Which is why almost none of them have ever been successful at building a recruiting company. Because recruiting is a people problem, not a technical problem, and people problems are vastly harder to scale.
If you don't believe me, go talk to third party technical recruiting firms and ask them what the backgrounds of their top contributors are. Unless they're being spoonfed with a sweetheart deal from a former employer, virtually all of these people can barely operate their own PCs, much less know anything about engineering. I used to be a recruiter. The three most successful third party technical recruiters I know (with incomes in the 300k+ range in Texas, not SV) are a high school dropout, a college dropout, and a theater major. These people know how to communicate with other people, are as nice as can be over the phone, and they will fight you over a stray nickle.
Being great at recruiting is about being great at building relationships. Building relationships is built upon trust. Clients need to trust that the recruiter can find candidates and convince them to sign on when the time comes. Candidates need to trust that the recruiter is motivated to find them a position. As long as the effort is there, the better the recruiter balances that trust factor, the more money they make.
Triplebyte sucked at the people game. They sprang public opt-out on people and couldn't figure out why that would be such a big deal to potential job seekers. They assumed that their process was so great that experienced devs who had jobs and lives and lifestyles and families would just jump at the chance for impersonal coding evaluations. They started a company with no idea how to recruit and hired people to build a recruiting product who had no background in recruiting. And no, founding a company does not give one recruiting expertise, even if they're doing the hiring. It is both a talent and a skill.
Oh, and here's a fun one: they failed to recognize that along with being a people game, recruiting is a numbers game. Being highly selective in your pool of candidates is a genius way to run out of money as a recruiting firm, especially when you've already chosen to place numerous self-own barriers in front of you.
The average third party technical recruiter with a year of experience actively maintains and prunes their own database of 3-400 candidates. More than that is too many people to manage. You can either pay a bunch of engineers to attempt to automate a small part of their process, or you can simply hire people to do the work and pay them according to what they produce.