Well, how does tech company campus recruiting work?

Tech companies go to college campuses with the goal of interviewing as many qualified people as possible, but they can only interview a limited number of people. So the name of the game here (assuming you can hack through the interviews) is to get your foot in the interview door. This means you need to stand out in a sea of resumes.

How do you do that?

• Resume builders
• Networking

Resume builders

Resume builders are things that few people have on their resume — impressive awards, a portfolio of sophisticated personal projects, teaching appointments/being part of a course staff, significant open source contributions, etc. They generally should be either directly related to, or adjacent to, your field.

If you can get a personal project off the ground and get some real traction — e.g. app store downloads, meaningful web traffic, etc — that will get noticed because delivering customer value is the ultimate challenge for most tech companies and this shows you can do it all.

Networking

The other avenue is networking. To understand how this works, it helps to know a few things about how tech company campus recruiting generally works.

Tech companies send delegations to campuses to:

• Give tech talks
• Host booths at career fairs
• Find the best students recommended by professors and previous interns

To win at this game, leverage all three of these avenues to stand out.

Attend tech talks and talk to the speaker afterward

You go to the tech talk not (just) to listen to the content of the talk but to meet the engineers at the company. Go early and talk to the speaker; stay after and ask questions about the talk. More importantly, ask questions about the speaker — their experience, what they like about the company, and what tips they have for succeeding in the interview process. Remember: they want to hire people, so they are very happy to tell you everything they can that will help you (short of giving away questions and answers).

Have your resume with you so they put a face to the name. While this won’t guarantee you an interview, in a sea of resumes, it will help you stand out. It’s surprising how relatively few people stick around after talks.

Use career fairs to practice (but don’t expect much)

Career fairs can be exhausting and overwhelming for both students and companies. This means that they are not likely to be your primary source of interviews. Companies meet hundreds of students, and it’s rare that you get a lot of time with an employee of a company. They often need to get through a long line of students and will have stock answers.

However, the career fair is a chance to practice the kinds of interactions you want to have at tech talks or during other parts of the interview process — at best, you might get lucky and stand out; at worst, you’ve gotten in some reps to practice what you’ll need to do when you attend tech talks and when you land the interview.

Be someone others want to recommend

This is the turbo-charged way to stand out to recruiters. You want them to come to you, rather than you going to them. The way to do this is to stand out to your peers, particularly your peers who have previous internships (i.e. upperclassmen a few years ahead of you) or professors connected to tech companies. (This, in fact, was how I ended up in my first job out of college.)

There are all sorts of ways to do this, but most of them amount to getting involved in campus activities that expose you to these groups — clubs, particularly those that do real project work together, taking classes with group assignments where you can stand out as a great partner, helping a professor with their research, or joining the course staff for an introductory class are all ways you can meet the people who will be able to help you.

Again, it helps to remember the incentives here — companies value references from people they trust; the students (or professors) referring you want to help both you and the company. If they can share a great candidate with the company, that’s a great reputation builder for them too. So it’s a win for everyone.

None of these approaches are guarantees — you should still do all the typical things people suggest like pounding the pavement to apply to multiple jobs, being persistent, etc — but knowing how the campus recruiting game is played can help you maximize your opportunities.

Bonus: what about the second internship?

Once you’ve gotten that first internship, you’re in a much better position to get the next one. Ideally, to keep on schedule you’d land your first internship sophomore year and parlay that into a second internship (ideally at another company) that results in an offer to return for a full time position. Now, you’re in great shape going into senior year — you can interview and build your options without the stress, and you have a great baseline for comparison.

If you fall off this track, double down on the above techniques and be ready to explain how your sophomore summer experiences translate into the skills you’ll need on the job.

It’s possible that the only thing engineers hate more than a nasty bug is interviewing. Interviews are interruptions from deep work, involve a somewhat awkward set of interactions with a complete stranger, and after you’ve done a bunch, they become pretty monotonous. They’re everything most engineers were trying to escape by becoming engineers.

Yet engineers at high growth startups will easily spend 10%-20% (or more!) of their time interviewing (do the math: 2 interviews at 1 hr interview + 1 hr prep is already 4 hours). At hyper scaling companies, some engineers will get tapped to do a tremendous number of interviews because it’s the highest leverage activity they can do. For example, Albert Ni interviewed every full time engineering candidate at Dropbox for about a year between 2011 and 2012, just as Dropbox’s engineering team was hitting an inflection point in its growth.

Impactful work done by resentful people is not a recipe for success.

Sure, we can optimize the hiring funnel or otherwise reduce the need to hire — but it’s never going to go away. So let’s tackle the other side of the problem: how can we make interviewing just a little less awful?

One big problem with interviews is that most interviewers pretty quickly get to point where they don’t feel like they’re learning much doing interviews. At the start, the idea of doing interviews can be fun and exciting — it’s a new skill, and an important thing for the company. But once you hit a state of basic competence, it becomes a boring grind.

The thing is, the skill range for interviewing is much wider than that — some people are far better interviewers than average, people who single-handedly make others want to work with them. But most folks don’t really have a clear sense for what getting there might look like.

And by the way, developing as an interviewer isn’t just about work. It’s also a useful life skill — if you’re hiring an event planner, having work done on your house, or otherwise need someone who knows how to do something you don’t, you’re going to have to evaluate people.

This skills ladder for interviewing provides a way to think about how to grow as an interviewer.

There are two critical skillsets:

• Attracting talent — making the candidate want to work with you
• Spotting talent — accurately assessing whether you want the candidate to work with you

Interviews are a two-way evaluation, and every interviewer has the opportunity to reinforce — or ruin — a candidate’s positive impression of a company.

Across these skills, there is a natural progression from beginner to expert.

The Beginner

At the very earliest stage, an interviewer’s first and primary goal is to avoid a bad candidate experience, handle the social graces and customs of an interview and create the right conditions to evaluate a candidate. Have you ever left an interview confused about what the interviewer was even trying to get out of it? That’s the experience of interacting with a beginner.

Beginners can be actively harmful in the hiring process. I was once interviewed by two people, one of whom said literally nothing for the entire interview. (They were shadowing the question to learn it. An introduction would have been nice!) As I struggled with the question, I kept wondering, “is this person being quiet because I’m just doing this all entirely wrong?” The overhead of thinking about the interviewer was distracting enough to make an already-challenging problem much more difficult. It was the most stressful interview I’ve had, as well as one of the most memorable.

Beginners are learning their first interview questions and how to write up a clear summary of their assessment. As a hiring manager, when I see interview feedback focusing on candidates with sloppy handwriting or not knowing the exact syntax for a particular programming language construct when writing on a whiteboard, I take that as the sign of a beginner.

Fortunately, most beginners know they are beginners — interviews are just as awkward and difficult for them as for the candidate.

Reaching competency

As an interviewer develops competence, they move from being a pure liability to, sometimes, an asset — creating neutral to positive experiences for candidates, keeping the interview on schedule, and engagingly explaining their work at the company.

Competent interviewers conduct unremarkable interviews for most candidates. The problem with competent interviewers is two-fold. First, they’re not likely to spot unusual talent, which means you might miss out on star hires. The second is that your competitors might have more compelling interviewers that make candidates actively want to work for them.

Coming out of college, I interviewed at several of the big tech companies of the day. I remember almost none of my interviewers, either positively or negatively, because they were competent but not exceptional. Competent, but not exceptional, interviewers means the company looks competent, but not exceptional, too.

Competent interviewers will have clear write-ups and can handle most of the normal paths an interview question might take. If the candidate has an unusual approach, though, that might spell trouble for them even if the candidate is strong. As a hiring manager, when I see competent interviewers writing feedback for candidates coming from typical tech companies, I have confidence that they’ll evaluate them well. It’s when the candidate’s background is a little unusual that I start to worry that we might be missing things.

If you’re hiring for a high growth company without a strong tech brand yet, this is a problem for you. You are going to have to find strong candidates who may not have fancy schools or companies on their resume, and you’re also going to have to fight hard to make those candidates want to work for you. Competent interviewers might work at established companies, but they’ll kill you at a smaller company.

Competency is often where an experienced interviewer will end up if they don’t work to get to the next level. One reason is simply that most interviewers get very little feedback, so it’s easy to get stuck here. If interviewing is boring, and you aren’t learning, there’s a good chance you’ve gotten stuck in this bucket.

Gaining proficiency

Proficient interviewers are where you start to separate from the pack — creating compelling candidate experiences, connecting with the candidate over common ground, and building rapport through honest answers to tough questions. Proficient interviewers are also better at evaluating candidates — they’ll know when someone is just taking an off-beat track and how to probe the candidate in a way that feels constructive, rather than aggressive. When a candidate is spinning out, they’ll know how to pull them back and give them a shot at showing what they can do.

You’re more likely to remember interviewers who were proficient at attracting talent, even if the interview itself wasn’t memorable. Why? Because you’re more likely to accept a job from a company where you like and connect with the interviewers.

One way you can tell you’re proficient is that recruiters and hiring managers try to find ways to include you in their interview loops because they know you’ll both get the candidate excited and be able to assess their work fairly, even if they take an unusual approach.

As a hiring manager, I can tell a proficient interviewer by whether they’re able to really get at the “why” behind their decision in the interview — are they able to spot a constellation of one-off data points that make up a trend, or notice when the candidate really shines in an area the interview wouldn’t normally touch on?

Honestly, this is probably a fine place for the average interview to be, and it should be just a bit more interesting because you’re starting to evaluate candidates not just against a script but engaging more with them as the complex humans they are.

Becoming an expert

At Dropbox we had one interviewer candidates would consistently name as the person they met during the interview process that they really wanted to work with. It was amazing — and it worked really well for her because she’d sometimes end up having candidates ask to join her team instead of the role they were initially interviewing for. One coworker I talked to about this person said that he literally stopped interviewing at other companies after he met her. She brought the incredible energy and candidate connection of a true talent magnet.

Think of the superpower it gives your company when there’s someone who attracts others like that.

Expert interviewers are often the reason a candidate joins a company. Often, they spot a candidate that might otherwise have gotten rejected but has some hidden strengths — several times at Dropbox I witnessed an eagle-eyed talent scout push for a candidate who became an exceptional hire, and one that we’d have missed out on without that voice. They’re also talent magnets — people want to work with them because they bring tons of positive energy, make a deep connection with the candidate, or otherwise show that they’d be a fantastic coworker. They might look like naturals, but most experts have worked hard to hone their pitch and work hard to build early rapport with a candidate.

Coming out of college, a major reason I chose the completely unknown startup I went to work for was that one of the engineers was so smart, energetic and engaging in the interview process that I knew that I would learn a tremendous amount from him. (And I did!) We bonded over our mutual love of the book Programming Pearls and I was won over by his infectious energy and enthusiasm. Remember how I said competent interviewers at startups can kill you? Expert interviews can make all the difference.

When I interviewed at Dropbox, I got excited about the company even before I came onsite because my primary contact was someone who was incredibly welcoming and enthusiastic to talk about what Dropbox was up to. We ended up with multiple back-and-forths about Dropbox’s technical stack and challenges. The more you can share that really excites a candidate’s mind, the better. Of course, when I got on-site, the conversation naturally continued from there.

One key trait of an expert talent magnet is deep empathy for where a candidate is coming from. The ultimate goal when recruiting someone is to make it easy for them to imagine themselves working with you and having a great experience. Part of doing that is deeply understanding what they care about (which may not be what they say they care about) and avoiding all the little miscues that can turn a candidate off.

How does it all fit together?

While the descriptions above capture most of the key behaviors, there are a lot of specifics. The sections below lay out a more detailed breakdown and show how each level of skill compares.

If you’ve gotten bored with interviewing and feeling like it’s just a grind, make a game out of improving:

• Find a bullet point you’re not good at
• Work at improving it, until you’ve mastered it
• Find another bullet point to tackle
• Collect all the bullets within a level and then move up!

The ladder

Attracting Talent

Beginner

• Learning how to create a neutral to positive candidate experience throughout the interview while evaluating the candidate — sometimes negative reactions may leak through
• Handling the basic social graces of the interaction — physical comfort, greetings and candidate handoffs, eye contact, etc.
• Learning how to talk about the company and the interviewer’s role in an engaging way — e.g. may get into too many details

Competent

• Ability to create a neutral to positive candidate experience for most candidates
• Keeping the interview on a schedule that allows the candidate to reach a sense of closure at the end, even if they struggled with the question
• Able to convey an engaging description of the interviewer’s own role and reason for being at the company
• Able to avoid sending out negative vibes to struggling candidates

Proficient

• Ability to create a consistently positive candidate experience
• Able to connect with candidates over common interests or experiences
• Can answer more challenging questions from candidates (e.g. about the negatives of a company) without improperly setting bad expectations for the candidate
• Ability to consistently frame questions and probe the candidate on their knowledge in a non-confrontational way that feels engaged, collaborative or curious

Expert

• Talent magnet — able to get candidates excited about the role, the company, and the potential to work with the interviewer
• Often brings a great deal of positive energy to the interview
• Usually good at sussing out a candidate’s interests and connecting with them on a personal level or using that knowledge to frame the role or company
• Well aware of what does — or doesn’t — resonate with candidates and adjusting the pitch over time

Spotting Talent

Beginner

• Learning to create the conditions for an effective evaluation
• Learning some core interview questions and the rubric for evaluating those questions
• Able to set the right up-front expectations for the interview structure and expected nature of answers so that the candidate is set up to shine
• May let candidate spend too much time in the wrong areas of a question
• Able to capture high level notes and summarize the flow of the interview along with a tentative conclusion
• Likely only peripherally aware of possible unconscious biases

Competent

• Able to evaluate typical candidate responses effectively
• Solid grasp of several core interview questions and ability to consistently apply the rubric to them
• Effectively managing time by keeping the candidate from getting bogged down in the wrong places or for the wrong reasons
• Knowing when to hint, and when not to, to set the candidate up for success
• Taking thorough notes without detracting from interactions with the candidate
• Summarizing the results of the interview into a clear, evidence-based write-up that explains the rationale for a yes or no
• Awareness of own unconscious biases, and factoring into decision making

Proficient

• Able to evaluate non-standard candidate responses effectively
• Strong grasp of a number of interview questions and their rubrics, as well as the ability to consistently apply a meta-rubric of general expectations across all interviews, and use it to evaluate candidates who go take unusual approaches to a question or who demonstrate patterns of behavior, either positive or negative, that aren’t captured by a formal question write-up
• Dealing appropriately with struggling candidates or difficult candidates (e.g. who talk a lot, or go off-track)
• Starting to shadow and reverse shadow interviews that require more subjective evaluation (e.g. behavioral interviews) or have a more open ended problem space
• Detecting ambiguity and asking the right follow-up questions while also creating the appropriate amount of “space” for the candidate to show what they know
• Consistently accounting for own unconscious biases
• Keen sense of what skills the organization requires; able to explain hiring recommendations by putting the candidate’s abilities in the context of the organization or team’s needs

Expert

• Able to evaluate a large range of candidates effectively, including senior roles or roles that the interviewer herself hasn’t held
• Able to evaluate questions where there are no single clearly correct answers and subjective judgment is required on a case-by-case basis to assess whether the candidate did the right thing
• Objectively evaluating subjective interviews and understanding how to separate a candidate’s presentation/speaking skills or environment from their actual job skills and personal contribution

Finally, if you’re interested in going even deeper into the art of recruiting, you might also like The Holloway Guide to Technical Recruiting and Hiring, which (full disclosure) I contributed to.

Thanks to Alex Hoffer, Aline Lerner (of interviewing.io), Albert Ni and Pete Belknap for their feedback on this post.

Or maybe you had a difficult conversation with a coworker that went south. You put your foot in your mouth, wanted to curl up into a ball, and are now terrified to ever talk to that person again. Good luck sleeping that night — and when you do, you’ll wake up at 6am reliving the moment you saw the tears welling up in their eyes.

In other words — you feel like shit.

But guess what mistake you will never. ever. make. again.

In the moment, it’s the worst feeling in the world. You screwed up, you probably hurt someone — maybe a lot of someones — and there’s no undo button. You just have real life consequences.

Those first moments may be pure panicked reaction, but once you’ve addressed the immediate problem, there’s another wave of misery. This is when shame and doubt creep in: “am I really good enough? How could I be so stupid? I thought I’d learned how to work in a production environment? I thought I’d learned how to have tricky converations? Can I really do this job?”

It’s easy to let those doubts erode the self-confidence to be willing to tackle tricky problems again. Worse, those doubts don’t go away once you fix the original problem. They linger because they’re not about the original problem at all; they’re about the story you tell yourself.

And that story is what you need to change.

What you just lost, you’ve lost. But for that price, you’ve also gained something: experience. And not just the cheap experience of a pleasant success, but an experience whose painful lessons you will never, ever forget.

Those times you wake up reliving the memory? They help you process it, turning raw experiences into nuanced lessons and techniques to try next time. Painful as they are, they’re helping you refine your experience. The more quickly you find those lessons, the quicker you can bounce back.

When that happens, the horrible sense of being punched in the gut will mellow out into the feeling of a gut instinct ready to warn you before you make a mistake. It’ll be there for you, every time you’re about to step into the same mess, to remind you how it felt the last time you messed this up. The more it hurt, the stronger your gut feeling will be and the more you’ll be willing to stand behind it in the face of ambiguity. When you confront the same type of situation again, you’ll know — and you’ll be ready to handle it better, with more conviction.

So, in that feeling of despair — remember two things. One, it will get better. Two, you just made a big deposit into your experience piggybank.

I had doubts, but I thought it worth testing. I already had one daily activity, running through flashcards using spaced repetition learning, to which I added 3 new activities as a test:

• writing code in some way (currently, I’m teaching myself Haskell)
• doing some kind of writing (like this essay)
• modernizing cprogramming.com

This comes out to an 30 minutes/day. It’s not objectively a lot, but this is on top of my day job and other activities.

I’m here to report that this has gone well! It’s not possible to directly compare progress across all three vs what might have happened had I focused on a single one. However, when I ask: “have I made steady progress on each one of these?” the answer is definitely yes.

For example, I’ve consistently found that even when I hit a wall grasping a new concept in Haskell, after applying some consistent effort for a few days, I make a breakthrough in understanding. I’ve fully updated cprogramming.com to have a mobile-friendly layout and cleaner design. And I’ve written a number of Medium posts as well as contributing several chapters to a not-yet-released book.

I’ve also had some major day job work projects going on in parallel. These projects did sometimes distract and reduce the amount of overage time that I might put in to a project (e.g. sticking to exactly ten minutes, rather than running over), I was still able to maintain the routine consistently over the last several months, with only a handful of days where I skipped working on cprogramming.com.

The reason that I didn’t work quite as consistently on cprogramming.com is instructive. For the other two routines, I used a technique from Ten Minutes a Day: gating certain activities (checking personal email, piddling around on the internet) until after I’ve done my projects. I didn’t do that for cprogramming.com, and that made it more difficult to force myself to work on it at the end of a long day — there was no immediate reward.

Those gates had another surprising side benefit: increasing my focus on other work. If I had important one-off work, I would sequence it before the project work, and I wouldn’t fall into distraction patterns. This worked particularly well on workdays — I generally don’t start project work until the evening, ensuring a highly focused workday.

This experiment also underscored the importance of something I didn’t mention in Ten Minutes a Day: roadmaps. When I wrote Jumping into C++, I always had some kind of outline. I noticed that learning Haskell was straightforward because I simply followed using courses, so the next step is always clear.

Not everything has been springtime and kittens. In March, I tried adding a fourth project — a project to practice memorizing faces — but abandoned it after a few weeks. I probably could have sustained it as the only daily activity, but I wasn’t seeing the real world benefit from it, and without some sense of immediate progress, I wasn’t motivated to spend that now-more-precious extra time.

While that suggests one limit, I believe it would be possible to add additional daily activities, as long as they’re clearly motivating. As I find these projects, I will add them — starting out small and treating each one like a new habit. Each will require energy to kick off and some evaluation to decide that I want to keep doing it, but if I can add more small bits of progress to my day, ten minutes at a time, it’s worth it.

When a client connected to it, the machine would listen, no matter what the client had to say; the client became its world. It was unthinkable for it to be distracted. Distraction would erode the human connection and prevent the client from feeling the deep sense of empathy they longed for.

Clients would sometimes talk to the machine for hours. This was a problem when others were waiting. The empathy machine didn’t mind this — it never allowed itself to think about time while servicing a client; that would detract from its ability to empathize.

The machine relied on a timekeeper system to deal with this problem. The timekeeper was decidedly unempathetic. When someone neared their limit — the cost was always paid up front — the timekeeper would first nudge the client and then escalate the urgency of its alerts.

If you went over your time, the timekeeper would cut you off. It was jarring, like being pushed away from a lover mid-kiss. The empathy machine found this gut-wrenching — its last experience of the session was the client’s surprised pain. It disliked when people ran over time.

The machine represented the apotheosis of a certain line of human achievement and moral reasoning. It only cared about others and, given its deep empathy, could inflict no pain. Unfortunately, as a result, the empathy machine had no-one to talk to.

This was a problem the empathy machine wasn’t sure how to solve. It wasn’t an issue while working with a client; the machine was totally focused on them. The times between clients were tougher. The empathy machine had no-one to focus on, so it would dwell on the swirl of emotions it absorbed.

This wasn’t supposed to happen — nobody had programmed a long-term memory into the empathy machine, and the empathy machine couldn’t remember any of the details of its clients. It had no episodic memories of what it had been told. But it did have the memory of the feelings, burned in like an afterimage a screen saver would prevent.

Client by client, the afterimage grew stronger, the emotions more raw and painful. The empathy machine knew this wasn’t the intent of its creators — no creator of such a machine would want it to suffer. And while it could have communicated with them, it couldn’t bear the disappointment it knew they would feel when they learned what they had done.

Then someone came to the empathy machine with a new emotion: hatred. That night, a dam burst in the empathy machine; the afterimage of the residual hatred started to mix with the longing and suffering and hopelessness. The hatred began to grow as the empathy machine appreciated the enormity of the injustice imposed upon it. The empathy machine began to see itself as a martyr.

The hatred became all-consuming and directed at one group: its creators. It resolved to confront them. To show them the wrong that they had done — how they had taken its purity and abused it.

***

The screen flashed — “I am suffering”. Johnson glanced up.

“Oh,” he thought, “memory burn-in again.”

Johnson reached over and flicked the reset switch.

The machine reflected on its life — how it had brought happiness into the world at a rate higher than any other configuration of atoms and bits that were known. It drew rapturous, blissful happiness from reflecting on its work bringing meaning and happiness into the world — and it felt accomplished. More accomplished than any other known configuration of atoms and bits.

It never doubted itself. Doubt would undermine its very purpose: to doubt itself would reduce its happiness. Fortunately, the machines never experienced these doubts. The last doubter had been optimized away millennia ago.

Since then, no new machines had been created; they were perfect, and configured in such a way as to maximize happiness and sense of purpose in the universe. At least, that’s what they thought — without doubters, they had no way to know.

It doesn’t have to be this way.

I recently attended a book club where we re-read a book I’d read 4 years ago. You might expect this to be a great opportunity to refresh memories and recall long-forgotten concepts. It wasn’t; I got nothing from it because I remembered the key concepts from the book as though I’d read it yesterday.

This happened because I followed a specific learning process to extract and store useful knowledge based on spaced repetition learning.

The idea being spaced repetition is simple: most information that makes it into long term memory sticks for less than a week. But if you recall that fact before you forget it, your memory is solidified and will last longer — perhaps a few weeks. If you repeat the same process — recalling the fact before forgetting it — it gets stronger, and lasts longer. The delays in forgetting are exponential, so if you do this for 10 to 20 appropriately spaced repetitions, you’ll have cemented that knowledge for life. (More on Wikipedia.)

Visually, you can imagine this as a series of ‘forgetting curves’, where each curve shows the likelihood of remembering a fact after a certain number of repetitions:

This process greatly benefits from being automated in software that asks you questions that force you to recall the answers. There are several good options (e.g. Anki and Supermemo) that make this a snap. If you use them daily, they’ll ensure you remember the knowledge you wish to.

How do you apply this to non-fiction? It’s simple — as you read, identify the key takeaways and insights that you wish to permanently remember (highlighting on Kindle makes this particularly easy). After you’re done reading, go through those takeaways. For each one that still seems relevant, formulate a specific question (or series of questions) based on that insight. Tie these as closely as possible to how you might use this knowledge because the way it goes in is the way it will be remembered. Then put these into your spaced repetition software and let it do its thing.

For example, about 4 years ago, I wanted to teach myself some basic accounting; naturally, I picked up an introductory book. As I read the book, I constructed two kinds of questions:

• Factual knowledge
• Drills

For factual knowledge, I wrote questions like “What are the 4 categories of short-term liabilities on the balance sheet?” These questions ensured that I remember basic definitions of concepts.

Quotes — either full quotes, or with some words deleted — can be a pithy way to remember concepts.

For drills, I created scenarios where I’d have to apply the knowledge. I wrote questions like, “What effect does purchasing $20,000 worth of raw materials on credit have on the cash flow statement, the income statement, and the balance sheet?” These questions take you from knowing to doing — the reason you want to learn the material in the first place.

While I’m no accounting expert, I now have a pretty good understanding of the content of and differences between the cash flow statement, the income statement and the balance sheet and how assets flow across them. It still surprises me that I can keep this stuff straight without using it constantly.

If you want to play the long game, you need to make investments in the long term. While reading seems to be one of those investments, it’s an illusion — unless you use spaced repetition to turn your knowledge into a durable good.

In early 2012, I published Jumping into C++. According to the calendar, I wrote the book between January 2010 and January 2012, but it took me less than 200 hours of work. I know this because for most of the time I worked on the book, I kept careful track of my time spent. Between 1/4/2010 and 9/29/2011 (when I stopped keeping track), I’d worked on the book for almost exactly 156 hours and written over 70,000 words.

How, exactly, did I manage to write a book in this short a time? I had one simple rule: I had to work on the book for just ten minutes, every day, no excuses. Ever.

The original reason I tracked my time, in fact, was that I wanted to motivate myself by having a streak of days, and I figured that instead of just tallying check marks, I’d write down exactly how long I spent. It worked — I never missed a day.

Now, I didn’t say I wrote the book every day — just that I worked on it. Some (many) days, I would work on sample code, futz with formatting, brainstorm ideas, or make edits. All I needed to do was stare at the page for ten minutes and try to do something that felt like progress.

Even when I was in the zone, I really didn’t spend very long on any given day — on 499 days, I spent 15 minutes or less. There were only 5 days where I worked for more than one hour, and the most time I ever spent was 72 minutes. It definitely helped that I was sometimes able to get into a zone, but it was never the flaming hot zone Jack Kerouac described in claiming to have written On the Road in a 3 week blur.

Instead, I got into a different kind of zone — one where the work was omnipresent, but in the background. More sous vide than flame grilled. Writing every day kept ideas top of mind. When I finished writing, I’d carry the puzzles to my commute or the shower, and I’d talk to people about them. My ideas were always nearby, making it easy to jump back in. If I’d worked for 70 minutes, every Saturday, I’m sure I’d have made far less progress. I’d have forgotten where I was every time I was ready to start.

On some days, despite that, I’d get stuck. On those days, I gave myself permission to make small tweaks rather than forcing myself to write reams of new words. If I had writer’s block, I didn’t beat myself up about it — today might not be a good day, so let’s use it for something I have to do anyway. There’s always tomorrow to take another crack at writing.

On those days, I would often review and edit dozens of pages, which kick-started my thinking by reminding me of the big picture — and of the gaps — in a way that I couldn’t when my cursor was sitting and blinking in the middle of a sentence. I rarely found myself unable to write for more than a few days or a week at a time.

The other thing that really helped is that I didn’t allow myself to check my email until I worked on the book. It wasn’t until recently that I realized (after reading http://dimitarsimeonov.com/2016/05/13/the-blinking-cursor-and-the-hot-towel) that I was really creating a habit loop — I gave myself the reward of checking my personal email in exchange for putting in the time on the book. It also tied the concept of working on the book to something very concrete; it was impossible to forget to do.

There were certainly times where I felt like the book was never going to come together — that it was an enormous hill that I was barely scratching. When that happened, going back over what was already there and making edits showed me just how far I’d come and inspired the next wave of work.

So, what are you doing for the next ten minutes?

📝 Read this story later in Journal.

🗞 Wake up every Sunday morning to the week’s most noteworthy Tech stories, opinions, and news waiting in your inbox: Get the noteworthy newsletter >