[jjeaff]

I think if you tie to revenue, be sure to create a cap. Because if your price is going to go up and up forever for something that everyone knows is mostly fixed cost, then people like me are going to be calculating the point that it would be cost effective to do in house and then jump ship when we get to that point.

[cdolan]

I agree here - if you tie to revenu and get too expensive (to me, over $500-100 a month), I’d find the dev time to replace this before we “overpaid”.

[runako]

> over $500-100 a month

Businesses come in different sizes, and it's tempting to project startup-level finances onto your customers. But $100/mo is not an expenditure of note at most businesses. (Neither is $1k/mo, but that starts to be a different market.)

Some back-of-the-envelope math on costs for building a billing portal:

- First pass of Stripe integration including displaying receipts, upcoming bills, edit of payment info, etc. : at least two solid weeks to more like a month of dev time. Yes, most devs can get something shipped much faster. But to achieve a battle-tested, secure, reasonably complete implementation will take a few revisions. A month of dev time in the US will vary, but $10k fully loaded isn't out of the question. If you're using an agency, this could be much higher.

Other features your business may want will drive this cost higher: Salesforce integration, QuickBooks integration, operator-facing tools so customer service reps can see what the customer sees, etc.

But let's just say $10k+ of dev time for the implementation. Then don't forget you just added a bunch of code to your codebase that will need to be upgraded, tested etc. when you rev platform versions. If you added any dependencies specifically for the billing portal, congratulations on having your billing portal now drive your release schedule. The billing portal adds permanent technical debt to your project without helping you meaningfully differentiate.

For a small app, the billing portal code could reasonably be bigger than the rest of the app. I have seen this too.

Crucially, while you're building this code, you're not building code your customers care about. Zero of your customers would upvote "build billing portal" over anything relevant to why they are your customer.