Improved optimum bid markup estimation through work load related bid distribution functions

Research literature and professional journals were reviewed to determine the present state of knowledge of the use of bid models by construction contractors. The Friedman and Gates models were explained in detail because of their potential significance to the construction industry. A construction company's actual three year bid results were compared to the results they would have achieved during the same period of time if they had used the Friedman and Gates bid models. The primary difference between the process of estimating the cost of building projects, and estimating for heavy and highway projects where estimates are both equipment heavy and highly production rate dependent, was explained. Contractors' actual bid markups on a large highway construction program were used to show that the bid markups have less variance than cost estimates. Contractors involved in earth moving and other equipment heavy work tend to mark up their bids a consistent amount, but change their estimated production rates when they need work. Three years of a competitor firm's bids were analyzed to show that his bid distribution functions vary as his work load varies, and a method was proposed to develop work load dependent bid distribution functions for bidding against this competitor. The assumption that competitors will have the same cost estimate, and will vary their markup, was challenged. Although this assumption has been a part of all previous bid models, for a significant part of the construction industry it is not an appropriate assumption. Cost estimates are lowered by the use of more efficient production rates when heavy construction contractors need work. A model was presented in the form $F(r)$ = $1-e\sp{-a(r-r\sb0)\sp{m}}$ where a and m are Weibull parameters, to fit a theoretical bid distribution function to a competitor's actual bid distribution function. The theoretical bid distribution function was then used to develop three separate work load related distributions for a known competitor. The purpose of developing three separate distributions is to provide a method for a contractor to utilize the bid distribution function that best fits a competitor's current business situation at the time a bid is submitted. Finally, a Utility Model was developed to demonstrate that competitors change their bid objectives as their work load changes. Using the theoretical distributions developed earlier in the paper, a competitor's three work load related utility functions were computed. This method to define the utility function of a competitor will permit a contractor to improve his ability to evaluate the competitor's bid strategy.