Performance Motors has just launched a new model that is being produced at a rate of 10,000 vehicles per month (and is flying off the dealers’ lots as fast as they arrive). You are one month into customer sales (10,000 units sold and delivered) and already there are some reports of a specific failure that impacts safety. NHTSA has opened a preliminary investigation, and your engineers are participating in the analysis. It is expected that the results will be known in one month. At that time, the findings will be:

 a) that there is no defect (just some bad luck and statistical outliers), or 

b) that there is a quality defect, meaning that a batch of some component was not manufactured correctly, or 

c) that there is a design defect requiring a redesign and replacement of the part. If there is no defect, then there will be no further action and no additional costs incurred. If a quality lapse is pinpointed, several scenarios are possible, but the most likely case is that it would be confined to one of the four weekly batches of the component that went into the first month’s production (2,500 vehicles, all of which are now in customer hands). NHTSA would require that those vehicles be recalled immediately but at least it is known which vehicles received which parts. On the other hand, if a design defect is found, then all cars will be recalled, and no more cars would be sold until they are fixed with an approved solution. As a precautionary measure, a crash redesign of the component could be started immediately and would take one month and cost $3M. Because the key design features are embedded in firmware, replacement parts could be available immediately upon completion of the redesign. Your engineers estimate that the likelihoods of the possible findings are: 20% that there is no defect, 40% that one batch will be affected, and 40% that all parts must be recalled and replaced with a new design.

You have two key decisions to make right now; one is whether to keep selling vehicles or not, and the other is whether to start a redesign right away. Stopping sales is not expected to affect demand in the long run, but it does delay cash flow, complicates logistics and is not good for customer satisfaction; a stoppage is estimated to cost $1M/month. Production would not stop – it is much cheaper to repair already-built units than to stop and re-start the line. The product redesign might ultimately prove to be unnecessary, but it would ensure that replacement parts would be available if NHTSA ordered a full recall. If a full recall were ordered, and the redesign had not been done, it would have to be started at that time, at the same cost. Sales would have to stop for a month because of the recall, at the same cost/month as above.

Recalling a vehicle from the field to have it repaired at a dealer costs $5,000, if repair parts are available   If parts are not available, there would be a 1-month wait and the owners would be provided a loaner for the month so the cost would be $7,000 per vehicle. Repairing a vehicle that is still at the factory costs $500. Currently there are 10,000 vehicles in customers’ hands, and in the coming month there will be 10,000 more vehicles coming off the production line at the factory. If you do not stop sales, in one month there will be 20,000 vehicles in customers’ hands, whereas if you do stop sales, there will be the same 10,000 with customers, and 10,000 parked at the factory. If there is an additional month delay at that point (a recall is ordered, but no parts are available) then there would be an additional 10,000 vehicles parked at the factory that would also need to be repaired.

What are you going to do?

Prepare an executive summary report that summarizes your recommendations regarding the best strategy to pursue. Your report should be a Word document and the first page should simply be a cover sheet with a title, date, team members’ names and any other identifying information you think necessary. The second page should be the executive summary report. You have one page to tell the company executives everything they need to know about the situation. It is important that you not only recommend the preferred strategy on an expected value basis, but also include a discussion of the risks associated with your recommendation, and how they compare to other choices that might be made. You would not be out of place to mention some of the qualitative factors influencing the decision (public relations, ethics, etc.), but you are not in a position to offer any recommendation on those points, so don’t get carried away on this.

Your report must be accompanied by backup materials showing how you arrived at your recommendations. How you show this is up to you, but it should be clear enough to stand on its own. The decision tree on this is not very complicated – just four branches, each having three possible event outcomes, so you can just do a set of four payoff tables and as long as they are well-labeled there is no real need to draw the tree although you are more than welcome to if you wish. Calculating the payoffs is a little more involved and requires that you think through the sequences of events carefully!