This study employed a Draper-Lin small-composite response surface design to determine the optimal dimensional-factor levels to minimize the door-closing effort for a new automobile project. It was found that the factor levels that minimize the closing effort did not lead to robust manufacturing conditions. Likewise, the factor levels needed to produce a robust process did not lead to a sufficiently low closing effort. Various compromises were considered; namely, the minimum closing effort obtainable subject to selected semirobust conditions. The net result was that it was not possible to obtain sufficiently robust factor conditions, while achieving a suitably low value for closing effort.