I know, cuz I am one.
I also looked at Stripe Billing for usage-based and it didn't meet my requirements either. (though I am using Stripe Billing with flat charges).
My exact use case is:
- I want to sell API credits upfront including for subscriptions (i.e. user signs up for $10/month of credits, they pay $10 upfront, and can spend $10 worth of credits). Stripe billing doesn't support charging upfront, for usage-based, they only bill after the billing period is finished. Some of my users have been gaming the system, canceling and not paying so that doesn't work for me. I believe Lago did support billing upfront.
- I want to freely mix subscription based and pre-paid credits. My users go over their quota one month, they don't want to upgrade their subscription to the highest tier, they prefer a one-off top-up. I need to control over which credit gets consumed first. Stripe billing and Lago both had issues with this, I can't remember exactly what.
- I wanted to support as many payment methods as possible, and particularly Chinese wallets for pre-paid credits (Alipay, Wechat). Lago has no plans to implement this, I was half considering implementing it myself inside Lago. I don't think Wechat and Alipay will matter much for B2B businesses.
- I'm also a huge fan of massively regression tested code, and Stripe Billing with its test clocks blows Lago out of the water here. Lago has no ability to walk the clock forward for subscription lifecycle testing. Though maybe this matters less if you have faith in the product, you can just expect to get the right callbacks on time.
I did notice that the Lago devs on slack took time to answer questions down to the deepest technical level. If I was running a B2B startup, I would probably try really hard to fit Lago in the picture, particularly at at time when stories of Stripe account bans are so prominent.
If the user cancels their subscription, I run it through the next payment period for their usage based billing period and then cancel it.
Also, this is kind of their thesis, isn't it? That developers want open-source software to do their billing, which they can modify themselves if they need certain features instead of relying on a proprietary provider like Stripe. Doesn't sound too outlandish to me.
The fact that even Amazon has switched to Stripe shows this.
And that is not even getting into the PCI, PCI DSS, SOC, I and II, etc compliance soup of self hosting.
Basically, here are the worst of the worst of Amazon's transactions, please take this risk onto Stripe's platform, then we will cut cost with full knowledge of exactly how much we will save by spending developer resources to cut out Stripe for this particular card type in this particular country.
I would not call this a healthy relationship, but rather stripe taking the marginal additional volume that Amazon brings to their platform to net slightly more money and get slightly more data points on weird payment methods.
If anything, the mind fallacy isn't the GP but on the commenters pretending that they love losing money. If we could pay $0 and get everything we want, we would. Everything else is just performative commenting to save face.