So, I am well aware that it's not true that in menu's etc. one should show only 7±2 options, because the original research by Miller was about working memory. However this does not necessarily mean that more is better (though in some cases it might), given too many options it does become harder to choose and choices that are too complex will be postponed or even skipped. Thus my question is: has there been any research done as to what the optimal number of choices is when a user is actively comparing the various options?
And to give a concrete case to discuss:
There are 200 products of a certain type. The user wants to buy one product of that type. The system knows which products are best and worst (put differently, the 200 products are sorted by likelihood that the user will want to buy them). How many products should the system actively present to the user to maximize sales?
The thing is just that in a case where a user is bound to make a choice I don't see any downside to showing more choices (except a worse UX), but in cases where a user already is in a stable state: for example when he already has a mobile phone subscription and is vaguely interested in switching giving too many choices can easily lead the user to postponing the choice to another time.