Stop using videos for marketing material. Especially among the type of user you're trying to attract, video demos or marketing material are often a no-go; they either don't feel like taking the time to watch them, or would prefer text for higher information density and consume-at-your-own-speed. Post text, screenshots, annotated demos of an
actually usable guest profile on the site, heck, even short animated GIFs if you absolutely must, but not full-length marketing/demo videos.
The sign-in flow might suck, as many others have commented, but even if it didn't, $29 just to see if something works out for them might be too much for many users. Even if they have the money to burn, that doesn't mean they'll spend it here. Offer a trial or a freemium service.
I think a lot of folks on this thread are commenting about the specific technologies, UX "feel" of the product, or how it provides the features that it does. Those might be valid criticisms, but I think they're missing a much bigger point: you have a few hundred people that said "interested" during a proof of concept, a few hundred more that started sign-in, and a few dozen after that who eventually didn't make it out the other end. The narrowing of that funnel speaks to sign up problems, sure, but the input numbers themselves are still way too small. You're marketing this offering aggressively, and they're still small. Unless you're doing B2B marketing for some huge, high-dollar product, numbers that small mean either inaccurate marketing (there's a product/market fit, and people come to your site looking for something that does what you advertise, but it turns out to not actually do it) or a lack of a market.
Consider that your offerings fall, for many freelancers, into the "perceptual cost hole": things that are minor hassles for many people, but that, at the end of the day, they often don't consider to be costs of their business. Like, DocuSign is miserable, and dealing with companies' NDAs sucks. But when reading your offering, I have to remind myself that those annoyances are (maybe) not things I have to waste time doing; that there are things that could make life easier. My gut reaction to those features is "yeah but I spending time/money/hassle on those things is something I already do and that's just a part of freelancing". This is obviously wrong after thinking about it, but, if shared by others, this is a perception that may seriously damage adoption of your product while causing the initial "interest spike" you saw.
As alternatives, consider: free initial offering/trialware; freemium services; a pivot to emphasizing one of your features as a primary offering (e.g. NDA hosting or something).