i've done this and do this, with mixed results, and will continue to do it. one spectacular-for-me failure. one small win. several not-great and never-got-to

i've done this and do this, with mixed results, and will continue to do it. one spectacular-for-me failure. one small win. several not-great and never-got-to-production projects. my last time i specifically took a new tack -- farmed out the same test project to 3 diff people - or tried - only ended up with two diff teams - chose one. both were good-ish, but one guy was too-inexpensive (guy i chose), and the other guy was trying to overengineer, etc. i'd def recommend my current guy, tho i don't have work with him right now, and i'd prob recommend the other guy, too. both upwork. ping me if you want a name. this was a real simple php/laravel/postgres/bootstrap project. and i've been hacking on it a bit since, which is somewhat typical. my guess is i'll try to add a bunch of new functionality, at which point i'll send it back to my dev. i typically provide at least a Balsamiq, then a few must haves and/or must not haves, then hope for the best. i prefer to not be involved until something is just done, and have set up 'milestones', which you can do in upwork, and that can work, but it's all kind of a shit business, imo -- there def needs to be a better way.