I Treated My Job Search Like a Product to Build

I graduated into a difficult market. Not just difficult in the "hiring is slow" way. I was an early career person trying to move from software engineering into a product role. Two years of prior experience. Competing with people who had been PMs for five years. The conditions were already stacked a certain way and I knew it.

So I made a decision pretty early on: I was not going to out-apply everyone. I was going to out-think it. What happened next was honestly one of the most chaotic, interesting, frustrating, and skill-building periods of my life. And I want to write it down.

Job applications are time consuming. That part is obvious. But the real problem is not just volume. It is that everything requires the same amount of effort even when it should not.

Writing a cover letter. Researching a company. Finding the right person to reach out to. Following up. Keeping track of who you talked to. Every single task eats your day and most of it is just process, not thinking.

That is the problem I kept coming back to. Not just "how do I do this faster" but "how do I do something that other people are not doing."

What I learned from building it was more valuable than the app itself. I learned how to think about what the actual bottleneck is before I start building. And that became the seed of everything that came after.

I started using Clay on the free plan. You can only export 100 people maximum so I had to be very strategic about who I was pulling and what filters I was using. This alone taught me something real: when you have limited resources you get very good at defining your criteria precisely.

I also used Clay to generate personalized LinkedIn messages. And then I stopped, because I noticed something. Everyone was doing the same thing. The messages felt generated. Recruiters and founders could tell. The personalization was surface level because it was all coming from the same type of prompt. That was a pattern I kept running into. Every time AI made something easier for me, it also made it easier for everyone else. The tool was not the advantage. How I used it was.

Did it save me time? Honestly, not that much at first. But it taught me how to think about user flows. How to use AI not as a replacement but as a layer. How to build something that respected the human judgment that should still be in the process. That ended up mattering a lot when I was talking to hiring managers about how I think about product.

At some point I went deep on automation and it got kind of wild.

What it did teach me though: how to think in systems. How to chain tools together. How to build something that runs without you watching it. Those are real skills and I use them constantly now.

Sending cold emails is also a time consuming task. When you are emailing three people at the same company for the same role the emails are almost identical. So I built a Google Sheet with columns for name, company, title, subject line, and body. The body column used Excel formula syntax to pull the person's name and company into a template dynamically.

I also tried Phantombuster for LinkedIn connection requests. Fourteen day free trial. It was reaching people I never would have found manually and the acceptance rate was around 10 to 20 percent. The subscription is expensive though. If you are doing very targeted outreach it can be worth it.

This one was a little unhinged but I loved it.

When there was a role I really cared about, especially at a startup, I would build a quick personal website using Lovable specifically for that role and sometimes for that specific person I was reaching out to.

Not everything worked. Some of the automation was honestly busywork dressed up as productivity. There were weeks where I was building workflows instead of having conversations, and conversations are what actually move a job search forward.

The search took longer than I wanted. It was hard in ways that were personal and not just logistical.

But I came out of it with skills I genuinely did not have before. Technical skills, yes. But more than that: the ability to look at a messy, overwhelming process and ask "what is actually broken here" and then go build something for it. That, I think, is the thing worth writing down.