Stacks Job Portal

Background

HireVibes was a job board for the entire blockchain industry. The sharper opportunity was smaller: let a single ecosystem run its own branded version of the board, embedded on their site, while every job and candidate still flowed back to our central database. Onboard one ecosystem and you grow their pool and ours in the same motion.

Stacks signed on to be the first. They were paying, there was no precedent to copy, and there was a deadline. My job was to design the embedded product, sit with the client, and turn what they told me into shipped iterations.

Putting it on someone else's site

You can embed a product two ways: an iFrame or a real API integration. Engineering chose the iFrame — faster to ship, lower risk, the sensible call with a signed client and a clock running. I didn't make that decision, but I had to design inside it.

The problem showed up fast. Picture a developer on the Stacks site, browsing roles inside our widget, who finds one they want and taps apply — and the page throws them out to a different website, HireVibes, a brand they've never heard of, to finish. Every interesting action wanted to eject the user off the client's site. The whole pitch was keep the ecosystem's people in the ecosystem, and the frame underneath was working against that.

Opening up the talent pool

The first version worked the way the old product did. A candidate applied to a specific role, and only then was their profile shared — with that one company. A connection only happened if the right person found the right posting at the right moment.

That's a slow way to make matches, and it didn't fit the business. Our model was crowdsourcing candidates, and the people who make connections for a living are third-party recruiters — not individual jobseekers hunting for the exact right listing.

Results

While we shipped these changes, the Stacks pool went from 191 to 275 — a 44% jump, on a base I'll admit was still small. Six months on it hit 935: a 3× increase. That was during a deep bear market, when most ecosystems were shrinking, not growing — which is the number I'd point to first. It was enough to bring on a second ecosystem, Mundocrypto, on the same model.

What I'd do differently

The iFrame is the thing I'd revisit. It was the right call for shipping on time, but it fought the white-label promise the whole way — every decision in this story was partly me designing around a frame that wanted to send users away. Next time I'd make the case for the proper API integration earlier, before the constraint was locked in, instead of spending design effort working around it.