One of my favourite mental models for answering questions like this one is Shishir and Matt's eigenquestion model of framing problems. Framing a difficult question correctly makes everything else downstream much easier to answer.

I think the right framing for the question of 'what makes a good SaaS business' is these two questions:

1. What is the intensity of use?

How frequently is the product used by your customer?

2. What is the breadth of use?

How many people and organizations within your customer's business use the product?

More is better, of course, in both cases. I don't think you necessarily need to have both - Salesforce has done just fine with the Sales org using it very intensely (though I believe one of the reasons to buy Slack was to go 'wall to wall' across their customer's internal orgs). Having both also doesn't guarantee success. But if you don't have at least one of the two, there had better be something else very special about the business for you to get excited about it.

On a related note - I have long found seat-based pricing to be problematic (and frankly a bit lazy), because it introduces major friction into broadening the use of your product in an organization. Maybe AI will finally be the thing that changes how SaaS products are priced!