Limiting your usage-based product

Published in Case Studies
August 22, 2023
2 min read
Limiting your usage-based product

You have a product that is working well, but you’re looking to generate more revenue out of it. A way of doing so, especially if your product is usage-based, is to limit a few features on each tier. But how do you decide on the limit to implement to generate some revenue?

Follow along if you’re interested:

How to limit your usage-based product for maximum revenue while keeping users satisfied?

  1. Get a table showing how each user uses your product’s different features: how much data they’re storing, how many messages they’re sending, how many templates they’re creating… You want to ensure your data is clean: use only activated users who have been using your product for at least a certain time.
  2. For each feature and each tier, you want to create a frequency distribution table: how many free users have only one template? 2 templates? Ten templates? Etc, etc.
  3. Draw a line on feature usage where roughly 80% of users will stick to the plan they have and 20% will want to upgrade. See Growth Hunhinged’s guide to reverse trials for more information about this 80/20 rule.
  4. Check how many accounts would be impacted if you added this new limit, and decide which features to base your price on to increase revenue.
  5. Rinse and repeat for every tier of your pricing you want to put limits on.

This works for API requests, file storage space used, messages sent, users added, or templates created…

Let’s say we have the following frequency distribution table for the free accounts template usage:

N of templatesN of accounts% of accounts

If we sum the number of accounts having 0, 1, 2, or 3 templates, we have 81% of accounts. We should draw the free tier limit at three templates.

Formula to get the frequency distribution on Gsheet

If your data is on the A column (A2:A11), and the classes on the C column (C2:C8), then your formula is: =frequency(A2:A11, C2:C8)

See this GSheet example.


Now, you have the usage distribution for each feature. You may have 1, 3, or even ten features you could limit, but which makes more sense?

List the impact for each limit: How many accounts would be impacted if you implemented this limit now? What’s the proportion of those accounts based on the specific tier you’re limiting?

See the following example:

TypeLimitN of accounts% of free accounts
File space2 MB562.86%
API tokens3201.02%
Rate limit400 requests/min301.50%

Here, we can see the limit that would have the most significant impact is the number of templates: almost 10% of the free accounts would have to move to the next tier if we limited it.


Previous Article
I published an NFT collection with no Web3 experience


Case Studies

Related Posts

Should your B2B SaaS offer a Black Friday deal?
October 18, 2023
3 min
© 2023, All Rights Reserved.

Quick Links


Social Media