I’m not sure why I keep trying to get into running. It’s never worked out for me. I was promised guilt-free eating, health benefits, a smug sense of moral superiority, and a runner’s high. I eat more but that’s about it.

Cardio fitness generally isn’t a limiting factor for my lifestyle. Other than the time my girlfriend’s elevator broke and I had to walk up 19 flights of stairs1.

I’m not a running kind of guy. My 23andme results back it up. I have the genetic potential of an elite powerlifter - not my fault I’m bad at running. Watching a treadmill count down is like water torture. The worst possible video game. Everything hurts. Running sucks. Why am I doing this again? So my past running attempts failed.

But something had to change when I realized I’d gained 15 pounds working at Rippling2. I decided to try again.

Slower

I was complaining about running to a hardcore marathoner friend, and he introduced me to the concept of Zone 2 training, which focuses on long, slow runs at 60-75% of max heart rate, to build endurance. And at the beginning, this means going slow. I mean really slow. I was getting smoked by grandmas with strollers down JFK Drive.

As it turns out, the body uses different energy sources and different muscle fibers at different intensity levels. And you have different amounts of these energy sources - 50-100x the calories in your body are stored as fat vs. carbohydrates, but they convert much more slowly, so carbs (glycogen) gets used for high intensity work.

Low intensity workouts stimulate Type I muscle fibers, which help you get more efficient at using fat, for longer, and save the limited glycogen for later. But interestingly you also get benefits at high intensity because Type I fibers help clear lactate. This reduces the feeling of “the burn”. More here if you’re interested.

This means the main way to get better at running is not by trying harder, but doing more miles. But it’s slow. You can’t ramp up your mileage too quickly. Your body’s not used to it. So you need to run slowly. Multiple times a week. And ramp up gradually. Your body eventually learns. As you plod along.

Hacks

At first this discovery was DEEPLY FRUSTRATING.

I prefer to find tricks and hacks to do things faster and more efficiently. I don’t think zone 2 training really counts as a hack. It’s like you asked someone how to get better at writing, and they told you to sit down and write for an hour every day. The preference for efficiency is what attracted me to the cult of HIIT, which promises fitness in a few hours per week. My usual workout involves a gay man yelling over club music for me to work harder, while doing something different every minute3. I don’t have to think, and I hate slowing down.

In high school I had the concept of grade-to-work ratio, which, for example, led me to not do any of my AP Bio homework. Midterm scores could reset our overall grades, and I was pretty sure I could cram it. My teacher, understandably concerned with my (pre-midterm) F in the class, called my parents, lectures ensued, I did homework, took the midterm, and got my A4.

This strategy was well adapted for school, but is pretty badly adapted for life. It’s fun to discover an arb or hack a system, but most of life's best things require steady persistence rather than cleverness. You can't shortcut your way to meaningful achievements. Being a good partner or friend, staying healthy, even developing a craft — these all require regular consistent effort.

There are marginal ways to improve at running that don’t require you to run, like the right shoes, nutrition, and sleep. But the main input is still consistent miles. No adrenaline, no ways to transmute money into results, just the grind.

Showing up and not hating it

Running slow is harder. Just when you're vibing, feeling good, outpacing that grandma we mentioned earlier, your watch beeps, telling you you’re working too hard. What the fuck? Working too hard? And so you slow down to a pace that’d feel uncomfortable in the Tenderloin. And it feels terrible. But this is steady plodding lets you add miles without crashing at mile 2.

And you just have to show up. I’ve tried a lot of things. Out and back naturey routes, candy, procrastinating on more important things, lying to myself that I just need to show up. Though I’m not at the cutting edge yet.

This doesn’t even seem to get better as you get better at running. The author Haruki Murakami has this amazing quote in his memoir, “What I Talk About When I Talk About Running”:

“Once, I interviewed the Olympic runner Toshihiko Seko… I asked him, “Does a runner at your level ever feel like you’d rather not run today, like you don’t want to run and would rather just sleep in?” He stared at me and then, in a voice that made it abundantly clear how stupid he thought the question was, replied, “Of course. All the time!”

The you’re-never-done quality of exercise used to bother me. I recently found a good framing: toothbrushing vs. diploma problems. Once you’re done with a diploma, you’re done. You don’t have to do anything after that, no matter what the fundraising people say. But you have to brush your teeth every day. So you might as well learn to enjoy brushing your teeth.

Changing the framing from “I must suffer. Existence is suffering. I am condemned to pushing my meat sack up Hayes Hill like Sisyphus because otherwise I’ll get fatter and also lose the $100 I spent on a half marathon” to “I have the privilege of developing my bodily strength and beauty to the highest limit” was very helpful. Both are true. But one of them makes it easier to put on my shoes.

I’m still very mediocre, and have a long way to go before my first marathon, but I’m happy that I’ve found a way for running to be fun instead of painful.

I’m not telling you to start running. It hasn’t made me a better person. I’ve never gotten the high, and the smugness is just to get likes on the internet I promise. But if you want to get into it for whatever reason5, and haven’t gotten it to stick, going slow to go further has worked for me.

After successfully demonstrating that I could swim in a square, clockwise and counterclockwise, my instructor drilled into me when to ignore what I learned. Instruments can be used wrong. Your count can be off. The current may be stronger than you thought. Blindly following your instruments is an easy way to end up in a bad place. A compass doesn’t lie, but it might not be telling you what you think it’s telling you.

Instead, he said, the number one thing, if conditions allow, is to understand the dive site, visually keep track of true things, and navigate relative to them. If you're on the wrong side of the reef, it doesn't matter whether you counted correctly. The beach is always up the slope, and you can navigate by keeping it to your left or to your right. Bubbles always travel toward the surface. Fish and turtles and boats can move, but buoys and reefs and shipwrecks don't.

If you know you’re swimming between the beach and the reef, with the reef on your right, and you can see both, you don’t really need the compass.

In the modern world, though, it's increasingly easy to navigate the world through instruments and proxies. But they have a loose relationship with reality. Apple Maps used to send people to the Australian outback, and Google Maps tells you to take illegal left turns. Police and schools juke the stats. The teen depression crisis might be due to an Obamacare measurement change. A PowerPoint deck confidently tells you what a feature does, and whether it's in production.

This isn't to say that instruments are bad. It would be impossible to function in the world without them. But I try to treat them as what they are - proxies - rather than ground truth. And only measure things that you’re using to make decisions.

I used to be really into sleep tracking. My watch, ring, and bed would give me rigorous quantitative evidence that I did not sleep well on Friday nights, supported by the anecdotal evidence of nausea, headache, and waking up on the floor1. By carefully logging ‘Drank alcohol’ into the apps, I could get a monthly email telling me how I could improve my sleep by drinking less. Measuring sleep in this case wasn’t wrong, but also was not very useful.

I would also sometimes wake up, feeling fine. I’d check the sleep tracker, and it would say that I didn’t get enough REM sleep or whatever, and I’d find it irresistible to conclude that I was going to have a bad day because I didn’t get enough REM sleep. So now I don’t measure sleep, drink less, and try to sleep around the same time every night, and it’s working out pretty well.

Treating proxies as proxies generally works well at the personal level, where you can look at a road sign and refuse to make an unprotected left across three lanes of traffic. You can try different foods and eat what makes you feel good, regardless of what your watch tells you. You can get your own oven thermometer.

Trusting eyes over instruments works less well at work, where you're often already dealing with abstractions of abstractions, everyone is performing for everyone else, and observing that the emperor has no clothes gets you executed.

One of my friends used to spend a lot of time on urgent fire drills from his CEO because one of the metrics was off forecast. He’d do a couple days of round the clock investigations before concluding that the DAU forecast was off because somebody hard coded Diwali last year. By itself, this isn’t necessarily the wrong thing to do. But the product that this was for hasn’t materially changed in the last 10 years, and the metric wasn’t being used to make decisions. My friend did get promoted.

It may not matter, for many values of 'mattering', that your metric is flawed and doesn't actually measure (e.g., likelihood a customer churns) - do you want to die on the hill campaigning to get it measured differently, or make number go up and get promoted? Usually it’s far outside your power to change it and your campaign will be viewed with suspicion.

I haven't got an answer for you. Maybe you can do both. Or neither.

But if you're looking for true inputs to make better decisions, I encourage you to look with your eyes rather than your instruments. Use the feature. Talk to a customer. Test your kid's algebra. Taste the soup. And steer clear of drivers who stare at their phones instead of the road.

Subscribe now

Every SaaS company at scale, and usually every product, eventually gets a slack channel where people shout a product question into the void because they don’t know who to ask directly. This happens when the company is big enough that people don’t actually know who all the other people are, or what their roles are.

The first channel that pops up is usually Sales/CSM/Support/Marketing asking questions for PDE (Product/Design/Engineering). Customers ask questions, and sales people don’t know answers. Sometimes the question’s already been answered in something formally documented. Sometimes the question was answered last week in the same channel. Sometimes nobody knows the answer!

It’s pretty expensive attention-wise for everyone who could possibly answer a question to read every single question, so usually a few cross functional people who haven’t burned out yet1 step up and play the role we’re going to call “traffic cop”.

The traffic cop and channel provide value in a few ways:

1. They maintain a bank of canned answers, both formal (FAQ doc) and informal (knowing the best way to answer a particular customer question), which saves everyone else from spending tokens on thinking up answers

2. They route questions to correct(er) people and reduce misdirected messages - turns out that question was actually for the Platform team, not the Apps team. This reduces both unnecessary distractions and also latency for getting answers

3. The traffic cop saves everyone from having to check the slack channel

The primary activity is taking and processing semi-structured inputs (“Is this a bug?” “What happens if the customer puts in a negative value?” “Do we support backdating?). They then have to figure out what the path to resolution is. Usually it’s one of five things:

1. This has a formally documented answer - read the support doc here!

2. This has a formally documented answer that’s stale - oops! maybe we’d update it more often if people read the docs…

3. This was informally documented in Slack two hundred threads (i.e., one month) ago

4. This is not documented, but I, the traffic cop, can answer it.

5. This is not documented. But I, the traffic cop, know who should have the answer, and will @-tag them.

Some of these are harder than they sound at first. How do you know a document is stale? How do you know if something was asked before in Slack?

Traffic Cop is never someone’s formal job responsibility, but projects grind to a halt without them. Usually the traffic cop has the job title of Product Manager, BizOps, Product Ops, Product Marketing Manager, or Early Employee because they are cross functional enough to know both the ask-ers and the ask-ees, and are also often actually writing the documentation.

I have never been thrilled with my performance in this role, but there are some trusted techniques for doing it more scalably:

1. Track inputs like tickets. Lightest weight but least reliable is using a slack react for each message, heavier but more robust is importing into Linear/Jira as a ticketing system and pushing responses back into Slack.

2. Don’t be a ball hog. ****It makes you feel important but trains your teammates to let you do their jobs for them. When you inevitably drop the ball, you will only get credit for dropping the ball.

3. Batch responses. Resist the urge to jump on every red bubble. You’re probably already fried from context switching, no need to do it more.

4. Share DMed questions in the channel, and process the responses - they should be integrated into enablement meetings, support docs, etc. Better yet - gently ask people who DM questions to ask them in the channel.

5. Self serve question answering (e.g., docs, FAQs, etc.) has a minimum activation energy. You need to get them to be “good enough” that people will look there first. In my experience it’s roughly covering ~50% of their questions. You also then need to maintain these docs for staleness.

Best of luck.

Author’s note: if the below role sounds like something you do, I’d love to talk to you. We can do some group therapy and swap notes. You can reply to this email or send an email to karl@lightpage.com

The basic theory here is that, most of the time, dependencies slow you down. There are ways to ac/decelerate the process that are problem-agnostic. Usually, you are communicating back and forth with other people until you have reached some form of mutually acceptable agreement, and there are ways to do that faster.

A lot of these tactics get 0.01% of the project done a few hours or days faster, but they really add up. Taking hours off of each sub-chunk of the core dependency of a giant project can make the whole thing go way faster.

Summary

Time and attention are precious. Fewer cycles means faster resolution. Faster cycles also means faster resolution.

1. Communicating in full chunks allows people to process your request instead of following up for clarification

2. Do the extra work to communicate clearly and with empathy

3. Spoon feed requests and to make them happen faster

4. Align to their schedule to reduce time between cycles

5. Accelerate tempo with interruptions

Minimize round trips with full chunks

This is the same principle as not sending a single Slack message saying "hi". Every time someone gets around to replying to you, you are magically at the top of their queue for that period of time

• Give enough information that the other person can resolve your request immediately. For example, don't say "I'm mostly free Tuesday afternoon", say "I'm free 1-4pm Tuesday".

• Give people detail on whether you've actually done the thing or not. Instead of "I can reach out to the vendor", you can say "Unless you tell me otherwise, I will reach out to the vendor about X by EOD and cc you so you know it's done".

• Basic stuff, like making sure someone has access to a doc when you share it with them, you attached everything, you named files clearly, etc.

• Try to have the correct people in a meeting or on an email chain so that you don't have to revisit. There is value in a group sitting together and seeing all the other people nod.

People are busy. Do the work to be clear.

Different companies have different communication cultures. But most of the time, you will be rewarded for communicating clearly, and making it easy for other people to understand (and do) what you are requesting.

• Be clear (at least with yourself) about what outcome you want in a given interaction. A lot of people just show up to meetings (and life, tbh) without intention, yet they are surprised their goals were not met.

• Imagine the other person is distracted, sleep deprived, and thinking about lunch and/or not getting fired. Chances are those things are true.

• Context matters - remember that while you might have been thinking about a particular thing for the last few hours, somebody else has probably been worrying about something else.

Spoon feed to prevent choking

The more you can spoon feed something (e.g., "please send an email containing the below paragraph to these 3 investors so they can help with introductions"), the more you will get what you want quickly.

It takes activation energy, time, and attention for someone to compose an introduction email, to go figure out how to run a workflow, or dig into a tool they use once a month and find an approve button.

Also, the longer it takes them to do the thing you're asking them to do, the more likely it is they'll get distracted by something like reading a really overly long sentence and/or get interrupted by a slack message or an email or a and forget what was going on at the beginning.

Keep the ball in their court.

People (including you) have rhythms to their work. If you work within or adapt your rhythms to whoever you're working with, you can make things go faster. The cost is usually disrupting your own rhythm.

• Reply to messages immediately as you see them. You can often get a couple rounds of communication in. If you wait, they might get into a meeting and you won't hear from them for hours.

• One way to get things moving on a daily cadence is to make sure the ball is always in someone else's court at the end of each day.

• Avoid getting buried. If you send an email on a Saturday, your email might get buried when they check their email on Monday, and they won't get to it until 4PM. If you schedule send it at 9AM, they might reply at 9:05AM.

• If you're dealing with time zones, work with their rhythm. Let's say you're working with a partner in India (IST), and you're in San Francisco (PT). If you send an email at midnight PT (12:30PM IST) as you're going to sleep, they might reply to you by 2AM PT (2:30PM IST) . If you reply at 9AM PT (9:30PM IST), their day is already over, and they won't get to you until the next morning. But if you get up briefly to reply at 3AM PT (3PM IST), you can often get in an extra round of back and forth.¹

Pick up the damn phone

You're probably a millennial or gen z who has some sort of vague phone anxiety. Get over it. If there's something you (likely both) care about, and someone has given you their phone number, you have the magical ability and usually have explicit permission to interrupt them. You are allowed to call them, they are allowed to not pick up, and you will get things done faster.

A few important caveats:

• This does break focus for the receiver, so use it judiciously. You can also send a slack saying "hey can you jump on a call, want to talk about X".

• When a call could've been an email, and especially when a call scheduled for next week could've been an email sent today, send the email.

• This is especially true when choosing between email today and call next week.

Keep driving things forward

• You're allowed to follow up on things. You're allowed to follow up weekly or daily or hourly depending on your own judgement.

• You can also backchannel things, play good-cop-bad-cop, ask your CEO to send angry emails to slow partners, etc.

• You can ask for (un)reasonable timelines, even if they are rejected. Sometimes you can demand them.

• While doing this, you should generally be polite but firm. I've seen other techniques work, but they're off-brand for me and so I rarely use them.

Warnings and conclusion

Most of these tactics are computationally expensive on your end. You are working 2x harder to go 4x faster. This is not "leverage".

1. Communicating in full chunks allows people to process your request instead of following up for clarification

2. Do the extra work to communicate clearly and with empathy

3. Spoon feed requests and to make them happen faster

4. Align to their schedule to reduce time between cycles

5. Accelerate tempo with interruptions

6. For things you want people to understand, repeat yourself, a lot.

Neal Stephenson, “Fall; or, Dodge in Hell.”

My sister, a doctor, has no idea why anyone pays me money for my time. I, however, have an exact idea of what she gets paid to do1. It's much harder to classify "oh I take random projects that executives don’t want to do”.

I’m annoyed that when people will ask "what exactly is it you do around here?" and I won't be able to give a clear answer. People will ask “where do you work?”, and they won’t know where it is2. Friends won’t understand. Their parents certainly won’t3. It probably doesn’t serve me well on dating apps.

This goes against the social training that pushes toward categorization. Most of our social technology is creates cogs in the machine that are standard shapes and sizes. A third year biglaw associate is a third year biglaw associate, and while lawyers probably have intense arguments about whether Skadden or S&C is currently cooler4, those fights are internal to the guild. Every bigtech engineer seems to be solely focused on making L(N+1).

Illegibility makes it hard for other people to know where you fall in the pecking order, which is terrifying for most people.

It’s funny when Parker Conrad’s linkedin job title is Customer Support at Rippling, or Patrick Collison is a Content Strategist5. But they’re already famous.

Another thing is it’s surprisingly easy for people in relatively illegible positions to seem more competent than they are. The distracting smooth talker and the person holding everything together are maybe distinguishable in a meeting, if you have enough context to identify the bullshit. It's rather harder to tell on Slack, let alone in an interview. People can get away with solely adding half-thought comments to Google docs for a very long time, as long as they occasionally do it around 9pm.

The high context, difficult to classify work is illegible by nature. There is variance in illegibility. Measuring support tickets per week closed, or sales quota attainment, is easy. Making a chaotic project go less chaotically than it would’ve otherwise is a lot harder to objectively measure, though people at the ground level usually know who was helpful and who was not.

The people who work on special projects6 are usually unsung, sometimes heroes. They work hard and few have heard of them. There is a clannish quality. People who have done it immediately click, and people who have not have their eyes glaze over.

People describe special projects as wearing multiple hats7 or a SWAT team, but in my experience it's more like wearing a red shirt at an understaffed Target. The cash registers are down, Glenda has the flu, and someone shit in the dressing room.

Part of my journey has been learning to embrace and accept the illegibility. If most of the company doesn't know what I do, that's fine. But every time someone external asks me what I do, I debate whether to give a Starbucks job8 to avoid the conversation9.

It’s a job unlikely to get a tick-tock promotion every year to put on linkedin. There are few opportunities to empire build. But if I can demonstrate to the people who matter that I can get difficult things done, that's it. If my teammates want to work together again, that's it. If I'm proud of what I've accomplished, that's it.

There are few real rules1 in the game of life. And there are no admins.

People get this confused in part because we spend 12+ years trained by an education system designed to produce assembly line workers. We are led to believe there’s only one answer, and someone else figured it out. We were told as children that we could be whatever we wanted to be. Then in the next breath we had to ask for the teacher’s permission to urinate.

At my high school, there was fierce competition in grades. Dozens of us took 15 AP classes, and of course got straight As. One kid (or his parents) realized that any grade in non-AP classes (out of 4.0 vs. 5.0) actually decreases your GPA, and asked admin to not count his honors credits. He got valedictorian and was airlifted into Harvard.

In retrospect, this was really good optimizing slightly outside the box. But I was mad when I found out. Since most of my friends were second generation, going to a public school, we didn’t know about higher-class techniques like bribing lacrosse coaches, pretending to be LGBT, hiring essay writers, or claiming a disability to get more time on tests2.

Everything in this category, however creative, is still playing someone else’s game, plus or minus some willingness to cheat.

The hardest part of adulthood, for me, has been defining the game, who I’m playing with, what I want, what I’m willing to do. I can’t change the basic starting conditions, or the past, though I can change my interpretations of both. I see my peers still trying to please an admissions committee, and worry I still need to too. Fuck ‘em.

Many people conflate the inability to choose our starting conditions or outcome with an inability to choose our action. This is wrong. But it’s much easier to be a victim and blame factors outside of our control. It’s frustrating that the game isn’t fair. You can to walk away. You can petition other players to change the rules. You can even flip the table3. But sitting around complaining about why you aren’t winning as much as you deserve is unlikely to change your situation.

A consequence of defining your own rules is realizing how arbitrary the rules you think you’ve been handed are, and how other people aren’t even reading from the same book. Spend 30 minutes with a lawyer, and you’ll discover even the letter of the law is much fuzzier than most people think. Then there’s enforcement, which can be charitably described as inconsistent.

There are also many softer ways that other humans introduce rules. Decisions made and presented as fait accompli, or non-negotiable. Systems in place that take significantly less effort to be obeyed than to be disobeyed or circumvented. False dichotomies. Our desires are constantly manipulated through advertising, social validation / shaming, and the media.

But, to my knowledge, there are no stone tablets from God saying Thou Shalt Have A Bedframe By Thy Twenty-Third Birthday4.

Deriving everything from first principles quickly becomes unmanageable, not to mention there are higher costs of social coordination, so it’s probably not worth your time to define a new pricing structure with the power company or get people to wear kilts.

“There are no rules, only guidelines” is not a call to abandon your family, move to another country and live on a diet of cocaine. You may find that what you’re already doing is what you want to be doing given your goals, and the constraints that you’re under.

The upsetting implication, and why I sometimes avoid recognizing my own agency, is that if I have a choice, then what I’m doing is actually what I chose to be doing. This can be kind of upsetting for one’s self-image. Today, I’m choosing to continue being a middle manager at a software company instead of trying to become a pop star, artist, astronaut, comedian, or romanticized small business5 owner. Later, I’m going to choose to watch some mindless TV. All of that is okay.

This might be obvious to some people. But for me, accepting that what I’m doing is my choice forces me to reckon with my revealed preferences. My job isn’t preventing me from being a pop star. My utter lack of interest in the work of becoming a pop star is preventing me from being a pop star.

On an individual level, things you’re allowed to do is one of my favorite lists because it’s full of actions on an appropriate scale for an empowered human.

But the scope of what a small number (even 1) of dedicated people can do is quite large. Most people, confronted in childhood with the lie of “you can literally do anything” and the truth of “if you don’t ask for permission to pee, adults will attempt to punish you”, settle somewhere around the truth they experienced and never unlearn it.

There’s the famous story of how they train elephants in the circus. The general idea is that the baby elephants are restrained when they’re small, tied to a pole and beaten when they try to escape. When the elephants are adults, they can still be tied to the same pole, though adult elephants can easily break both the rope and the pole.

You can’t do everything, but you can do a lot more than what you’re doing today. I’ve only really started wandering off over the last few years, and it’s been wonderfully rewarding. I don’t always feel the rope.

If you’re reading this, you probably overthink more than you underthink. In writing this, I’m probably overthinking. Overthinking about underthinking is an appropriate amount of headassery for me…

Behaviors serve some sort of purpose, they just might not be the purpose we tell ourselves. For me, overthinking serves a few purposes:

1. Procrastination. Overthinking is just like endlessly reorganizing your desk. For me, there’s twoish drivers of procrastination:

   1. I don’t actually want to do the thing but am unwilling to admit it.

   2. Preventing emotional pain. If I actually do the thing, I might fail. If I strategize enough, I can simultaneously increase the chances of success (narrator: not true) while also delaying a true evaluation of whether I’ll succeed.

2. Cleverness and/or perfectionism. If I just do the obvious thing1 what’s the point of deploying the Super Elite Special SWAT Team member? Wouldn’t it be great if I could anticipate every objection and have a slide that mitigates each2? If I get this perfect, nobody can criticize me. I’ve also worked for enough ex-consultants who sold packaged pearls of wisdom for a living, and have absorbed some of their neuroses.

3. Labeling the unknown is comforting. It usually doesn’t actually change the outcome, since most of the time what happens is fairly obvious. This might be a sub-form of procrastination.

There’s hundreds of productivity and note-taking apps out there that encourage overthinking. The personal-SaaS-industrial-complex would like you to think that for $9.99 a month this can be solved. The truth is most tools can only help you do what you already wanted to do, marginally faster or better.

I’m not going to build or sell something to encourage underthinking. But if there were something for underthinking, what would it look like? What does it look like in the real world? I’m deliberately spending 10 min thinking about this, and that’s it.

• The writing industry actually has some interesting takes on this problem. There’s the writing app that deletes everything if you stop typing for 5 seconds. iA Writer has “focus mode” that only shows what you’re working on.

• In recruiting, exploding offers timebox overthinking (usually not to your favor).

• Twitter (sometimes) prevents overthinking with character limits, though they’ve steadily eroded that mechanism with new features.

• Stories (Snap, IG, etc. etc.) lower the stakes, which helps lower the activation energy required to make a post.

• TikTok’s duets, remixes, lenses, songs, etc. all provide rails that make it easier to create content.

What’s are common threads?

1. Artificial deadlines

2. Templates and/or rails

3. Lower psychological stakes

4. Remove additional context

I’m going to try some of these on the meat computer and we’ll see how it goes.

Thanks to Matt MacInnis for coaching me to think less ;)

If you say something similar on the Internet, you risk a tidal wave of complete strangers throwing FEEDBACK at you for days (and intermittently for years!)3.

There’s a few reasons why this can range from disorienting-to-horrible:

1. Engagement lacks nuance; they don’t know you, you don’t know them, and nobody is giving each other the benefit of the doubt.

2. Sheer volume. A few friends agreeing that you’re wrong is not threatening. Dozens and hundreds of strangers telling you you’re wrong overloads brain processes designed to handle a universe of 150.

3. Friendly fire. These strangers usually aren’t actually yelling at you. They feel unheard and are yelling at someone way behind you. You just happened to catch the bullet because you’re nearby.

Spend enough time in the war zone and most reasonable people will exit. The survivors develop an immunity. Numb from criticism for mundane bagel opinions, it’s easy to conclude that all feedback is equal, and become unmoored, pushing away even the well-intentioned.

You might even harden your opinions. You’ve been voted and catapulted off the island, never to return to the fold. Might as well start your own.

In the worst case, you might even get addicted to the fight. Gladiatorial combat is a rush, and on the internet, you can always find an audience to back you up.

People who are immune to criticism and are addicted to combat start to behave like narcissistic warlords and/or cult leaders, which is part of why the Discourse is the way it is today.

There’s a couple ways you can make the internet a kinder place4:

• Avoid carpet-bombing. Tweeting “all men are trash” or “women only want your money” is not going to get bad men and women to improve. It will get sensitive people to numb themselves, or conclude they might as well engage in whatever crime they’ve been convicted of.

• Don’t feed the trolls. There are many people who have decided to pollute the commons for their own benefit. It’s bad on purpose to make you click / QT. Just keep scrolling.

• Avoid piling on for your own satisfaction.

• Try to write kindly. Remember that you’re often not talking to a friend who knows and trusts you, but writing to a stranger whose mailbox you just barged into.

• Accept you’re not responsible for everything, including educating random strangers about bagels.

• If you’ve gotten this far, realize you’re probably in the upper quartile of well-behaving online people already.5

Related reading:

• SSC’s Toxoplasma of Rage

• Fan-in, or “attention lensing” by Coda Hale. Thanks to Justin Duke for the tip.

Source: XKCD #386

I gained a lot of appreciation for people who make things, and lost a lot of tolerance for people who only pontificate.1 I found myself especially frustrated with my past self, whose default was to complain and/or comment, then wonder why things didn’t magically get better.2

It is, of course, much easier to complain about how things are bad rather than do anything about it, which is why people prefer to complain. 1/100th the satisfaction, but 1/1000000000000th3 the effort. Plus, when someone eventually fixes the problem you can pat yourself on the back for having brought attention to it. You can even complain about multiple things in the time it would’ve taken to fix one thing.

If it’s that bad and it would be so easy for people to just fix it, why don’t you?

It’s also easy to confuse being “helpful” with being helpful. A lot of people think they’re “adding value” by nitpicking, or supplying unsolicited takes, when they’re actually just draining energy and momentum. Even the legendary VC phrase “let me know how I can be helpful” is a meme precisely because it’s generally not helpful, vs. “I will connect you to five potential customers tomorrow”, which is helpful.

Making is so much harder.

One of the things they teach you in lifeguard / surf camp is that people who are drowning will attempt to climb on top of their rescuers, killing both. Simply try to do something about a problem, and many people will think you are responsible for the problem’s existence.4

At internet scale, simply by doing anything, you expose attackable surface area to furor you never imagined.5 Most people are shocked and drop out of making things, preferring to stay safe. Those who make it through often look like startup founders, battle-scarred and tone-deaf from being attacked for so many years.

Even if you learn to appropriately weigh complaints, it’s entirely possible to get DDOSed by attempts to help.

“Have you ever thought of X?”

“Cool! Yes, but there’s 12 other things we thought should come first. You’re welcome to go and do it. We’ve got a big tent and a lot of shit to clean underneath.”

“Actually, can’t help, gotta take my dog to therapy, bye!”6

This type of casual drive-by advice is astonishingly common. I’ve mostly learned to ignore it when received. I do still love to provide it. It’s well-meaning, and you get to feel smart without actually needing to do any work.7

I’ve come to believe that working through something is the only way to explore the idea maze. Everything else is commentary. I’ve mostly stopped sharing unsolicited “helpful” just-a-thoughts and comments at work. I save them for Twitter, the primordial soup of commentary from which living things8 occasionally emerge.

Now, when I get the urge to comment, I try to be on the maker’s side and ask,

“So, what are we going to do about it?”

twitter

1. above ~2k followers you can have a short discussion about pretty much anything

2. you can DM or reply to anyone, but be ok with being ignored

3. letting your audience train your content negatively affects quality

4. I generally learn more from accounts with fewer followers

5. it took 28 years for someone to unironically call me “Satan” on the internet

6. sometimes people are looking for a fight and you just happened to be around, don’t take it personally

7. real face + alt > anon

8. tech twitter has more ads than youtube

9. mute notifications from people who you don’t follow or who don’t follow you

10. many things that feel obvious to you are not to most people; conversely many things that feel insightful are not to most people

11. most of what you read on the internet is by insane people

12. calling out grifts is usually neither effective nor helpful

13. people beef and troll for clout

productivity/work

1. focus heavily on what’s actionable, everything else is just noise

2. write down 3-5 things you want to get done in a day, the night before. if you get it done then that’s a successful day from @pmarca

3. reducing work in progress dramatically increases speed

4. usually you can be faster just choosing to be faster

5. abandon more stuff instead of leaving it half finished and blocking

6. a crude but effective heuristic is just to say reflexively say no the first time anyone ever asks for something; wait for them to ask again

7. changing other people’s incentives to increase alignment is far harder than just getting them to like you

8. most of the time the obvious thing works, and it’s usually better to just try it

9. usually better to give people what the evidence says they want vs what they say they want

10. the real process != what management says it is != what people think it is

11. sometimes people who seem smarter than you just have access to higher quality information. also 99% of the time it doesn’t matter who’s smarter

relationships

1. would you rather be right or happy?

2. mirrored rituals and gifts are powerful

3. commitment is powerful and often a feature not a bug

4. thank, appreciate, and tell people you love them more than you think you need to

5. when texting, use more emojis than you think you need to

6. gestures inherently have meaning

7. semi-structured questions are useful when meeting people - including “how should I engage with you? What are you interested in getting DMs about?” h/t @krzhang

8. you can just randomly text or call people

9. the best friends are the ones who understand when you’re signaling play

10. the hardest part of maintaining relationships is somehow scheduling

11. internet-enabled weak ties are an ok substitute for unplanned irl interactions

self

1. I am a lot less charismatic over audio/video than irl

2. every year I think I’m getting more intentional, and it’s true, but this will also be true for next year

3. easy ~30% mood improvement from a hot drink, shower, walk, or sitting outside

4. cynicism is rooted in feeling like you “know better” and protecting the ego, and is self-fulfilling h/t @wgalyean

5. when I optimize solely for being clever and/or right, I end up being an asshole

6. I am not so smart that I can’t get hacked by crazy conspiracy stuff

7. I’m fine for 2 months with a laptop and pb&j sandwiches

8. trying to blend in with everyone else since puberty was successful, but the wrong thing to do for my happiness

9. physical pain (e.g., exercise) reduces psychological pain. exception: most of my physical pain is actually stress

10. I can shape my physical and digital environments to fit my needs (e.g., set up a desk outside, mute ig people who give fomo, mute words on twitter)

11. I’ve only really gotten ok at meta-programming my desires in the last two years

12. simply choosing to enjoy something or be grateful actually works

,etc.

1. nearly every culture thinks food and family are uniquely important to them. if you agree and ask nicely, you get to try their food and meet their families

2. instead of ___ you can always just… not

3. a lot of tests are actually just storytelling exercises

4. america has the civilizational equivalents of tapeworms, diabetes and gout

5. america is still one of the best places in the world to live because of costco

6. ads are involuntary spaced repetition reprogramming

7. some trad stuff is lindy, just not necessarily the 1950s american trad stuff

8. whether the world is meritocratic or fair depends on your abstraction level. at the highest abstraction level the universe doesn’t care. at the lowest level it’s up to you

   

If you work in tech, you may have noticed a confusing process where, every few years, there’s a new type of person whose title/JD has some permutation of “product”, “associate”, “general”, “strategy”, “business”, “operations”, “finance”, “manager”, and the symbol “&”. 

Their role is to “lead key business priorities”, “unlock new growth”, “drive alignment on strategy”, “help the organization scale”, “launch new markets”, and possibly “make sure the smoothies arrive on time”. They describe themselves as “special forces” and “general-purpose athletes”, though they’ve likely never touched a gun and generally aren’t very athletic.

What gives? Why does the title keep changing? Because of the Status Cycle.

The Status Cycle

The cycle creates a new cross-functional role, which has a status/comp arbitrage. As the new role gets increasing visibility, adoption and status, the arbitrage goes away and another New Title must be created.

Phase 1: New Title

Pioneering executives have trouble recruiting talented young people, often because their comp bands aren’t competitive. Unwilling or unable to pay more, they promise things like “high visibility”, “high autonomy”, and “opportunity for growth”, along with a New Title. This is mostly done by rebranding old roles, or by porting them from other industries (see: CoS and politics).

Ambitious youngsters see that it’ll take 2+ years to get promoted because their current manager isn’t making Director any time soon, and therefore they won’t make Senior Manager. Or they don’t want to be management consultants anymore. They decide to start climbing a new ladder.

Phase 2: Manifesto

The New Title hype has built enough that a Manifesto arrives, which increases publicity, visibility, media coverage, and solidifies the definition of the role. At this point, there’s usually a small community, and there’s a pipeline developing from the SF/Silicon Valley offices of MBB and elite investment banks. Most hot companies have 1-2 people with New Title.

Examples:

2015: Why Bizops is the Hottest Team In Tech

2019: The Chief of Staff Role in Silicon Valley,  Hail to the Chief of Staff

Phase 3: Harvest

Competition intensifies as New Title attracts more and more interest. Momentum from the high status of the earlier phases begins to be overwhelmed by declining exclusivity. Hiring managers and recruiters have caught on and start repainting old job descriptions.

At this point, there may be a big enough community for a conference. The first few cohorts have usually aged out into management roles. There are entire teams of people with New Title. Fresh MBAs usually show up around this stage.

Phase 4: Decline to Steady State

We are currently entering this phase for Chief of Staff.

Interest in the role slowly declines. The role is no longer unique, and uncool companies have started to use the title. Status begins to trend toward Project Manager, which is a large part of what all these roles do. 

At this point, it becomes necessary to add qualifiers, because there is again a need to define hierarchy between roles. For example:

• Chief of Staff to the CEO (reports at exec level)

• Chief of Staff, VP Operations (sparkling bizops)

• Marketing Chief of Staff - Western Division (?????)

Note: there are amazing project managers out there who do a lot of the same work, for a lot less money, because they went to a non-target university when they were 18.

The Aggregate Cycle

I suspect this process applies for jobs like Analytics -> Data Science, but don’t have enough personal experience to be sure.

As we play through multiple cycles, what results is a mishmash of different titles that do variations on the same job. Even if CompanyA and CompanyB generally pay the same, you might be better off being an L5 Chief of Staff, VP Operations at CompanyB than an L5 BizOps Manager at CompanyA. You also don’t know which team actually makes decisions, has exec sponsorship, or gets the interesting projects.

Overall, though, the process is probably net good - some illegibility and new opportunities create room for people to grow outside of existing hierarchies.

No need to worry about it too much. No matter how successful you are, your mom will still wish you went to med school.

Related reading:

Status as a Service by Eugene Wei

Geeks, MOPs, and sociopaths in subculture evolution by David Chapman

The fragmentation of the internet is continuing. Top of mind for most Americans is the possible nationalization of TikTok. The Trump administration is also saber-rattling with a “Clean Network". This behavior is dangerous for the future of American companies. We are losing the moral high ground on open markets, and fewer markets will be open to American companies. Fragmentation will accelerate.

There was already some natural fragmentation

Natural fragmentation. In some internet-enabled businesses, locals have a strong home-field advantage. They tend to not need any protection. There are millions of small complexities not translatable across borders. Or there is a need for a strong ground game. Examples include ridesharing/food delivery, accounting/tax, logistics, non-card payments, e-commerce, SME services.

Natural oligopoly/monopoly. Other businesses are much more global. They don't need much local customization. This means providers can scale, bottoms-up, even without an office or local entity. Returns on scale mean winners take most. Examples include social media, communications, search, most SaaS, card-based payments, and digital infrastructure.

The current global providers of naturally oligopolic (love that I can do this in English) services are generally American. AWS (but not Amazon.com), and MS Azure will continue to dominate. WhatsApp is everywhere outside of the US and East Asia. Instagram is everywhere except China. Google search and Android will be near-impossible to unseat. However, America has burned through the goodwill generated by programs like Internet.org and Next Billion Users.

Regulatory fragmentation is coming for the natural oligopolists

"How Asia Works" suggests a few key differences between successful and unsuccessful modernization. Successful countries protected and nurtured domestic manufacturing, but with mandated export discipline. Locals were protected at home, and got a lot of help, but had to compete globally. Unsuccessful countries opened access to domestic markets, but did not up-skill. Their economies are still dominated by multinationals. Oil, mining, auto, electronics.

China and Russia applied this framework to the internet. They've built protected local internets. India is starting, too. Governments will try to reduce the influence of Silicon Valley/Washington, and political stability / economic sovereignty will take priority.

Facebook has seen the fragmentation accelerating, and is scrambling to find local allies. It recently invested in Reliance Jio (India) and Gojek (Indonesia), both national champions. These also happen to contain 22% of the world's population. Reliance Jio and Gojek will then encourage local lawmakers to remain open to FB.

What’s next?

Real answer is ¯\_(ツ)_/¯, but that’s never stopped thinkbois like me from speculating. Generalizing from the FB example:

• Politicians, regulators, state media (especially large, nationalistic, or authoritarian ones) intensify targeting American companies for: 

  • Taxation

  • Nationalization / divestments

  • Shakedowns

  • Jobs programs

  • Censorship

  • Scoring points

• Favoritism toward local champions become more overt

• “Regulatory consulting”, campaign donations, direct bribery grow

• Multinational (esp. American) corp dev teams get busier yet face more uncertainty

• Development dollars used as currency for US/Chinese alignment

Who benefits?

• Local competitors to natural oligopolies in large markets; in the short term. If markets open up again and they can’t get legal barriers, they’ll get stomped

• Politically-connected entrepreneurs, “regulatory consultants” and lawyers

• Consumers in small, still-open, or previously-ignored markets

• VCs who can’t get into X, but can get into X for Region

• State censors

Who loses?

• Local competitors in small, still-open markets, who get stomped in the short term

• Consumers in restricted markets, who get less choice

• Companies with X for Everywhere baked into valuation

• American multinationals on average. $FB? $MTCH? $NFLX? Someone else can tell you how much international growth is baked into valuations.

Some other examples that helped form this thesis:

List is definitely incomplete. I've ignored Africa and most of Central/South America, which I know little about. I would love to learn more.

Natural fragmentation:

• Ridesharing - Uber burned billions then sold local ops to Didi, Grab, Yandex

• Ecommerce - Amazon.com losing in Asia (seems ok in Europe)

• Food delivery wars are not settled, but mostly local players

• Stripe has given up on India

Regulatory fragmentation:

• TikTok nationalization in the US

• Huawei’s troubles

• GDPR

• Reliance Jio's fundraising bonanza. It included Facebook, Google, a bunch of American private equity, the UAE and Saudi sovereign wealth funds

• India bans on Chinese apps, bounty for a local Zoom clone

• Lark, Bytedance's Slack competitor

• Whatsapp Pay blocked in India and Brazil

• Indonesia’s “foreign digital goods” tax

• Facebook censoring in Singapore

Thanks to Turner Novak for reading a draft of this essay.

If you want to talk about this more hit me up @chiefofstuffs on Twitter

As they grow, companies naturally develop immune systems as a part of their risk management and risk reduction practices. Most of your opposition isn’t malicious… it’s just emergent behavior from a system designed to restrict the rank and file to legible, approved sets of activities and thinking. There’s a reason why “internal startups” fail.

Most of this comes from my experience at a public “tech” company working on projects with the c-suite, product, strategy, corp dev, engineering, compliance, infosec, finance, fraud, a call center, and two different risk teams. 

— if this is feeling too much like an online recipe you can skip this section —

Congratulations! You’ve been selected as part of a secret internal committee whose goal is simple: prevent anything from changing at The Company. We like things just the way they were when we joined. All the music on the radio today is bad.

To be the most effective, you have to remain undetected. Block too many projects in a very public way, and people will know to go around you, or avoid you. You need to preserve your role in the hierarchy; a silent killer.

Despite your limited range of action, there are still many weapons at your disposal. Be warned: these techniques are like dimensional strikes. They eventually salt the earth so that nothing can grow here. But that’s ok. I’ve already vested and cashed out.

Framework

This piece will focus on a specific subcategory of project-killing: Strangling In The Crib. At this stage, when there are few resources committed, it’s more effective to sow Fear, Uncertainty and Doubt (“FUD”) to slow projects down and kill momentum. Eventually a few things will happen:

• The leads will give up, submit, move on, or leave

• The organization will develop a natural immunity. “We’ve been talking about this for forever. If it was a good idea, wouldn’t we have already done it?”

Advanced readers looking for techniques like Accelerate Into The Wall or Going In Another Direction (for projects that already have momentum) will need to wait for Volume 2, should it ever be published.

The Analytic Paralyzer:

This technique is most effective on absentee and/or anxious executive teams. This can prevent projects from being approved. As the name suggests, this merely involves driving the project deeper and deeper into the weeds until it dies of old age or its sponsors leave. There are endless dimensions that you can paralyze around, but here are a few of my favorites:

“How does this affect [insert other project]?”

• Reduce your enemy’s agility with endless Gantt charts, increased complexity and sync meetings.

“Is Project X on strategy?”

• If there are >2 former management consultants in the room, this is an insta-kill

• “What is our strategy, anyways?”

• “We shouldn’t commit to any projects like this until we figure out what our strategy is, otherwise Project X might be off strategy. We can’t afford any more whiplash, morale is low enough as is.”

“What are our competitors doing?”

• There’s usually just enough public information to feel like more could be known with some more digging.

• But there is usually very little definitive evidence. Plus, execs always care about the competition.

• Played correctly, this can suck 1-2 months from a project with zero change to any decisions made

Follow the Instructions

“Normally, if given the choice between doing something and nothing, I’d choose to do nothing. But I will do something if it helps someone else do nothing. I’d work all night, if it meant nothing got done.” - Ron Swanson, Parks and Rec

This is best used to destroy the lead’s will to live, while maintaining plausible deniability. It’s a form of malicious compliance indistinguishable from zealously “protecting the company”, which is why it’s quite useful. This is a known technique for strikers.

If you work in compliance, legal, or infosec, you probably don’t need me to tell you how to destroy someone’s will to live. If you work in a different department, but still want to destroy someone’s desire to live, all you need to do is draw a pentagram on the whiteboard, speak the invocation. “We need compliance’s approval to move foward on this.”

Pro tips:

• Your naïve victims may assume there’s a rational basis for the rules you’re rigidly following. Subtly indulge that assumption so that they try a bit before you shut them down, ideally in front of leadership. Try not to show the schadenfreude.

• To be most effectively ineffective, soft block projects by highlighting risks without providing any help mitigating them. It’s the business’s decision, of course! But it’s not your job to help make it…

Author’s note: despite the stereotypes about these extremely important functions, I’ve had the pleasure of working with some wonderful people who do their jobs conscientiously with a view toward achieving our shared goals. I’ve found that people in these functions are generally not set up to succeed, and/or are given the wrong KPIs, and they’re just acting according to their incentives.

Compliance/Legal:

“We’re at [size stage larger than in the past] now, we need to start requiring [DNA samples from everyone who works at a vendor].”

“The law isn’t really clear so we’re going to need to wait for [case in state with <1% of your customers] to resolve itself before we can give a recommendation”

"Based on [Law], we could be liable for up to $2 trillion in damages. You, the CEO, and the CEO’s family two generations above and below could be flogged and executed.”

InfoSec:

There’s basically two flavors of infosec people. The first flavor is the ex/white hat hackers. They’re paranoid and have an anime Slack avatar “as social engineering defense” but will solve problems. The second flavor is actually a bad Compliance person, used to work at Ernst and Young, and cargo cults what the first flavor says without understanding the principles behind it. Your goal as a Change Preventer is to be like the second flavor.

“What if we get hacked?”

• If unchecked, this question can create an endless list of requirements that effectively kills a project from its sheer weight.

• Bonus points for mentioning PII, HIPPA, and Social Security numbers even though companies that leak this data largely go unpunished

“Let’s go with the [vastly inferior] solution from Dell / EMC / Oracle, they have all the certificates.”

• This is a signal that you don’t actually care about security and actually care about a proxy for ass coverage.

• Congratulations, you’ve successfully continued the tradition of “[I won’t] get fired for buying IBM”

“What if quantum computing breaks the current encryption paradigm, and all our security becomes useless?” (This is how my manager would joke about this category of question)

The Goalpost Switch / Scope Creeper

Why fix what ain’t broken? Things are going so well / the wheels are barely staying on. Sure, we love innovation here at BigCo. Just fill out these 7 forms and you can use Bob’s old airgapped ThinkPad to try this project out.

“What if [group of customers] commits [bad action like fraud]?”

• This works because of the insidious human tendency to attribute any negative impacts to the change, rather than the baseline.

• The question works well if 1) [bad action] is a scary word and 2) the answer is unknowable.

• Pro gamer move here:

  • Imagine [fraud] is currently 1%, and [fraud] with Project X is correctly estimated to be 0.6-1%.

  • Focus the discussion on that 0.6-1% and how it could be reduced, and ignore that there’s an expectation that [fraud] actually decreases

“Why can’t this also solve [tangentially related problem]?”

• Ideally you rope in an under-resourced, slow-moving team that can complicate requirements.

• This synergizes very strongly with the Analytic Paralyzer

“Why isn’t everything automated? This won’t scale.”

• This is, of course, just a different flavor of scope creep - we can’t do this unless it’s perfect.

“We need to limit the size of the pilot" AKA “No oxygen for you!”

• As an innocent young analyst, I once led a machine learning project where compliance and infosec limited the sample data to 50 documents.

  • Amusingly the startup I was working with (at the time, 3 people) found product-market fit through this project, and is now valued at 2x my former employer.

• You can also try to limit the project through unreasonable restrictions on risk.

  • Kill the A/B test the moment conversion changes, but before they can figure out why

  • Restrict the rules engine to be as strict as possible (e.g., levenshtein distance of zero for name matching) Remember. What if [group] commits [fraud]?

The Classic Sabotage

Back in World War II, the OSS (predecessor to the CIA) created a document called the Simple Sabotage Field Manual for ordinary citizens in occupied territory. The first half focuses on physical sabotage, like putting sand in the gears at a factory. The second, in my opinion, is more interesting.

“A second type of simple sabotage requires no destructive tools whatsoever… It is based on universal opportunities to make faulty decisions, to adopt a non-cooperative attitude, and to induce others to follow suit.

Simple Sabotage Field Manual

Further reading:

Be Slightly Evil by Venkatesh Rao

If you liked this, please share. If you’re looking for more I also tweet at @chiefofstuffs

Share

This content is provided for informational purposes only, and should not be relied upon to make legal, business, investment, tax, strategic, or career decisions. You should consult your own advisers as to those matters. While adapted from sources believed to be reliable, Chief of Stuff has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. 

I worked at a struggling tech company for a couple years, and spent a lot of hours with my friends (also in similar situations and colleagues examining what we would’ve told our past selves. Below is a quick distillation of wisdom if:

Your company is struggling

1. If you get a retention offer, take it and start looking immediately. DO NOT act like you’ve already earned the money. Those 6-12 months will destroy your soul. At one point, I was calculating how much extra I was making, per remaining hour, to keep myself in meetings. “They’re paying you $200 to be here, just shut the fuck up and nod…”

2. The company is not your family. Some of the people in the company are your friends in the current context. It’s like your dorm in college. Hopefully some of them will still be your friends after. But don’t stay because you’re comfortable.

3. Don’t think that there won’t be politics because there wasn’t politics before. Politics emerge when the players believe the game is zero sum. In a recession, the players are more likely to believe the game is zero sum.

4. Maybe the turnaround will work. Maybe it won’t. You have to decide whether it’s worth waiting for. Your existing time is sunk costs.

You’re not sure of how much risk you want to take

1. It’s okay to go for the safe job. Sure, Airbnb was founded during a recession. 50% of the time, you’ll survive the charge across no-man’s-land and reach the enemy trench.

2. It’s also okay to take risks. Staying at a company that’s slowly dying has its costs too. Stick around too long and you’ll lose your belief that you can build, that change is possible. Try not to learn the wrong habits.

3. Just… try not to fuck up the ratio. You can take the risk but if you do so, demand some of the upside.

You’re finding a new job

“When a management team with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact. - Warren Buffett, maybe

1. When picking a job, yes, your manager matters. But if you have an amazing manager at a shit company you’ll still have a shit time. In some ways, it’ll actually be worse. If they’re good at their job (including retaining you), they’ll keep you at a bad company for too long. And then they’ll leave, because they’re smart and competent. Maybe they’ll take you with them.

2. Your equity package is a lottery ticket with expected value of zero.

3. VCs are salespeople for a living, are constantly talking their own book, and only have to be right 1/10 of the time. Even the ones you trust to have your best interests at heart are often wrong. Don’t ever join a company solely on the recommendation of a VC. Even if they’re a board member. It’s the nature of boards that they ignore (externally) problems until their hand is forced.

4. “Take any role, at any pay, on a rocketship and everything will work out” is only sort of true. It’s the Hollywood model… do it for the exposure and it might be great! Technically true, but it’s also a way for people with more information to offload the risk of bad companies onto… you.

5. Seriously, you have to form your own opinion. Hyped = hiring good PR != successful. Every time I’ve outsourced my thinking for a job change (n=2), I’ve missed out on 5-6 figures of equity appreciation. But, who the fuck knows.

Lastly, it all tends to work out, on average, eventually.

If you thought this was useful, please share it. Or tell me your own stories, and I’ll share them in a follow-up.

Share

The personal loan market has blown up in the last ten years. 20.2M Americans have unsecured personal loans as of 3Q19 with $156B in outstanding balances, thanks in large part to QE, LendingClub, and Prosper.

Over the years, we’ve learned a few lessons:

• P2P doesn’t work at scale because there are very few individuals who both have money, and want to actively pick loans

  • After Prosper failed to raise/sell in 2016, it essentially became a captive origination platform for institutional buyers

  • Upgrade (LendingClub founder’s second try) went full institutional

  • LendingClub has gone from 16% self-directed retail (1Q16) to 4% SDR (3Q19)

• Institutional pickers, however, are still quite hungry and opinionated

  • GS decided to stop buying from LendingClub and build its own consumer division (Marcus)

  • Entire businesses like Theorem have emerged to buy consumer credit

  • LendingClub has launched LCX, a real-time bidding platform for loans

  • Lesson from Product School re: online car sales: Who’s the real customer?

• Alternative data for underwriting is, so far, a dud. Everyone above a certain scale basically uses FICO.

  • ML is tricky to use in a legibly/provably-not-racist way

  • A lot of data are still unavailable in easily consumed ways and introduces friction

    • Plaid / {utility,cable,cell bill} / TurboTax needs consumer consent and a login

    • However, Credit Karma and LendingTree are slowly accumulating this same data

  • What also matters is how the product is sold - if you’re securitizing a block of ‘prime’ it’s a lot easier for your capital markets team to point to FICO than to say with a straight face ‘our unseasoned model thinks that this person who’s at 620 FICO is actually more like 710 equivalent’

• All the money that was saved by not having branches… went toward marketing expense.

  • Retail/branches are CAC. Or CAC is the new branch cost. Same as e-commerce.

  • In 3Q19, 44% of LC fee revenue went to marketing expense

So, what is the value that these originators provide? They get a ~5% fee on originations, of which ~3% is marketing expenses and ~1.5% is origination expenses. And a lot of those origination expenses are either undifferentiated/commodities (credit pull, The Work Number), or not core competencies / outsourcable. Looking at LendingClub, ~40% of its non-core workforce is outsourced, which is probably most of the call center.

In the third quarter, 19% of our total workforce resided in our BPO partnerships, while 48% of our workforce was located outside San Francisco - LC earnings transcript, 3Q19

So, in summary:

1. Buyers of consumer credit will largely be sophisticated bidders. The real customer in this market is the institution, not the consumer.

2. There’s not much secret sauce in underwriting, and a lot of the useful data are difficult to aggregate

3. Originators (or origination partners of WebBank / Cross River) fail to keep most of the value (approximated by revenue) that they generate, losing it either to customer acquisition (Credit Karma, LendingTree, Nerdwallet, Google, Axciom, credit bureaus, and the good ol’ USPS ) or origination expenses (BPO)

What does this lead to? What other market does this start to look like? Stay tuned for part 2 (hopefully tomorrow)…

Share

Hey, just as a heads up, no big deal, it turns out our whistleblower system wasn’t recording submissions properly, so if you submitted anything in the last couple months, could you please re-submit it?

Everything’s totally fine, btw.

A few weeks later, the CEO was fired for ‘not cooperating with the board’s investigation’. What’s came out in the later lawsuits is that CEO had done all sorts of things - including some self-dealing, gray area stuff, and technically-not-fraud-but-smells-like-fraud for years. The all-star board tolerated it until growth started slowing, and the stock price went into reverse.

Sketchy stuff happens at pretty much every company, venture-backed or not. Startup or not. Especially when you have incredible pressure to grow. Leadership makes (hopefully) conscious decisions to take on certain risks and tradeoffs in service of the company’s goals.

Typically, investors do too. Their tradeoff is reputational risk vs. chance of returning the fund. Nobody shit on Facebook until recently. I’m sure there were other Susan Fowlers, early on, who disappeared into the void because growth papers over everything. There were few benefits to calling out Theranos. We tolerate a degree abusive behavior, sexual harassment, and fraud as a cost of doing business.

We publicly state that we have zero tolerance. We believe we have zero tolerance. This makes it easier for us to justify our uglier motives. As the music stops, sketchy behaviors both become more tempting (just juice the numbers this quarter, and we’ll make it up in the next!) but also less tolerable.

Employee behavior changes too. Competent players recognize the shift from positive-sum to zero-sum, and either exit for greener pastures, or switch tactics to value extraction. I suggest you get out before this happens.

In my observation, a few behaviors start to emerge:

1. Savior projects that happen to look good on a resume

   “What’s holding us back is that we don’t have a {microservices architecture, machine learning capability, splashy brand marketing campaign}. If only we had that, all our problems would be solved. I can launch an MVP in ~3 months, I can see if it works in 4 months, and I’ll be out the door in 5.”

2. Extremely well-compensated savior hires and/or consultants

   The sociopaths have fully recognized that it’s time for value extraction, and line the pockets of their friends. The board hires a niche consultancy and/or their friends from McKinsey. New VPs of Strategy, Product, Chief Architects, CISOs and similar come in to fill the shoes of the departed positive-sum players. They often look like this:

3. Conversation shifts away from the medium term

   People either start looking for projects that get them immediate reward (short-term) or talk about flashy long-term projects they have no intention of fulfilling. Marketing starts paying more and more for volume, and the exec team gets bogged down in strategic planning reviews.

4. Empire building

   Because American corporate culture values {having a lot of reports, managing a large budget, having large scope}, the remaining players begin trench warfare fighting over remaining resources. While management is distracted, OPEX inevitably creeps up. Admin teams like Compliance, IT, and HR balloon as they all think of infinite risks to mitigate and how they need more headcount to do so. More ‘offsites’ are scheduled to address the ‘high turnover’ problem. This is also the perfect time for an enterprising Oracle salesperson to start buying people dinner.

Finally, cost cutting. A PE shop comes in, dumps the LaCroix, and manages remaining cash flows until the carcass and IP can be dumped elsewhere.

In the process, largely by accident while finding ways to serve better ads, we've managed to use software to enhance our empathy, to grow our social sphere, and to make ourselves better people. But our implementation has been poor.

Most of our energy in social media has been devoted to moving the offline online, in order to make it parsable to computers that serve advertising. Online creates scale, which is wonderful. But with scale comes metrics that don’t work with our brains. To the hunter-gatherer brain, engagement ratios are the terrible, terrible knowledge that you are unimportant and being ignored.

There's a reason why Instagram is experimenting with hiding likes. Illegibility is a social lubricant - or rather, social legibility through software is like being knurd.

Being drunk is to be intoxicated by alcohol to such an extent as to be unable to perceive the world clearly through the senses.

Being sober is to be able to perceive the world clearly through the senses, yet humans are quite capable of giving themselves illusions and little stories to make life more bearable.

Being knurd is to be (un)intoxicated with Klatchian Coffee to such an extent that all such comfort stories are stripped away from the mind. This makes you see the world in a way 'nobody ever should', in all its harsh reality.

- Discworld and Terry Pratchett Wiki

Theory says there's a hard cap to the number of social contacts we can have - Dunbar's number. 150 relationships. We can do better with technology. There's a hard cap to how much weight a human being can lift, but that doesn't stop us from building levers and forklifts.

But I have a few observations:

500 is roughly the number of relationships young millennials can sustain with the current social media stack. Above that number of Instagram followers, and behavior approaches a broadcaster/audience dynamic. This is pure anecdata.

The one-to-many opacity of Snapchat’s messaging (especially pre-Stories) was powerful. When someone sent you a Snap, you didn’t know whether it was just meant for you or for many - and that is in essence what allows for many friends at scale. Instagram Stories are powerful not only as a broadcast medium, but also as an effective entry point to Instagram DMs, which are 1:1.

The curation (potentially) necessary for higher-quality interactions restricts growth. See: Coffee Meets Bagel starting from 1 person per day, now at 20+ per day. At the logical end of that path, Tinder’s infinite swiping drives interactions, screen time, revenue.

There are now so many more ways to find your tribe. Now you can download the official Jeremy Renner App (yes I know it got ruined) and connect with fellow fans. There will be more communities like this. And I think that’s generally good thing.

Anonymous vs. pseudonymous vs. real names has tradeoffs. Anonymous and you’re interacting with the void. Real names require bravery for genuine interaction. Pseudonymous provides optionality but at the cost of inconsistent behavior by different actors.

We have an ever-spreading array of ways to gain self-perceived status. For most, the economic game is still fucked; but now you can have a cool build on your minecraft server. You can be perceived as extra rational on rationalwiki. You can get thousands of upvotes for Reddit comments that all involve your dad beating you with jumper cables.

Here’s some elements I think might work for a ‘bicycle-for-the-heart’:

New units of engagement. The social media giants are fundamentally architected around posts, likes, and comments. Not at all how humans naturally interact… but nice and machine parsable.

The birthday reminder -> birthday box -> click to say happy birthday evolution, while telling us exactly how many friends said Happy Birthday or clicked the button, is a quintessential example of why Facebook is broken. The metrics are empty.

Force a higher bitrate. A click on the like button gives very little information. 280 characters is better than 140. Videos are better than photos. It’s harder to hate real people.

Illegibility as a feature. Disguise the scale. Hide whether a message is one-to-many. Prevent screenshots. Hide the number of likes. Hide the number of followers. Track karma internally, but don’t share it. Stats are bad for users.

I’m not an angel investor, so please don’t intro me to founders who are building the next social X unless they’re cool

bonus paragraph left on the cutting room floor:

The paperclip maximizer is already starting to reach out of the factory. So long as social media is about making things more legible, we will get more extreme, more bifurcated, more segmented. Your objective function is effective targeting. It's all good to draw circles group people into nice, neat segments. Isn't it better to get them to sort themselves into segments? Isn't it even better to widen the distinctions between those segments?

In this piece I attempt to argue that the uniquely high-cooperation culture (“let me know how I can be helpful”) of Bay Area Tech is driven on an ecosystem level, but that might collapse if/when growth slows.

Let’s start with context:

Premise 1: Bay Area Tech culture cooperates by default; other cultures don’t necessarily

Premise 2: The rules can change

Instead of trying to be unique, I’ll just quote from Jerry Neumann’s excellent The Deployment Age (emphasis mine):

Everything you’ve learned in your career has to be re-examined every once in awhile to see if it will be as true in the future as it was in the past.

Some things we’ve learned over the past 30 years–that novelty is more important than quality; that if you’re not disrupting yourself someone else will disrupt you; that entering new markets is more important than expanding existing markets; that technology has to be evangelized, not asked for by your customers–may no longer be true. Almost every company will continue to be managed as if these things were true, probably right up until they manage themselves out of business. There’s an old saying that generals are always fighting the last war, it’s not just generals, it’s everyone’s natural inclination.

On to the argument…

Most startup advice I read is either from already-successful entrepreneurs or VCs, and both the fortune cookie tweets and essays heavily lean toward the “cooperate”, “infinite games” type of thinking - play the long positive-sum game and in aggregate everyone wins. This is absolutely the right advice on an ecosystem level.

In the prisoner’s dilemma, it is rational for each criminal to defect. It is also rational for their hypothetical crime boss to encourage cooperation. It probably also helps established players to have a no-shivving culture. The pool of public advice is also massively skewed away from the ‘defect’ column because nobody wants the reputation of being a defector! It’s left to Robert Greene and (still!) Machiavelli.

Incentives for powerful players aside, the pro-cooperation advice is probably mostly genuine, coming from personal experience. Historically, cooperation has been the good general strategy in the Valley. Historically, it HAS served you well to be founder-friendly in the venture business.

Right up until you need to fire a CEO to ensure that your 10x-returning-the-fund-investment goes public. Quoting from Ben Thompson’s excellent The Uber Dilemma (emphasis mine):

That almost assuredly changed Benchmark’s internal calculus when it came to filing this lawsuit. Does it give the firm a bad reputation, potentially keeping it out of the next Facebook? Unquestionably. The sheer size of Uber though, and the potential return it represents, means that Benchmark is no longer playing an iterated game. The point now is not to get access to the next Facebook: it is to ensure the firm captures its share of the current one…

…In other words, an iterated game is good for founders: it ensures venture capitalists are nice. Single move games, though, which Uber has become, often end badly for everyone, particularly founders.

In the current, stable era, conflicts have been infrequent. In the Valley, pretty much everyone’s been winning. Our industry is so young that we’ve only really seen a boom. Few of us actually went through the dot-com bubble. Our culture has been shaped by that.

It’s not clear to me that the Valley attracts especially cooperative people anymore. Even before you get into the whole invasion-of-the-MBAs rhetoric - it’s just too big. You have to have reversion to the mean.

We can see the cooperate-by-default strategy isn’t as common in other places, in other industries. It’s not as common in Asia. It hasn’t been common on Wall Street. It’s not true in politics.

I spent some time at a company that went from growth to cost-cutting and was astounded by behavior changes from people I’d previously thought were the nicest. Maybe I was just naive.

I’m not trying to scare anybody, we should just be careful to note that public advice and perception will never catch up with private activities - even more so should the environment shift from growth. Don’t go into meetings expecting to get shivved, but maybe don’t label your vital organs.

Related thought that didn’t make the cut: In the broad context of history, the current era is uniquely stable

(i.e., absence of large-scale conflicts, reliable trade / contract enforcement, etc.) due to a combination of the following and other factors:

1. US-led geopolitical stability, globalization, etc. in the after WWII

2. Productivity gains from effective exploitation of fossil fuel energy

Further Reading: The Accidental Superpower, podcast interview