Proffer Jobs: Decentralized Recruiting on the Blockchain
App #3 of Proffer’s 5 apps in 5 days series
Proffer Jobs crowdsources recruiting by letting hundreds of recruiters of varying expertise choose the right candidate for a job posting — one of 5 apps my co-founder Anshul and I built for Coinbase’s Toshi/Token Hackathon to explore use cases of social search on the blockchain. These 5 apps helped us iterate on the protocol design for Proffer, and collectively won the grand prize in the hackathon.
The goal of this article is to discuss why crowdsourced recruiting is a valuable pursuit, and how it can be implemented easily with Proffer, the foundational protocol for social search on the blockchain.
Note: If you’re curious to learn more about Proffer before going through this article, read the tech spec here and a higher level walkthrough of a social search on Proffer here.
Why should recruiting be distributed and how can we do it better?
In a way, the fundamental process of recruiting in today’s companies is already distributed. The team that’s hiring will hand off the task of filling the role to the company recruiter who in turn can use other platforms like Hired, social networking sites like LinkedIn, or headhunters to find the right candidate for the job.
The problem with this is that soon the original hiring team has received a bunch of potential candidates who aren’t a great fit for the role.
They lose days interviewing the wrong candidates. It’s because the distributed recruiting avenues they are using don’t understand the team and job opening that well. Subject matter expertise, the working styles of each team member, etc. — there are so many factors that decide whether a candidate will be compatible for a job or not. Most of these factors require human experience. Experience in not just working on the subject matter, but also alongside different teams that come in a variety of shapes and sizes.
So it comes back to the same problem we see in the romantic matchmaking scenario we discussed with Proffer Pair:
- People are experience goods. It’s impossible to describe the ideal job candidate in a few words.
- Only people who have worked in similar positions and teams can judge whether a person is a good fit for the job opening.
- Proffer works great on problems that require offline expertise.
Implement Proffer Jobs using the Proffer Protocol
So let’s use Proffer to come up with a better shortlist of job candidates so that our teams don’t spend days interviewing the wrong candidates. We’re going to crowdsource the recruiting process through professional and amateur recruiters who have expertise in a given job opening.
- For professional recruiters, it’s expertise in the form of past success in finding the right people for similar roles as the one in the job opening.
- For amateur recruiters, it’s expertise in the form of having worked in similar roles.
Define the Seeker
The first part of running a Proffer search is figuring out who is the Seeker — the person, entity, or dApp looking for the answer. For Proffer Jobs, we’re going to make the Seeker the company who has a job opening. The Seeker starts the search and the Responders provide answers.
Most companies also have a referral fee/bonus for the headhunter who helps hire a new candidate. It’s usually one month’s salary — so if the job opening comes with a salary of $120,000, the headhunter will get $10,000. Likewise, if responders in Proffer Jobs can find the right candidate for the company, they should get the $10,000 and split it amongst themselves. To achieve this, the company tells Proffer Jobs (which informs the Proffer Protocol) that the SeekerStake is $10,000 (converted into tokens that is). The SeekerStake is added to the TokenBackingPool, the pool of tokens that are split amongst the Responders who are correct. Something we’ll get to a little later on.
Seeker asks a Question
Now let’s give the Responders something to answer — the Question. Every Proffer Search has a Question and for Proffer Jobs, our Question is whether a potential candidate is a good fit for the job opening. Proffer Jobs lets Companies and Job Seekers sign-up. So when a company posts a new listing on the Proffer Jobs dApp, it generates pairs matching the job opening to eligible candidates.
Responders submit Answers and Proof of Confidence
Now how can a Responder answer this Question? The Proffer Protocol requires a Proof of Confidence in the form of tokens (VoteStake). For Proffer Jobs, every responder can say Yes or No to a job-to-candidate pairing and they have to back their answer with tokens. For Proffer Jobs, let’s set it to $1 (converted to tokens). If the Responder is right, they get back the VoteStake, otherwise they lose it. All VoteStake is submitted to the TokenBackingPool, which now will contain both the SeekerStake and all of the VoteStake.
Complete the Search
Next, we can tell the Proffer Protocol how many Responders to ask before closing the search. For Proffer Jobs, let’s ask 100 recruiters. We can also specify a time / duration for which the search should remain open for responses. For example, setting the protocol’s ResponderTimeLimit to 1 week means search ends in 1 week regardless of how many responses have been collected.
View Results and Take Expertise Into Account
Once the search is closed, the company will be able to see the following results:
Take a look at the Skill values in these results. They are the special sauce of Proffer’s Global Expertise Bank. When Responders answer correctly for a Question in a certain topic, they earn Skill for it. When they get it wrong, they lose Skill for it. For Proffer Jobs, we can specify topics relevant to the job opening when we run the Proffer Search and gain valuable expertise data.
For example, we can add the topics of Solidity, Cryptocurrency, and OpenSource for the sample job listing we showed above for Proffer. When users join Proffer Jobs as professional or amateur recruiters, their resume gets vetted and their expertise gets initialized in the Global Expertise Bank. Proffer Jobs can also tell the Proffer Protocol to only use responders who have high expertise in Cryptocurrency jobs. This would mean that only recruiters who have recruited for Cryptocurrency jobs or have worked in a Cryptocurrency job will be allowed to vote on job candidates for our example job.
This will make it so that the Skill values reported above represent the aggregate expertise of the Responders voting Yes and No respectively. We get a taste of the power of this in the sample results. We see that even though there are fewer votes for Yes than there are for No, the company should trust the Yes responders more and interview Amanda because there is more expertise behind that answer.
Reward Correct Responders with Financial and Skill Payouts
If the company does decide to go ahead with the Yes voters and they end up hiring Amanda, all of the Yes voters will get to split the TokenBackingPool. With $10,000 of SeekerStake and $1000 of VoteStake, each Yes responder ends up earning $11,000 / 42 = $261.90. This includes the $1 they each had put in as VoteStake. So they earned $260.90 each. The same would happen with No responders if the company didn’t hire the candidate. If the company didn’t interview the candidate, then everyone gets their VoteStake back because no one was proven correct.
In this example, we were able to get recruiters who had high expertise in cryptocurrencies to help us find the right candidate for the job! Recruiters, professional and amateur, are able to earn in exchange for their expertise. Companies are able to get a better shortlist of candidates because all of the recruiters involved in the process have high expertise that has been vetted through previous performance on Proffer Jobs and through their professional background. This is all possible due to the Proffer Protocol and the blockchain’s ability to provide frictionless transactions.
