approaches to estimating focus on a base estimate that typically lacks a standard base estimating strategy, to which a more or less ‘deterministic” contingency is applied – in most cases in an over-arching manner due to residual amount probably already buried in the base. Many contingency quantification methods are generally not based on causal factors.
Organizations generally lack pragmatic cause-and-effect models that relate estimate outcomes to known sources of variation which have been documented through years of research in the field. Fundamental reason for this is generally due to inability of many practitioners to take a system’s view of the estimate and consequently model its relationship. Also, the supporting framework for this approach in terms of historical database application to manage a life-long of project data, understanding of inferential statistics and data analytics is non-existent.
Estimating is both a science and an art. Unfortunately the “art” component is treated by many folks as a subjective guesswork. Management Science methods such as parametric regression, simulation and optimization provide a more rational and systematic approach to eliminating subjectivity bias.
The goal of Estimate Modeling Inc. is to help organizations build practical estimating systems that are utilized consistently within each organization by creating awareness through online publications, training to empower practitioners with pre-requisite knowledge and providing online access to our sample conceptual & semi-detailed estimating tools and databases appropriate for early phase estimates