Appendix 2: Product Upgrades Based
on Minimum Expected Quality Loss
Introduction (Langford 2009)
Maintenance and sustainment costs are typically one-third of the development costs, as was briefed to the Government Accountability Office
(Chaplain 2008) for the Space Shuttle, to 70% of the lifecycle costs for the
general category of software (Boehm and Basili 2001). Often, managers
responsible for maintenance and sustainment target costs reductions of the
order of 15–20% to improve product profitability. Perhaps such actions
assume that customers are pleased by both the gesture to reduce costs and
the company’s interests in supporting fielded products. However, for customers, perhaps the most meaningful consideration of continued product
support is lower cost of ownership.
The authors posit that the upgrade cycle for fielded products could be
based on the expected quality loss that results from the period of the upgrade.
The consequence would be a Pareto-efficient determination of the upgrade
period. To achieve Pareto-efficiency (based on the principle that one-sided
benefit to a party to a negotiation results in an inequitable distribution of
losses), losses for all stakeholders must be considered and incorporated into
a cooperative exchange of benefits and losses. A common distinction between
the interests of stakeholders can be depicted graphically as leaning toward
either smaller or larger than some position that will eventually be the negotiated settlement. That is, the agreement between two stakeholders is defined
as the position whereby neither side to a negotiation has an unfair or disproportionate advantage. For the purpose of this paper, the mathematics
simplifies by assuming an idealized negotiation (Figure A2.1) where two
parties incur equal losses about a center point target value, m.
The minimum loss depicted as the quality loss function in Figure A2.1
defines the target value of the critical performance characteristic, m, as a negotiation between two strategies with opposite demands on quality for a given
investment. One party to the negotiation determines that more performance
is better (considered as larger-the-better (LTB) strategy) while the other party
considers that smaller-the-better (STB) demands on performance is required.
335
on Minimum Expected Quality Loss
Introduction (Langford 2009)
Maintenance and sustainment costs are typically one-third of the development costs, as was briefed to the Government Accountability Office
(Chaplain 2008) for the Space Shuttle, to 70% of the lifecycle costs for the
general category of software (Boehm and Basili 2001). Often, managers
responsible for maintenance and sustainment target costs reductions of the
order of 15–20% to improve product profitability. Perhaps such actions
assume that customers are pleased by both the gesture to reduce costs and
the company’s interests in supporting fielded products. However, for customers, perhaps the most meaningful consideration of continued product
support is lower cost of ownership.
The authors posit that the upgrade cycle for fielded products could be
based on the expected quality loss that results from the period of the upgrade.
The consequence would be a Pareto-efficient determination of the upgrade
period. To achieve Pareto-efficiency (based on the principle that one-sided
benefit to a party to a negotiation results in an inequitable distribution of
losses), losses for all stakeholders must be considered and incorporated into
a cooperative exchange of benefits and losses. A common distinction between
the interests of stakeholders can be depicted graphically as leaning toward
either smaller or larger than some position that will eventually be the negotiated settlement. That is, the agreement between two stakeholders is defined
as the position whereby neither side to a negotiation has an unfair or disproportionate advantage. For the purpose of this paper, the mathematics
simplifies by assuming an idealized negotiation (Figure A2.1) where two
parties incur equal losses about a center point target value, m.
The minimum loss depicted as the quality loss function in Figure A2.1
defines the target value of the critical performance characteristic, m, as a negotiation between two strategies with opposite demands on quality for a given
investment. One party to the negotiation determines that more performance
is better (considered as larger-the-better (LTB) strategy) while the other party
considers that smaller-the-better (STB) demands on performance is required.
335