The bidding process involves two crucial decisions:
first, whether or not to bid for a project and, second, the
determination of the bid price (Shash, 1993). While the
latter has been subject to intense research interest (for
example, Friedman, 1956; Gates, 1967; Willenbrock,
1972; Fine, 1975; Carr, 1982; Moselhi
Dozzi and AbouRizk, 1996; Li and Love, 1999; Chua
and Li, 2000; Dulaimi and Shan, 2002; Marzouk and
Moselhi, 2003), there has been comparatively little in
the way of objective research into the former.
The bid/no-bid decision is both complex and dy-
namic, involving many factors (Shash, 1993), while the
selection of the most appropriate projects for which to
bid is fundamental to a successful commercial strategy.
Moreover, the decision to bid, as with that of determin-
ing the project mark-up, is very important as success
or failure of a contractor’s business lies in the outcome
derived from those decisions. What evidence there is,
however, suggests that this decision is usually deter-
mined by subjective rather than objective information
(Fellows and Langford, 1980; Ahmad and Minkarah,
1988; Shash, 1998). Improvement in the contractor’s
selection of projects would give significant benefit to the
construction industry and consequently to its clients.
Further, a suitable decision support model can be a stra-
tegic tool in determining the most appropriate projects
to seek and for which to submit a bid.
The aim of the paper is to:
investigate and analyse the factors that influence
the decision to bid process; and
model the decision to bid process and in doing so,
to distinguish relationships within the decision
Previous research into the decision to bid
Construction organizations are required to be selective,
choosing which work they will tender for from a
Construction Management and Economics
A logistic regression approach to modelling the
contractor’s decision to bid
DAVID J. LOWE
and JAMSHID PARVAR
Project Management Division, Manchester Centre for Civil and Construction Engineering, UMIST, Manchester
M60 1QD, UK
Received 6 June 2003; accepted 17 November 2003
Significant factors in the decision to bid process are identified and a pro-forma to elicit a numerical assessment of
these factors is developed and validated using the bid/no-bid decision-makers from a UK construction company.
Using the pro-forma, data were collected from the collaborating company for historical bid opportunities.
Statistical techniques are used to gain a better understanding of the data characteristics and to model the process.
Eight variables have a significant relationship with the decision to bid outcome and for which the decision-
makers are able to discriminate. Factor analysis is used to identify the underlying dimensions of the pro-forma
and to validate functional decomposition of the factors. Finally, two logistic regression models of the decision to
bid process are developed. While one model is ultimately rejected, the selected model is capable of classifying the
total sample with an overall predictive accuracy rate of 94.8%. The results, therefore, demonstrate that the