> thought Notion's free-for-a-certain-number-of-blocks pricing was great. It let you try the platform without limits on functionality

The problem was that it was difficult to know what a block was or how much you could do with 1000 blocks until you had invested enough time in the platform to be able to estimate how many blocks you might need to evaluate it.

Then there was additional mental overhead of trying to conserve blocks to extend your free trial. Do you implement something the correct way? Or do you maybe cut some corners to use fewer blocks so you can stay on the free plan a little longer? This extra decision making step added a little bit of fatigue to every interaction, which isn’t what you want when users are trying to evaluate your tools.

Typical trials have well-defined and easy to understand limits in units that users already understand: Users, bytes, projects, documents. If users have to learn your project before understanding the terms, it’s mental added overhead.

I don't think the kind of people that are overly worried about not using up their free trial generally make good paying customers.

Here was my experience: my team signed up, started using it, tried out all the different features (none of which were behind a paywall), didn't hit the 1000 block limit, and started paying within a week of signing up because it was exactly what we were looking for at the time.

I was one of the people that saw the free 1000 blocks and chose not to signup, because I didn't understand what that gets me.

I am currently happy on the free-tier and would cheer-lead internally at my job if there was interest to migrating to Notion.

I was one of the people that saw the 1000 block limit and just signed up to avoid the hassle of even thinking about it.

I wonder how the total numbers of people that did each of these things compare.

You could always create another free workspace, and move pages between them. Once you got tired of that, you bought an account. Happened for us, and I think it was a great way of pulling us in.

Something related: more than a decade again, I was working as a sales engineer selling an enterprise search engine. A common pricing model we publish is charging $X per gigabyte of search index [1]. This sounds like similar rationale, because it feels like an honest pay as you use scheme. But it's hard for customers to understand how much they would be using. And of course since it's an enterprise solution, there is heavy switching costs playing into the equation — both into our product and if they adopt it, away from it.

[1] The size of search index depended on the amount of content indexed as well as heavily on the features that were turned on. E.g. pre-index search-term expansion would increase the size by a factor, and so on.