[moosebear847]

If you sold ice cream on the street and needed to figure out how to become profitable, would you be so opposed to trying out selling at $5 one day and $7 the next? Or would this be too unfair to the people that paid $5?

It's the intent I guess. The bartender story seems to imply a bad reason for the discrepancy (some type of unfair discrimination), while in this case it's to find out if you can charge more to eg capture back more of the value people are getting from you. Which doesn't seem so bad.

[webmaven]

Testing different prices on different days doesn't give you as much confidence (price sensitivity can vary by day of the week, weather, etc.).

Then there are tests that would be hard to run even on different days, like doubling your nominal price and splashing a "today only 50% off" or "buy one get one free" offer.

Real-world price testing is so much harder to pull off, it isn't as common, which is part of the reason web A/B tests seem jarring when the curtain is pulled back.

[nly]

You have the benefit with ice cream of a functioning market. You can look at hundreds of competitors across similar towns with a little research