Word-of-mouth. Say to your users: "if you want this to be useful, ask your friends to use it, talk about it online…". (You'll get low conversion rates from this, but it's free, and it doesn't promulgate an attention economy the way that billboards and banner ads do.) Advertising is effort, but it's not that hard, really: most advertising firms are extremely inefficient at their jobs, but they get paid because the alternative is nobody advertises the thing.

It also helps if your service is still somewhat useful to individuals, in the absence of network effects. So, once the core service (organise meetups between users) is implemented, get to work on aggregating and categorising social events that already exist. (Use standard protocols, to leave the ladder around for others to climb, but keep track of your API consumers: some of them might be hostile anti-social entities like Facebook, whose existences need to be nipped in the bud.) Once you have something genuinely useful, pick a small-ish city, and guerilla advertise in it. If you've done things right, that should be enough.

Now, leave a skeleton crew on it, and move on to the Next Big Thing. You should be able to maintain three or four pieces of infrastructure in this way, some of which you might even be able to charge money for! (But resist the temptation to monetise and enshittify, or even just excessively-tweak, what you've built.)