How could Soham Parekh have improved his overemployment scheme?
Soham Parekh worked four tech jobs simultaneously but typically got caught at each company. Thought experiment: what would it take to get away with it?
Thought experiment: you wake up and you are Soham Parekh. You want to extract as many resources out of the tech sector via overemployment as you can. What is your approach to becoming as overemployed as possible?
If you’re not familiar with the Soham Parekh story, Suhail Doshi put out a PSA that Soham has been running an overemployment scam by simultaneously working at multiple startups. He did this remotely from India. TechCrunch reports that he had been sweetening the deal by taking low salary and high equity at each of the jobs. He was caught once members of the Y Combinator community started checking each others notes.
A Hacker News thread on the topic was flooded with people saying that they hired him. In summary, he excels at LeetCode-style interviews. But once it came time to work, he would give wild excuses about why he couldn’t meet deadlines. He would skip meetings, have massive delays for shipping pull requests, and get nothing done until he was eventually terminated.
I think we can do better. He repeatedly got caught by companies, and since his compensation packages were always equity-heavy — which typically vests with a 1 year cliff — he was getting underpaid per job.
Let’s break down his strategy:
Focus on startups. These are companies that are growing, hungry to hire, and are small enough to make rapid decisions.
Be a good candidate. Master the skill of interviewing. Convince people that you are one of the top candidates.
Look enticing. Proactively ask for a lowball salary offer in exchange for equity.
Add uncertainty. Claim that you are working on your US visa renewal, but raise some doubt about whether it will go through.
Make a good first impression. Attend any in-person meetings or orientations.
Surf the long tail. Obviously your work suffers because you’re working at 4ish places at once. The company notices eventually and fires you. Cash the paychecks and start interviewing somewhere else.
If you truly want to become overemployed in the tech industry, there are certainly better ways to do this.
First, trying to fleece startups is a bad idea. They actually need things to get done! And everyone knows everyone in a small company, and it will quickly become clear that you’re not producing any work. As companies grow, they develop more places to hide. I know several people who ended up in situations at Google where they didn’t need to do anything for 3-6 months. I’m curious what the record is. It was hard to engineer; usually it happened when a large team catastrophically failed and the organization started picking apart its remains.
Second, the output of his work could be compared against the output of other engineers in the company. He did full stack work, but startups employ tons of people who know about full stack work. So his output isn’t keeping up and anyone can load up Github and see that he is slacking.
So how can we do better? Being a consultant in a third-party integration tech like Salesforce and targeting businesses with somewhere in the ballpark of 100-1000 employees. If you do this right, you could hold at least 2 gigs at the same time while raking in retainers. Yes, this does imply that you need to become an expert in Salesforce, but hear me out. This company size is the sweet spot where they might have decided that they definitely need Salesforce, but realized that they don’t have the in-house expertise to do a Salesforce integration and had bad experiences trying to pay Salesforce contracting fees to do parts of the integration.
First, you’re not going to do the integration yourself. You’re going to be the subject matter expert on a team that is building the integration so that there is no knowledge lost when you leave the team. You are going to host training sessions and produce documentation about how things should be done, but obviously you’ll leave out enough details that there’s room for error. You’re going to figure out what engineers on the project are completely clueless with Salesforce and assign them those parts of the project. Give yourself deliverables that are small and achievable. So at each meeting everyone will be underwater and you will have your part done and have suggestions for how everyone can proceed — namely, by telling them everything you left out of the training and documents. If people ask you for help, just schedule a training 2 days into the future and invite everyone. You’re just always trying to produce artifacts, and a meeting is a good artifact.
The goal is to drag the project out while doing minimal work and giving it the illlusion of forward progress.
Towards the end of the project, make sure that everyone on the team is trying to use Agentforce. That should create some great messes that you can fix.
And now the magic: your retainer! This will require some experimentation, but you want the company to buy as many hours as possible while using as few hours as possible. You might think, “I don’t want to sell 2 days per month and then have to work that,” but don’t be afraid! You already set the precedent that Salesforce work is really slow. You won’t be doing much, and if they give you something real to do, you can sign a separate contract for the overflow work.
Your goal is to pick up as many retainers as possible and have to work as few of them as you can. When you first start doing this gig, you might have to put some work in to generate buzz. You’ve gotta jump start the consulting pipeline somehow. But you will eventually find companies that sign your retainer contracts and keep you around for a year or two — or more.