from Guide to the Job Hunt on Jan 15, 2023

How to make big decisions.

Indecision tends to sound the same, no matter the situation: Company A is offering more, but I love the team at B. Oh, and C is closest to home, but … You're spinning in endless circles, and the underlying cause? You're flip flopping between criteria and rankings — two ideas you should separate.

A decision is "small" if it's easy to commit, re-evaluate, and switch decisions in a short period of time. For example, picking a YouTube video thumbnail falls in this category: Try a thumbnail, see how your video performs over the next minute, hour, or day; then, try a different thumbnail.

A decision is "small" if it's easy to commit, re-evaluate, and switch decisions in a short period of time.

On the other hand, a decision is "big" if it takes a long period of time to assess that decision. For example, a job takes probably 6 months to get trained, know your coworkers, and define your role within the team. At that point, you can finally say definitively if you enjoy your job and team.

In any big decision process, the root cause of indecision is two-fold:

1. Overload: You have too many jumbled thoughts in your head to comprehend. There is a slew of information: Which option excites you most, how your friends and family weigh in, timelines for every option, unanswered questions that you need answers to, and many more. In short, you're feeling overwhelmed.
2. Disorganization: In an effort to sort through the mess, you skip from option to value to option. One option excites you, but another option offers higher compensation. However, the third option has a generous vacation package, which none of the other options have. In short, you're thinking in circles.

The above criteria are both hallmarks of indecision; if you feel overwhelmed and start thinking repetitive thoughts, this is a red flag: It's time to sit down, and follow the advice below.

Break down your thoughts: Write then organize.

The solution to the two problems above — being overwhelmed, and thinking circular thoughts — is fortunately fairly straightforward. Basically, break down your thoughts into pieces, in the following way:

• Write your thoughts down. Offload information from your head onto paper, then organize your thoughts in the next step — instead of struggling to remember and organize simultaneously. Right now, write down at least 3 thoughts related to your decision. Any relevant thought counts, no matter how random.
• Organize thoughts by criteria and ranking: (1) Your criteria are the rubric items that you'll make a decision by. These can range from excitement to swag. (2) For every criteria, rank your options, not score. In short, assigning absolute numbers is hard (Is this a 10? A 9.5?) and unimportant, given your goal is to pick the best option — not grade all your options. So, don't score. Rank the options. Right now, write down a criteria, and rank your options for that criteria. Ties are acceptable.
• Optionally quantify your rankings. If an option ranks the highest for a criteria, assign a score of 10. If the second highest, assign 9 — so on and so forth. Then, sum scores across all criteria. Finally, sanity check the numbers: Does the ordering roughly make sense? If not, increase the weight of more important criteria, and re-add up all the scores. For example, if mentorship matters to you more, double the values for that criteria. Right now, convert rankings to scores, and sum up scores.

The above will help you sort through your thoughts, but at the end of the day, there's one meta-criteria that takes precedence over all else: minimize regret. To minimize regret, trust your gut. Really think deeply about whether or not this decision is one you believe in. If your gut is revolted at the thought before you even commit, that gut feeling will only intensify as time goes on.

1. Understand your gut: If you don't know why your gut is angry, use the criteria and rankings to understand your gut — maybe there's a criteria you care about that you're afraid to admit. Maybe prestige? Job security? Work-life balance? Diversity? Maybe your rankings reflect other people's opinions, and not your own. Or, they reflect what you think should be correct.
2. Trust your gut: If you know why your gut is against your decision, do your best to reason with it. If your gut isn't convinced, stand up for your gut. If you don't stand up for your decision now, it'll be even more difficult when you've sunken time and effort into committing to a decision you regret, later on. If your gut tells you no, it's a no.
3. Check by pretending: One way to check with your gut is to pretend you've decided, for 24 hours. Are you excited to share your decision with the people you love? Are you proud of your decision? Did you feel more and more that this decision is "right," as you think about it?

To understand your gut, you may also use the values you distilled from your thoughts in How to make New Year's resolutions..

Now that you have a rough guideline for decision-making, I'll walk you through my own process, when I was deciding on my next steps after my undergraduate. Most importantly, I now have 20-20 hindsight, 5 years later, to assess my then-assessment of each option.

How I broke down my thoughts, for my job hunt.

In my senior year of undergraduate, I applied for companies big and small, as well as the PhD. This left me with two major sets of decisions to make. There are certainly other ways to split these options, but at the time, these were top of my mind. Here were my thoughts at the time:

1. Academia vs. Industry: At heart, I'd always been at home in industry, and I knew that I preferred engineering to more school. Yet, at the same time, I was curious about the PhD, and research seemed to be the only reprieve from being a code monkey, at the time.
2. Big vs. Small company: I had loads of fun at my only big-company internship, and there were plenty of mentors to look up to. However, the work wasn't satisfying. My only full-time small-company experience was neither fun nor satisfying, so I wasn't fond of startups either.

To answer the above questions, I used both on-sites and visit days to make sense of my gut feelings, about each of my options. Here are summaries of my thoughts, after each of my on-sites and visit days. These thoughts are largely jumbled, but they accurately reflect my confusion at the time:

• Aeva: There were some interesting interview questions — sparse vector representation and inner products — but the CTO missed the closing interview, so I wasn't able to ask high-level questions. It left a bad "first" impression, and since the company was focused on LiDAR, it seemed like software took the backseat.
• Lyft L5: Interviews were very interesting — a custom data structure, a "Trapping Rain Water" puzzle — and I had very interesting conversations with potential coworkers and various managers, over lunch. The team seemed strong, and the entire organization was early in its development.
• Nvidia: Interviews were the closest to the work I wanted to do — applied problems for autonomous vehicles, specifically for computer vision. I distinctly remember my potential manager being very charismatic, but for whatever reason, I wasn't able to meet my future teammates in-person. My gut was not terribly excited.
• Rex: I was incredibly impressed by how friendly and knowledgeable the CEO, future CTO, and engineers were — not to mention how willing they were to spend time talking to a candidate during the onsite. This was the only company to support me pursuing a PhD simultaneously. I was excited about the company, the vision, and the leadership.
• Berkeley PhD: Visit days were underwhelming. The vision faculty were not particularly interested in talking to me, and the faculty I did get to speak with didn't share many interests. However, there were many PhD students I got to work with and looked up to — Vlad, Bichen, and (now MIT prof) Cathy, just to name a few.
• Google AI Residency: This was one of my top choices, and this was by far the most exciting offer I received that season. I enjoyed hanging out with the other candidates after our on-sites, but we weren't a perfect fit. More importantly, everyone at the on-site — candidates, interviewers, organizers — all asked "Why consider the residency, if you already got into a PhD program?" I didn't know PhD programs were ultimately the goal, and I also didn't have a good answer.
• CZ Biohub: This was the most unique of my options at the time. I got along extremely well with my would-be manager. Furthermore, Prof. Terry Johnson and (now prof.) Ashwin Pananjady voted for this option. I was worried that machine learning would take a backseat to biology, however — as it should, given the organization's goals.

You can see above that my thoughts were fairly jumbled, so at the time, I sat down to enumerate my criteria. Here are the criteria that I extracted from my thoughts above:

• Excitement - How excited I am for the opportunity. This includes how much pay, potential living conditions, free food, swag, and other minor things that sway my gut desire.
• Connections - The potential for connections and many varying institutions are involved. Lower score if my connections in this place are already established. Focus on opportunity for more connections
• Learning - How much I'd learn. Is this a new field? Will I start from ground zero?
• Personal time - How much time I could have to myself, to work on side projects and play with my own ideas. Similar to work-life balance but also reflects how tightly employer tries to control my free time.
• Mentor - Are there people that I could or do look up to? Are mentors people I've already met and I know I get along with?
• Inertia - Do I already know what I'm doing and how to get around? Do I already have connections?
• Location - Is the location in a desirable and familiar location? How close would I be to friends and family? Staying in the bay area is a priority.
• Priority of Expertise - Is my expertise and work a priority within the company? Using 3 for top priority in the critical path, 2 for high priority, 1 for supporting team, 0 for bottom of the totem pole. Note: For example, computer vision is top priority at Tesla, as their car is vision-only, as of time of writing. On the other hand, computer vision is second priority at Waymo, as processing LiDAR and radar takes precedence.

Then, I tried to quantify each option, roughly following the guidelines above. When scoring each option, I tried my best to assign an absolute number between 1 and 5.

Per the ratings above, REX and Google AI Residency were tied for first. This fairly accurately reflected my gut at the time. However, REX allowed me to do the PhD at the same time, so I summed scores for the PhD and REX options together. With that, committing to REX and the PhD simultaneously was the highest-scoring option. Furthermore, I had resolved my two dilemmas:

• Academia vs. Industry: Fortunately, this decision allowed me try both industry and academia at the same time. I actually didn't think I'd finish the PhD, when I committed to starting.
• Big vs. Small Company: Ultimately, I favored big companies. However, that was the first time I had found such convincing startup founders — the first time I had heard a concrete plan rather than sky-high ideals and bigger-than-life solutions with no problem statements.

In the above table, you'll also find that the "Priority" criteria ranges from 1 to 3. This was my way of down-weighting "Priority," as that category mattered less to me. You may do the same for your own criteria as well, when tallying up scores per option.

It's now been 5 years since I made that decision. In the interim, I've finished the PhD, although REX shut down in 2022. Nevertheless, I'm still extremely happy with my decision, as I've picked up a number of invaluable skills from both the PhD and REX alike — all of which are integral to my job today. Due to some dumb luck, I've even dodged a few bullets.

Knowing this, I re-employed the same criteria-ranking system when deciding my first job after the PhD. To date, I've been at the new job for 6 months and am still happy with this life decision as well. In light of the outcomes of both job hunts, I wrote up this post to hopefully help you find the same satisfaction in your decision-making.

Don't freeze: No "wrong" decisions, and decide together.

The first evil above is information overload, which we solved already by writing down and organizing our thoughts on paper. There is actually a second evil in the decision-making process: paralysis, stunning you into a motionless stupor, with the gravity and stresses of a life-changing decision. To help with paralysis, I have two tips:

1. Every decision is a step towards your best option, so there is no "wrong" decision.

   1. Continue to iterate: Earlier, we defined a "small" decision as one where you can quickly commit, re-evaluate, and switch decisions quickly. By nature, "big" decisions make iteration slower. However, even and especially with big decisions, it's important to make sure you're still iterating, even if you're iterating over months and years, instead of days.
   2. Iteration is progress: Commit just long enough to re-evaluate your decision, but switch decisions when you've deemed it appropriate. So long as you commit to this process, every option you try is progress: For everyone option you eliminate, you can narrow down the remaining options and increase the odds that your next choice is the best one.

2. Consult others for criteria and rankings — anonymously, if you want. Advice doesn't need to come from friends and family.

   1. Anyone can help: Consider reaching out to experts in your field, relevant leaders, and peer advisors. A new perspective is advantageous and maybe even preferred, if your decision is a sensitive topic among your inner circles. The key is to realize that asking for advice can happen anonymously too. Let's say for example that there is sensitive information that cannot be disclosed, for your decision. In this case, seek relevant criteria and rankings advice from sources far and wide.
   2. Example: other faculty may have previously blogged about factors to consider when accommodating students with disabilities. Maybe a primary school teacher on the other hand tweeted about reasons to prefer one approach over another, when observing her own student's reactions to certain policies. Now that you know what information to look for, when making decisions, any source of information is invaluable.

In short, no decision is "wrong", and you can consult others, even anonymously.

For one last tip: Trust yourself. Trust your future self to try your hardest and make the most of your situation, no matter the decision and no matter how big the big decision is. Ultimately, you minimize regret not by making perfect decisions but by doing your best — both doing your best to make informed decisions, and by doing your best with the decision you made.

← back to Guide to the Job Hunt