A few thoughts. 1) Get a diverse group (geo + backgrounds) to use it. 2) Invite people to use your product in person (on their laptops), 3) Measure the impact using www.survey.io – iterate and fix all the user flows.

It depends. You do different things for an enterprise product than you do for a consumer product. Beta is not just a flash in the pan moment. You are trying to gather different feedback from users with varying levels of risk and reward. Gathering feedback is key. You need to make sure that you have an engagement and feedback process that is very explicit. Participants might choose to not participate which is ok if you app is well instrumented. But it probably means the CEO/Founder is dialing and smiling to talk to every single user and gather their input.

Basically you need to create a set of different phases:

1. Phase 1 - Friends and Family
You're checking to make sure the code works. Don't let everybody in. Make sure your stuff works. Beyond your nerdy engineering team.

2. Phase 2 - Private, Highly Segmented
Segment the hell out of your audience. You only want people that are part of your primary segment and that are very early adopters. Make sure to slowly invite more people.

3. Phase 3 - Targeted but Open
You may be able to test the virality of you customer acquisition. Most of these users won't provide real feedback. You can test your acquisition, virality, fall out points, messaging, etc.

Check out https://blog.serverdensity.com/running-a-successful-beta-trickle-effect-and-phone-feedback/ for more.

With clippPR, we were ultimately building a paid product, so we were only interested in learning from and iterating on feedback from beta testers who were willing to pull out a credit card. We actually charged $1 to get in.