PurposeThe ability to measure and assess “quality” is essential in building and maintaining a safe and effective transportation system. Attaining acceptable quality outcomes in transportation projects has been a reoccurring problem at both the federal and state levels, at least partially, as a result of poorly developed, inefficient or nonexistent quality assurance/quality control (QA/QC) processes. The purpose of this paper is to develop and implement a new QA/QC process that focuses on a novel double-bounded performance-related specification (PRS) and corresponding pay factor policy that includes both lower and upper quality acceptance and payment reward boundaries for bridge concrete.Design/methodology/approachThe authors use historical data to design different payment scenarios illustrating likely industry responses to the new PRS, and select the single scenario that best balances risk between the agency and industry. The authors then convert that payment scenario to a pay factor schedule using a search heuristic and determine statistical compliance with the PRS using percent-within-limits (PWL).FindingsThe methodology offers an innovative approach for developing an initial set of pay factors when lifecycle cost data are lacking and the PRS are new or modified. An important finding is that, with a double-bounded PRS, it is not possible to represent pay factors using the simplified table PWL currently employed in practice because each PWL value occupies two separate positions in the payment structure – one above the design target and one below it. Therefore, a more detailed set of pay factors must be employed which explicitly specify the mean sample value and the design target. The approach is demonstrated in practice for the Agency of Transportation in state of Vermont.Research limitations/implicationsThe authors demonstrate a novel approach for developing a double-bounded PRS and introduce a payment incentive/disincentive policy with the goal of improving total product quality. The new pay factor policy includes both a payment penalty below the contracted price for failing to meet a specified performance criterion as well as a payment premium above the contracted price that increases as the sample product specification approaches an “ideal” design value. The PRS includes both an upper and lower acceptance boundary for the finished product as opposed to only a lower tail acceptance boundary, which is the traditional approach.Practical implicationsThe authors illustrate a research collaboration between academia and a state agency that highlights the role academic research can play in advancing quality management practices. The study involves the use of actual product performance data and is operational as opposed to conceptual in nature. Finally, the authors offer important practical insights and guidance by demonstrating how a new PRS and pay factor policy can be developed without the use of site-specific historical lifecycle cost (LCC) data that include detailed manufacturing, producing and placement cost data, as data related to product performance over time. This is an important contribution, as the development and implementation of pay factor policies typically involve the use of historical LCC data. However, in many cases, these data are not available or may be incomplete.Social implicationsWith the new PRS and pay factor schedule, the Agency expects shrinkage and cracking on bridge decks to decrease along with overall maintenance and rehabilitation costs. A major focus the new PRS is to actively involve industry partners in quality improvement efforts.Originality/valueThe authors focus on a major modification to an existing QA/QC process that involves the development of a new PRS and an associated pay factor policy undertaken by the Vermont Agency of Transportation. The authors use empirical data to develop a novel double bounded PRS and payment schedule for concrete and offer unique operational/practical insight and guidance by demonstrating how a new PRS and pay factor policy can be developed without the use of site-specific historical LCC. Typically, PRS for in-place concrete have only a lower tail acceptance boundary.