So it occurred to me that running a product selection is like running a horse race. No, wait! Let me explain!
Horses for Courses...
Horses, like products, have pedigree and form. Each has a publicly known performance, and this will guide you to select your runners. You have to start somewhere! But remember they may also have some hidden secrets which the trainer/owner (or supplier/creator) would rather you didn't know about. These aren't always negative and could be positive, but the more you can find out the more likely you'll be to back the winner.
Courses for Horses?...
Every course (or customer) provides a number of challenges to a horse (or product). Sometimes these are flat (price-based) only, sometimes they are jumps (features) which the horse must clear. Whichever goes the distance succeeding at the most jumps should win the race.
It's not always a fair lineup, though
Of course in any high-paying sport, doping and other forms of cheating are not just possible but likely, hence there are groups of people who keep an eye on things to make sure the competition is run fairly. As with high-value sport, so prejudices, bad hypotheses and incorrect measurements can cause sneak in to high-risk decisions and cause an outcome to be unfairly favoured or excluded.
Keeping it clean and fair
So, make sure your products selection is made with the best possible data you have, make sure you set requirements around features and price clearly and asses all the products against these. And keep an eye out for anything which smacks of 'rigging' the race because of bad behaviour, and you should win more often than you lose!