People often ask why TrueDelta doesn’t survey owner satisfaction. Well, owner satisfaction is a very tricky thing to measure, especially if you do not have a random sample (which we don’t), because of the role played by expectations. If a car is very good, but buyers expected it to be even better, it won’t score well. If a car has some obvious flaws, but buyers knew about these going in, it can still score very well. Sales duds like the much-reviled Pontiac Aztek have a tendency to score very well on satisfaction surveys because they were only purchased by people who cared most about whatever they happen to be good at. Which leads to the question: what aspects of a car are most likely to vary from expectations? One candidate: seats.

As discussed in an early blog post, seats can be difficult to evaluate during a test drive, especially during the typical brief one. Some seat issues only become apparent after an hour or more on the road. Plus it’s not possible to rely on reviewers’ impressions, as different people are sized and shaped differently, and will consequently have different impressions. More than any other aspect of the car, seats must be evaluated by the buyer.

The solutions suggested in the earlier piece remain valid, but little practiced. You can request an extended test drive from the dealer. If they think you’re serious, many will let you take the car for a day. If this isn’t possible, try renting the car. The problem with this is that many car models are not available to rent.