Signal is a closed centrally controlled network that has a hard requirement on users to pay money to centrally controlled telco providers to get accounts making it impossible to be legally anonymous in most countries. It is also very unfriendly to custom clients or automation. In those regards it is no different from WhatsApp.

Matrix however is a great choice imo.

I think you're ultimately correct, the trouble is the same as what advertisers discovered when they tried to do the same thing years ago.

Facebook owned the users. Similarly, WhatsApp owns the users.

The choice between:

1. Tying yourself to a platform that could boot you at any time, while also reaching all the users they already have

2. Bootstrapping something from scratch with 0 users, when all the users you want are already using said platform and have little incentive to leave

Choice 1 is often the only viable option, particularly if you're financially dependent on the results.

I also think we repeatedly see, even on HN, that "it'll just be a matter of installing another app" is actually a huge mental hurdle for many people - especially those who aren't technologically savvy enough to know how to use something that isn't Facebook/WhatsApp for communicating with people.

While I don't enough about the problem domain and the app, "just don't use WhatsApp" is not necessarily helpful. The dev experience might be better, I dunno, but your product also needs users, and in my countries / communities, WhatsApp is the only thing they use, and meeting your users where they are is important.

Forcing your users to first install one more messaging app is going to cause people abandoning you immediately or slowly churn, which can make or break a product.

In many markets, avoiding WhatsApp if you need messaging kills your company on arrival.

WhatsApp has a critical advantage in Kenya from my experience: telecom operators such as Safaricom include quite generous WA bundles in their customers' subscriptions, allowing one to use it even though they would otherwise be out of credits. Being out of credits/airtime is not uncommon, since the economical bundles are sold as expiring in 24h/7d/30d.

It's more generic than that. If you build on other peoples turf or your operation depends on others you risk everything on a change of terms or a service vanishing entirely. The company could be big, could be small.

One classic joke is buying your competitors supplier.