Exactly. Deciding on some very expensive subscriptions that can cost 1k per month or so might be worth thinking about, but this is just meaningless optimisation.

not at all meaningless. unless you have money to invest, at the beginning you don't have an income. i could not afford to spend $300 a month to host a new product that doesn't make any money yet. i can afford the $20 however, but then once the product does make money, why should i change it if it works?