What not to do in CS PhD Application

There is already a ton of great advice out there for CS PhD applications. I don't think I am the one to give the best advice, especially considering how many times I actually failed to get a position. Instead, in this article, I am listing the traps I fell into and the mistakes I witnessed during my journey. My hope is that readers can find some encouragement in my missteps. The PhD application process is hard, but it’s not that hard. (Okay, it is, but you will survive it).

Do not apply just because the job market is terrifying.

I first landed in the US in Fall 2022. Just a few months later, the tech job market transformed from post-COVID heaven into absolute hell. I had planned to apply for a PhD before I even started my MS, but I watched many talented, hardworking peers—people who would have had multiple offers just six months prior—struggle to find jobs. I've had people tell me I made a "smart decision" retreating to a PhD in this economy, and I know others who explicitly chose the PhD route solely because finding a job was too hard.

To me, finding an industry job and doing a PhD are two entirely different paths to success, and neither is the "default" backup plan. Talented people will succeed in either arena, whether that means surviving an insanely competitive company or starting a PhD at a prestigious school. I have to admit, I sometimes held onto the comforting mindset of, "Well, I can always just find a job if I fail my PhD applications." That might be partially true, but ultimately, the only thing that matters is that you are doing work you care about. There are plenty of mediocre jobs, but there are also plenty of mediocre PhDs. If you choose to do a CS PhD, you need to be clear about why it is the most appropriate option for you—not just the least inappropriate one.

Do not do the math on your future salary.

We often hear about people at top industry labs making crazy money, and doing a CS PhD seems like part of the path to unlocking those roles. But then there are the counterarguments: choosing a PhD means losing around five years of industry earning potential. If we sit down and crunch the numbers to find the financially optimal strategy, calculating whether you make more in N years with a PhD or N+5 years with an MS, the math gets depressing fast.

The bloody fact is that we are human, and life isn't just an excel spreadsheet. Take my own timeline as an example: doing an MS at UMich cost around $80k a year. Then, I spent two years (2024–2026) grinding as a Research Associate, still needing around $30k a year in family support just to survive. Meanwhile, peers who chose the industry path earned at least $10k in their first summer internship alone, and upon graduation, stepped into salaries of $150k+ per year (and frankly, that is a conservative estimate).

To put it bluntly: it is financially stupid to pursue a PhD (at least in your 20s). Do not do it for the money.

Do not assume there is a guaranteed formula.

In fields like HCI, NLP, and Speech Processing, applicants usually have at least one first-authored paper in a top conference. Yet, every cycle, you will witness applicants with zero publications getting swept into top CS PhD programs.

Then there is the networking game. Many people successfully cold-email prospective PIs before getting an offer (in fact, for the only two offers I got in the 2026 cycle, I spoke to both PIs before submitting my application). But in every top school, there are also students admitted by PIs they had absolutely no prior contact with. We all guess that the Letter of Recommendation (LoR) is the secret sauce, but since we never see the contents, for all we know they could be a myth.

What I want to say here is this: you still need to maximize your chances. Publish strong papers, connect with PIs as much as you can, and try to get glowing letters from famous professors who actually know you. But understand that none of this guarantees an outcome. You cannot directly compare two applicants. Applicant A can have significantly more papers than Applicant B, and it is still highly likely that B gets admitted somewhere A got rejected. It feels incredibly unfair, but it is the truth of the system.

Do not STOP.

At the end of the day, I have to remind myself that a PhD application is just the starting line. The goal was never just to "be admitted as a CS PhD student." The goals are to "graduate," "deliver impactful research," "find a meaningful job," and, hopefully, "live happily."

The application process gave me a lot of painful memories, and I am still recovering from that grinder. But I always look back to 2021, when I first decided to pursue this path. I remember the role models I met along the way, and I strive to become the researcher I always hoped to be. I finally got admitted after five years, but the wait was just part of the process of achieving that larger goal.

The PhD application is just a small chapter. If you did well, that's great. But if you didn't (like me)—hey, it's okay. We will find a way out.