Sometimes we use products, no matter how difficult they are to use.
Friction is a good proxy for value. It's a simple cost-benefit equation, where cost is often more observable than benefit. And it's much easier to optimize and quantify cost rather than benefit – how quickly, efficiently, or easily we can do something (rather than how much we enjoy or value an experience).
Better yet, high-cost activities are more likely to have inelastic demand.
This principle applies to generational technology that looked like a toy before the costs came down. For example, the Wright Brothers Flier could only travel 200 meters. The Tesla Roadster could barely break a range of 200 miles, and there were no charging stations.
It also applies to startups. For example, users kept going to Twitter in the early days, despite seeing the Fail Whale 500 error page nearly every week.
The corollary to this is Schlep Blindness, an idea by Paul Graham that the tedious and unpleasant tasks we do anyways tend to be overlooked startup opportunities. His quintessential example is Stripe – many developers had to (painfully) figure out payments on the Internet, but few of them thought to solve that problem generically for everyone. Schlep blindness often means solving unsexy problems, but I don't believe it has to be. PG sums up schlep blindness,
Instead of asking "what problem should I solve?" ask "what problem do I wish someone else would solve for me?"
Then, friction as a proxy for value might be asking,
What solutions are people jumping through hoops to use? The corresponding problems might be more valuable than they appear.