Always Be EPIC: A Formula For Approaching Your Ph.D. Work

Sauvik Das
11 min readAug 5, 2019


I’ve talked about how getting your Ph.D. is going to be challenging, that it’s fraught with emotional lows, and that you are not alone in harboring feelings of self doubt.

I’ve painted a bleak picture. You might be feeling anxious. Maybe I’ve taken some of the excitement you were feeling about starting this journey and converted it into nervousness. But, look, I really enjoyed my time as a grad student. There were definite low points, but, on average, I was very happy to be doing what I had the opportunity to be doing. I’m grateful to my friends and advisors for that. The Ph.D. grind is not all bad, but it’s important to realize that it is a grind. Once you are aware of that, you can steel yourself and get ready to work.

But, I’ve said nothing yet about the “work.” You might be wondering “surely there has to be some practical advice out there about how to improve my chances at success.”

Of course! There is as much advice out there on how to do well in academia as there are academics. All of us who go through The Grind have our own ideas about what it takes to do well.

This post will highlight my perspective — distilled from years of cogitating on my personal experiences, the conversations I’ve had with mentors, and reading posts like this one written by those who came before me.

Standard disclaimers apply. My advice is not The Best. View it as one part of your own process of figuring out what works. I mostly wrote this for my Ph.D. students— i.e., students who work somewhere at the intersection of HCI, Applied Machine Learning and Cybersecurity. That said, I’ve tried my best to make the advice broadly applicable, practical, and actionable.

Your most important job, as a Ph.D. student, is to do excellent research. You have other responsibilities, of course, but your fundamental measure of success is your research.

So, what can you do to make sure that your actions are in line with this fundamental measure of success? Salespeople have a motivational saying: “Always Be Closing”. It is a pithy way of capturing a mentality of relentlessly working towards their fundamental measure of success: selling a product or service. Like you, I don’t like people relentlessly trying to sell me something, but I do admire the grit behind this expression — it’s a clear, guiding star that can help you when you get stuck in the weeds of your day-to-day. What’s the equivalent for academic research?

Let me propose one: Always. be. “EPIC”:

  • Executing: Channeling energy and excitement to finish what you started
  • Practicing: Employing deliberate practice to expand your skillset
  • Ideating: Exercising your creativity by generating new ideas frequently
  • Communicating: Sharing your work and your voice often and clearly

I definitely did not spend a bunch of time trying to come up with that backronym. Let’s unpack each of those.


To do excellent research, you need to to be an excellent finisher: someone who can take a good idea from hazy and amorphous to a falsifiable hypothesis and/or proof-of-concept prototype quickly and effectively. You need to be someone who can wade through the muck of the “trough of disillusionment,” make it through to the other side, and then do it all over again with a renewed enthusiasm.

‌I don’t just mean “work hard.” To generate impactful results, hard work is certainly important. ‌But, hard work is not the only thing that is important.

Then, how does one “execute”? In a nutshell, you must be willing to dedicate enormous amounts of energy to what you do. You and your advisor should find problems for you to work on that both of you think are important and exciting. You should wake up and think that what you are doing is meaningful and harness that energy to drive your research. You need to do this because research is hard, and you will get stuck. When that happens, you should be prepared to dig into this original motivation — this relentless belief that what you do is interesting and important — to unstuck yourself.

I promise my students that I will only ask them to work with me on projects that truly excite me, and to which I am willing to invest my time and energy. I ask them to do the same.

‌Be reflective: if you notice yourself slowing down, talk to your advisor! It happens to everyone. Trust me, your advisor will prefer to hear about this sooner rather than later — it is much easier to pivot at the earlier stages of a project.


Research is fundamentally about exploring what has never before been explored. No two research projects can be the same. So, as a researcher, you must be committed to expanding your skillset, learning and growing.

Writer James Clear defines deliberate practice as “a special type of practice that is purposeful and systematic. While regular practice might include mindless repetitions, deliberate practice requires focused attention and is conducted with the specific goal of improving performance.”

‌Deliberate practice underlies intentional attempts at growth. As you gain research experience, you will become a better researcher. But, that doesn’t happen automatically. That happens through deliberate practice.

‌It’ll be easier to employ deliberate practice at first: everything about doing research will be new and so you will grow out of necessity to handle the new stressors. But there will come a point — maybe not immediately — where you will have acquired a solid understanding of how to do research and how to publish papers. When you get to this point, it’s easy to become complacent. The easiest indicator of that is when the new ideas you think of pursuing are not that exciting to you but seem like solid ways to get another publication. You need to push through the temptation of staying in this comfort zone.

‌Here’s an example. When I started graduate school, I knew very little about data science. Facebook hired me as a software engineering intern in my first year as a Ph.D. student. But, recognizing a unique opportunity to expand my skillset for the betterment of my research career, I asked them to put me on the data science team instead. I worked under a social psychologist, Adam Kramer, and in one summer, I learned quantitative data analysis skills that continue to pay off to this day. Indeed, those skills enabled me to conduct research that made up some of the most important pieces of my dissertation research.

‌Deliberate practice could take on many forms. It could mean taking a course on a subject matter that interests you, or learning how to prototype with hardware, or learning about qualitative research methods.

Failing is an important aspect of deliberate practice. In weight training, there is a concept known as “training to failure” — that is, continuing with a particular exercise until you cannot finish your last repetition. This is one of the best ways to increase muscle mass. It’s not hard to apply this concept to your own practice. When learning a new skill, practice to the point where you fail. That’s the edge of your ability and that’s where you need to push. Having difficulty writing a paper? Go to The Most Dangerous Writing App, set yourself a 5–10 minute timer, pick a subsection that you need to get started on, and write.

More than anything, deliberate practice means committing to growth.


I mostly stole this from Michael Bernstein and it has been one of the most impactful pieces of advice I’ve received for doing good research. Early in my Ph.D., Michael, fresh out of grad school and on his way to start at Stanford, told me that the single most important piece of advice he could offer was to set aside 20-minutes each day to brainstorm.

Lot’s of others echo similar advice. Seth Godin, who runs of the most popular online blogs and has been publishing daily since the aughts, writes: “you can’t have good ideas unless you’re willing to generate a lot of bad ones.”

‌Nobel laureate Linus Pauling said: “If you want to have good ideas you must have many ideas. Most of them will be wrong, and what you have to learn is which ones to throw away.

‌Consider the following simple observation: the “goodness” of your ideas is normally distributed. Most of your ideas are going to be average. A rare, few, however will be good, and fewer still will be So Good That They Will Define Your Career. If you want to have a career defining idea, then, there are two ways you can approach it: (i) luckily stumble upon it in your first couple of tries or (ii) generate lots of ideas. I like luck. I don’t like to rely on it.

If you generate enough ideas, you are likely to have generated a few that are very good. The trick, then, is to learn how to pick those out of the batch. Your advisor should be able to help you with that.

‌A few corollaries to this:

‌1) As you gain more experience doing research and practicing deliberately, you should be shifting the “goodness” distribution of your ideas to the right: i.e., the average “goodness” of your ideas should continue to improve. That’s growth! However, the shape of that distribution will remain the same: your best ideas will be few and far between, and the only reliable way to extract them will be to generate many ideas. This is true for me, you and everyone else. You will only be able to pursue a handful of projects at any one time, so make sure you are pursuing the best ideas that you can generate.

‌2) Establish a routine of brainstorming every day. Your internal idea generator is a muscle: to strengthen it and prevent it from atrophying, you must exercise it regularly.

‌3) Don’t be too critical of your own ideas at first. This is very hard for researchers to do, in general, and is especially hard for new researchers. Why? Because when you start out, there’s a gap between your taste (which will be excellent) and your skill (which you will be honing). Refer to my post on the shape of your Ph.D. for more detail on this.

‌4) Assessing whether or not an idea is good is a skill you will develop with practice over time. In fact, it is one of the most important things you will learn as you get your Ph.D. Bob Kraut once told me that the hardest and most important part of good research is in asking the right questions. Many times, we get stuck finding the right answers to the wrong questions. The best way to assess if your idea is good is to solicit feedback from others. This is one way you the other Ph.D. students you meet in your lab, university and conferences can be force multipliers for each other: commit to engaging with your colleagues on seedling ideas. Adopt a mentality of constructive feedback: if you need to be critical of someone else’s idea, clearly articulate how it might be made better. Conversely, you do no good being a yes-person (i.e., an uncritical friend who will convey positivity no matter how they actually feel). If you have concerns about an idea, articulate those concerns. This feedback is how we improve.

5) Find inspiration in the world around you. Reading is a great way to find inspiration. You should read as many academic papers as interest you, but don’t limit yourself — non-academic writings can provide just as much fodder for a great idea. Art can be a particularly fruitful outlet for finding inspiration. My first Ph.D. student, Youngwook, pitched his starter project to me because he was inspired by the Spiderman movies. Incidentally, this is one of the reasons why it’s very important to strive for balance between your academic and non-academic pursuits. Giving your mind a chance to disengage from your research and focus on other things is, counter intuitively, one of the best ways to find breakthroughs and “A-ha!” moments.


Consider the age-old philosophical thought experiment: “If you find groundbreaking results in the middle of a forest, does your research still make a sound?” I think it goes something like that, anyway. A big part of research is communication — writing papers, giving talks, and even tweeting.

Scott Hudson once told me that research papers are your most important academic currency. There’s definitely truth in that: your papers are an attempt at formalizing what you’ve done, how you did it, and why it’s important. More generally, the most important thing you can do to ensure your research has lasting impact on the field is to publish at top-tier conferences and journals. To publish, you should learn how to write well. Here’s a short blog post I wrote in which I compiled a list of resources for academic writing in HCI. The best way to improve your writing is through deliberate practice.

However, publishing is not the only way to communicate your research (and it is not the only way to have impact). There are many ways to both practice your writing and communicate your research. A few examples of alternative communication mechanisms you might explore include: writing a blog (hi), tweeting (hi), creating YouTube videos (hi) and responding to relevant questions on Quora (never did this actually…). These are just examples: you don’t need to do all of these things, or even most of these things. But you should explore as many channels of communicating your research as you feel comfortable exploring.‌

Think of it this way: with every paper you write, every thought you tweet and every blog post you share, you’re purchasing a lottery ticket. With each ticket, you have a small but hopeful chance that your work finds an audience with which it will resonate and have impact. The more tickets you purchase, the more likely you are to “win the lottery” and find the right audience for your work.

At a minimum, I recommend: (i) writing a blog and (ii) maintaining an active Twitter presence. Writing a blog can seem intimidating, but the trick is to not worry about being original or insightful. Look at the stuff I write! It’s unlikely you’ll write about things others haven’t already written about, but writing blog posts can help you practice long-form technical writing with a tighter feedback loop (instead of months from written-to-published, it’s seconds). If you establish a consistent writing routine, you’ll find it’s easier to write other things as well —e.g., papers, essays, grants.

Maintaining an active Twitter presence is more pragmatic. For some reason, Twitter is the academic social media of choice. A good way to stay connected with your community is by engaging with other researchers on Twitter.

You don’t need to internalize all of this. But, if you get stuck, just remember: Always be EPIC. If you’re executing, practicing, ideating and/or communicating, you’re probably doing something right.

Thanks for reading. If you read this and thought: “whoah, definitely want to be spammed by that guy”, there are three ways to do it:

  1. Follow me on Twitter (spam frequency: daily — weeklyish)
  2. Subscribe to my YouTube channel (spam frequency: ¯\_(ツ)_/¯)
  3. Subscribe to my mailing list (spam frequency: In Theory? Once a month with links to whatever I’ve written. In Practice? Once a year, when I feel bad.)

You also don’t have to do any of those things, and we will both be fine.



Sauvik Das

Assistant Professor of Human-Computer Interaction at Carnegie Mellon University. Formerly at Georgia Tech. Ph.D. from CMU HCII. HCI, Security, Data Science.