From 0 to 10K Users: Our Product Experiment with JobPulse
A few months ago, Jerry Yu and I decided to take on a side project to see what it’s like to build a product from the ground up. That’s how JobPulse was born — a website that gathers entry-level software engineering jobs for students. We launched in June and by September, we had hit 10,000 monthly active users. Although the website traffic has exceeded our expectations, JobPulse is still a naive and unfinished product when compared to real-world standards. Whenever someone asked me what our goal was with JobPulse, I’d always say we just wanted to have some fun. The mere fact that we were able to launch something, to go from an idea to a platform with real users, feels like a win in itself. And we’re helping people? That’s just the cherry on top. It is such an ego boost when I receive LinkedIn messages from JobPulse users telling me how much they like the platform. That’s music to a Product Manager’s ears!
I see JobPulse as my personal playground where I get to do a bit of everything — from coding the website and marketing it to our audience to handling the day-to-day operations. This project has been a journey of self-discovery too. If you’ve read my blog about grad school applications, you’ll know I’m the kind of person who wants to experience it all. Doing a side project like this has been such a treat because I get to dip my toes into every aspect of building a business.
Unemployment Was the Inspiration
Fun as it might look right now, the idea of making a website like this came from our personal struggle. We’ve both been through the fear of being jobless, stuck in that endless loop of applying, hoping, getting rejected, and repeating. To me, it was a period marked by social withdrawal, despair, and a crisis of confidence.
What I didn’t realize then was that the suffering later became invaluable in the development of JobPulse. In a way, our experiences might have created a founder-market fit, much like what Partner Jared Friedman describes in his video How to Get and Evaluate Startup Ideas. The firsthand experiences we gained during our job hunts have significantly shaped JobPulse — we were the users. Our networks, consisting of other student job seekers, have provided a rich source of insights and support. Moreover, the social platforms where we used to vent and post content have surprisingly turned out to be ideal places to pitch JobPulse.
If I could time travel and speak to myself ten months ago, or to anyone who is actively job hunting, I would say: that hope is often hidden in places yet to be discovered. Keep moving forward and your efforts will ultimately pay off.
During my unemployment times, I started coding again and developed a small tool called JobLetter, which was obviously inspired by my own experiences. I also ventured into new territories, dabbling in design and marketing work for friends’ startups. At that moment, it felt chaotic and I was lost, desperately seeking something to cling to. Interestingly, the skills I picked up during that period are the very skills I now apply to JobPulse. In a sense, it is the needle that threads together all of my scattered experiences into a cohesive tapestry. And who knows, JobPulse might just be one piece of the puzzle that opens the door to another grander adventure.
Starting Off Petite
Hosting a website with 10,000 visitors might sound big, but the approach we’ve deliberately taken with JobPulse is to keep things small and manageable. That means a small team, a niche group of users, and focusing on just a subset of their needs. To make the scope even smaller, we only focus on job opportunities from a few tech companies. I’m not completely sure if this was the right move, and perhaps we’ve missed out on some bigger opportunities as a result. But there are definitely some upsides. It’s a project we can handle without getting overwhelmed, and it allows us to strike a good balance between work, life, and our passion project.
When we say we have a small team, it means just the two of us. We do turn to help sometimes, but for the most part, it’s Jerry and me. We often tell people that Jerry’s the engineer and I’m the product guy, but in reality, our roles are much more fluid. We both handle full-stack coding and are always ready to jump in and help each other out. We’ve completely bypassed the traditional process of writing PRDs, designing mockups, and then implementing them. Instead, we often go straight to trying out an idea to see if it works. In my opinion, it does help to speed things up and iterate fast.
As for user groups, we chose to focus on students because that’s a group we can relate to the most. In recruiting, students are usually at a disadvantage, with fewer skills and less bargaining power. With the recent layoffs in tech, students are now facing a new grad-unfriendly market that’s oversaturated with job seekers. To make things worse, we’ve noticed that many recruiting platforms seem to favor recruiters over job seekers, as recruiters are the ones paying for their services. That’s why we wanted to do something, however small, to make the job-hunting process a bit easier for students.
Out of the student group, we specifically focused on those seeking software engineer roles. The reasons are simple. Both of us have applied for SWE roles before, the market is large enough to attract users, and the application process tends to be more standardized. As Chris Gardner mentioned in his lecture on How to Build a Product that Scales into a Company, this group is likely to be our minimum viable user segment.
Users Call the Shots
Identifying the minimum viable user segment is just the starting point. Truly understanding the individual users represented by this segment is what drives our product development. The pursuit of user-centered design has been planted in my heart during my education at Carnegie Mellon. Therefore, right from the start, Jerry and I were determined to create a platform that truly serves people, rather than doing an engineering stunt.
Engaging with users has been an important part of our journey from Day 0. Whether it’s with current or prospective JobPulse users, these conversations have become a regular and enjoyable part of our process. They allow us to see whether the platform is meeting users’ expectations and how our concepts translate from theory to real-world application. To better gather these insights, we’ve set up a Discord community to interact with our users. It turns out that our users have been awesome at catching and reporting bugs even before we spot them ourselves. Plus, when we drop new features, we can get instant feedback from them, which really helps us understand their needs better.
Besides running a community, another way to learn more about the users is through user interviews. So far, we’ve conducted 4 rounds of interviews, engaging with 8–9 participants from diverse backgrounds every time. This iterative feedback has been crucial in shaping JobPulse and honing our problem statement. Some of the best ideas have come from our users, such as the option to report closed job listings or incorrect experience requirements. Initially, we planned to use hard-coded solutions or AI to identify closed jobs, using user reports as a temporary fix. Surprisingly, user reports proved more accurate and timely than we anticipated, leading to increased engagement and a united user community. So, the reporting feature stays.
In my view, the product truly belongs to its users. My role as Product Manager isn’t to dream up the most innovative ideas but rather to prioritize, validate, and iterate on the ideas that prove successful. The most challenging part for me is making good product decisions, or in other words, deciding which ideas are worth pursuing. Sometimes I feel lost when looking at the Mixpanel dashboard and sifting through user interview notes. Striking the right balance between our capabilities and what genuinely matters to our users is a delicate task. It’s especially tricky because we’re building a product from 0 to 1. Every decision could slightly alter the product’s tone or make us favor one type of user over another. Inevitably, sometimes our decisions are subjective and end up reflecting a bit of our own personalities. Perhaps that’s part of the beauty of creating something we truly own — we can leave our mark and call the product our baby.
