You know Tammy, that's a great question.

It really depends on the context, but if it's a situation where they want multiple user tests, I would say do ten.

Ten is a great number, it's gonna tell you basically everything you need to know as a theoretical sample, and any more more than that is normally just corporate self-flagellation for the purpose of saying "we did it."

Here's the problem: usability testing is NOT for determining product/market fit. If they aren't sure if it will sell, they need to be conducting smoke tests first, and if they don't want to invest in discovery, I am assuming they firmly believe that it will sell to their target user base.

If you can't do ten, do five. If it can be done remotely, do it remotely. If you can do it unmoderated, so much the better, because it minimizes the Hawthorne Effect, returns less biased data, and creates a much more realistic usage scenario than a user having a moderator sitting there watching them.

I have found, more often than not, that testing between 5-10 users told me the exact same things as testing many more.

- 5-second test for initial reactions.

- Walkaround tests for overall impressions.

- Critical task tests for testing specific task flows.

- Debriefing surveys for checking outcomes, expectations, and emotional takeaways.

TL;DR: just do ten. If they want more, do more, but it really isn't gonna tell you or them a whole lot more. They'll need to test consistently and revise after they've deployed anyway if they want to stay competitive long-term because both user's needs and expectations change over time as the entire landscape of their market vertical changes.