Maybe you want features x, y and z. If it takes a day then X is worthwhile, if it's going to take a month don't bother. Z must be done so if that's going to take a long time we need to drop y.
One of the best product owners I've worked for would often ask for estimates and base plans and features from those. Some things would be easy technically and valuable, others harder and just a nice to have (but worth it if it only pushed things back a day).