Description

The goal was to build a “LinkedIn for Freelancers” that would replace Upwork and have productized services offerings. We collected a structured dataset on a freelancer's historical projects, lead preferences, and other real-time data to match them with incoming client projects. We would also use the data to generate instant cost and timeline quotes.

To prototype the algorithm and figure out the data we needed for accurate results, we used GSheets to iterate quickly with our pilot group of freelancers and client projects. We tracked each client closely to identify holes in the funnel via a custom high volume CRM.

Once the product was stable, we built a front-end out of Vue.js for clients to self-serve with. If you watch the video below, you'll notice the quote and matchmaking results change with each user input. This enabled clients to refine their project scope without the need of speaking with each freelancer and agency, saving hours of combined time over weeks.

For the first few dozen clients using the self-serve interface, we were on the phone with them simultaneously to identify points of confusion. Their first reaction to seeing the real-time quote/matchmaking was often "wow, I had no idea something like this could exist".

After proving the concept with clients, we began to grow demand, using the same historical dataset powering the algorithm to create a growth engine — the Showcase and Insights products — which grew organic traffic by 30% WoW. Unfortunately we ran out of runway though before we hit breakeven.

https://www.youtube.com/watch?v=iz1Ob6U3nX4

My Involvement

• Led prototyping using GSheets and Zapier. GSheets was both the front-end UI for freelancers, and the backend to power the matchmaking algorithms.
• Led development for the matchmaking and predictive algorithms.
• Recruited a pilot community of 30 freelancers to iterate on the algorithm with, asking for new data points each week to improve match and prediction quality.
• Once the concept was proven, my cofounder built the frontend UI in Rails. To ship the product update quicker, he and I turned the turn the GSheets into a microservice using Zapier to return algorithm results in Rails (see the demo video below)

Traction

• Scaled to 200 freelancers, 200 leads a month, and $600k/yr run rate on Vue.js/Rails/Google Sheets/Zapier MVP stack.

Demo was created for the purpose of showcasing the product for low code/no code developers, but the overall product concept should still be clear. Demo is of frontend Rails/JavaScript UI powered by GSheet. Original GSheets prototype had 100% feature parity to this but was used as an internal tool by operations team (rather by clients) and couldn’t handle more than 50 freelancers in results without GSheets servers timing out. Freelancer-side tool and GSheets prototype demo available upon request.