I don’t have specific numbers to hand, so assume any percentages below are inflated placeholders.
Let’s say 10% of our customers explicitly ask for Zapier integration because they are already Zapier customers. A further 20% have the expressed need for some third-party automation that we don’t directly support, but is supported through Zapier. And a further 30% could benefit from it but have no idea this kind of thing even exists. 0% ask for n8n. 0% have even heard of n8n.
Well we need to build the Zapier integration to keep that 10% happy. Now that we have that integration, we can turn to the 20% that need something like this and tell them to sign up for Zapier, and then they will be happy too. Then we can publish how-to articles and give a nice surprise to the other 30%, who will also go and sign up to Zapier.
There’s friction here. Some customers will fit in Zapier’s free tier and others will have to pay extra. The process for hooking Zapier up to our product is clunky. And every time the customer wants to change some aspect of their automation, they have to leave their product dashboard and go to an external service.
The goal was to self-host n8n so that customers could keep doing everything within our product. Most of the 10% existing Zapier customers would carry on using Zapier; some would switch. We wouldn’t need to send customers over to Zapier to keep the 20% of users asking for something like this happy, and the how-to articles would help the other 30% without sending them to Zapier as well. Some of our customers would save money by not having to pay Zapier, for others it would make no difference. Our customers would be able to manage their integrations without going off to some third-party site.
You can see how this is a desirable thing for us to do. You can also see that the value to us is way, way, way below $50k/yr. We aren’t going to gain or lose any customers over this. The main difference for us is marginal UX improvements.
n8n received $0 from us. If n8n were open source, they would still receive $0 from us. The difference is that we wouldn’t be sending 50% of our customers to become new Zapier customers. n8n would have gained one small integration – us. There’s no point in us building an n8n integration when we have Zapier though, because nobody is asking for it and Zapier does everything we need and has more integrations. It’s also possible that we / our customers would add to the other n8n integrations if we needed them or contribute functionality or bugfixes, but again, that’s veering a little to close to the “payment in exposure” argument I dislike.
As I said before, if n8n want to play things this way that’s their prerogative. But somebody here was telling people that it’s open source when it’s not. It being open-source or not is a big deal for cases like ours; it’s not “purism”.