Design and Performance Specifications

One of my PM class lectures discussed design specifications and that the buyer assumes the risk because they have specified the equipment. On the other side, if the buyer defines the performance desired, the contractor assumes more risk. An IT example was given where all of the IT parts were specified, but perhaps the system as a whole didn’t work well because of unforeseen conflicts.
I mostly receive a combination of the two when I need to provide pricing. The client may provide a variety of design specifications (almost always in terms of how something should look, unless we are doing the design, and sometimes with actual pieces of equipment), but also include the performance expectations. Awareness of performance expectations is inescapable since we are often providing labor to make an event happen or to install a working exhibit.
Almost always in most RFP or RFQ’s is a statement somewhere that despite any owner suggested products, it is the shop that is responsible for fully meeting performance expectations.
First, this can be tricky, because sometimes it isn’t evident that something that was owner specified won’t work until significantly into the process. At that point, a lot of various resources have been consumed. Many times when this happened it was a project that has a decent portion of R&D. When I worked on the Sunlight part of the Science Storms at the Museum of Science and Industry, the owner specified a specific light fixture hung in the ceiling to focus on solar panels on the main floor. In the advent that the day was cloudy, these lights would turn on, allowing the solar panels to operate and allow slot cars to race along a track. The exhibit had a variety of performance and design characteristics, and a lot of R&D to determine what would work. Since the area where the solar panels fit where part of the design, this wasn’t easily changeable. While the sun easily powered the cars through the solar panels, the lights at 70’ (ceiling to panels) would not. This led to the eventual repositioning of the lighting instruments on the lower part of the exhibit. This required a compromise between us as the fabricator and the museum, and was based on the fact that despite best intentions of how something could work, this was what it took to make the exhibit function as required. While we did get an eventual compromise, getting to this point was a costly endeavor.
The above example was a learning experience for both my company (and myself), and for the museum. However, sometimes we see projects come in where the design specs and the performance specs do not integrate. I think these can be the most challenging to bid. You can’t ignore your knowledge, but if you price the project based on what you know will need to be done to make it function – the price may likely be much higher. Perhaps the simple option is to not bid. The other option is to care with the client the issues – but here the risk is that you are potentially sharing proprietary knowledge. I have seen a number of cases where we come in with a technique that is proprietary, only to see the client share that information with other bidders to gain a price advantage.
Finally, while we have taken on projects were we have offered the client what they asked for – and given them a product that was what they purchased, in the long run, no one exited the project happy. We didn’t make a product we could feel proud about, the client paid (what they felt was) a lot of money for an end product that met design specs, but not performance spec. Inevitably the client wants us in the end to take care of the poor performance because it has to be our fault – it couldn’t possibly be their design. Sometimes we can do a change order, and at least minimize losses, but that will still make the client unhappy.
Thoughts about how to manage the conflict between these two types of specifications or how you have dealt with it in the past?