A platform like this only works when both sides turn up, which makes its marketing site an unusual brief: it has to run two arguments at once. A warehouse manager who needs six people on Saturday morning and a student looking for a shift that fits around classes are not the same visitor. A homepage that addresses both at once addresses neither.
The second problem is that none of this is a national market. Nobody searches for gig work in general — they search in Grand Rapids, in Kalamazoo, in Tulsa, and usually for a trade as well as a city. The site had to exist at that level of specificity without becoming a page count nobody could maintain.
And the product is an app, not a website. Every page here is a step toward a download that completes somewhere else, so the structure had to carry someone from a search result to an app store without losing track of which side they came in on.