| Those of you who have had a dinner conversation with me inevitably know that I love bringing up the measurement of “micromort” and rattling off various dangerous activities and ranking them by risk. Now I get to do this in digital form. OK, what is a micromort? Micromort: A 1 in a million chance of death.
You can use this measurement to compare/rank the risk of any activity, and in fact your favorite LLM is quite good at generating tables of activities versus micromorts. You can even qualify certain activities, like “tell me the micromorts of road cycling, but only during the daytime in suburban/rural areas.” Here’s a handy graphic that shows some micromorts for a wide array of activities. (tldr; don’t mountaineer, BASE jump, or get old): This is an endlessly interesting measurement and you might ask yourself, isn’t this about exposure? If I cycle for 10 hours a week versus sky dive once in my life, how does that compare? Of course you need to estimate micromorts per hour — and an LLM is great at this — and normalize the exposure. Another fun prompt is to ask, “based on what you know about me, what do you think are my highest micromort activities?” Endless fun. I’ll leave it to you as a reader. Other ones I like: Cost per hour of pleasure (CPHP): Going to a concert or basketball game is high CPHP. Buying a super nice treadmill that you use 3x/week is low. Basically anything that you use all the time tends to be low, unless you’re fooling yourself that it’s an everyday forever product and it’s actually going to collect dust in your closet (yes, I’m very guilty of this). Complaints per Hour: This is as much for you as when you talk to your dear friends that are just going on and on. High CPH means high negativity. Maybe mostly a metric to lower? However, if it’s very enjoyable to complain (and sometimes it is) then it’s free, so it’s also effectively infinite zero CPHP, so maybe everything cancels out. Phone Pickups per Hour: If you are with boring people, or hanging out somewhere boring, the number of phone pickups is very high. As many of you know, when I wrote a book a few years back, the process is so intense (and sometimes very boring) that I had many many phone pickups. Eventually I just locked the phone away in a timed safe. Seriously! % Conversational Autopilot: We all hate boring conversations, but we can’t help ourselves. But we all end up stuck at group dinners or sometimes you take a random 1:1 meeting that just turns boring. You introduce yourself, you talk about your job, then maybe you talk about places you’ve traveled. It’s the lowest common denominator conversation, and you just put the conversation on autopilot because you’ve heard and said all the same things already. So what % of the convo is in this mode? Personally I find that great chats often hover around 20-30% — enough commonality to be comfortable, but then you tread new ground that will be memorable later. But let’s admit sometimes they’re like 75%+ and that’s when you want to dine and dash.
As you read through these, I think you realize (rightly so) that these ratios are mostly a nerdy way to complain about people and situations. It’s true. I think of my writing as usually low Complaints Per Hour but now you are getting a better sense of my inner self. The title of this essay promised you something more than micromorts and phone pickups though, it promised some new business metrics! And I have some for you. These are real metrics I have heard used or use myself: Lies per Second: In startupland, there’s a fine line between pitching and lying. Hopefully you’re just telling the truth attractively, but sometimes people are just making up numbers, saying they have customers when it’s just pipeline, removing labels from their graphs, and so on. In that case, the “lies per second” might start high and go even higher as the presentation progresses. When you find yourself in a high LPS meeting, it might just be time to end it early. Meetings per Decision Ratio: There are big decisions and little decisions. How many meetings did it take to make that decision? Take something simple like deciding to hire someone — some places have an organized, streamlined interview loop of 3-5, a presentation, and then boom, you decide. Other places it’s open-ended and you might have 3 initially, then another 5 then 1 or 2 more, then another 3, and so on. Why does it take so many meetings to make a decision? Sometimes there’s no owner to the decision so it’s unclear who has the authority to even give the greenlight. Sometimes it’s an ill-defined goal or process. Sometimes it’s constantly changing. Either way, a high MPDR tells you that something is broken, and the situation needs to be escalated to a decision maker to simplify. Time to First Excuse: You have a bad month. Why? “Seasonality.” Your customers hate your new product. Why? “Our marketing was poor.” You often sit in a review meeting and you can count the minutes before a poor result is presented, along with an excuse for why it happened. (It’s never the team’s fault, by the way). A low TFE often leads to a low TNE — that is, Time to Next Excuse. As the TFE approaches zero, this is when you know the team probably needs to get replaced/rebooted. Numbers vs Text Ratio: In startupland, one of my favorite stages is the very very beginning when the idea is being formed. That’s when it’s all story, and all text/narrative. There’s no numbers because there’s no revenue, no customers, and no product. Fast forward a few years though, and during a Series B, you should start to see numbers everywhere. You want a presentation that’s just chock full of retention curves, financial metrics, etc, etc. That’s why the “numbers vs text ratio” should start getting very high. Sadly, often that is not true. Sometimes you see startups that have raised $50M show up and just present a deck that’s all numbers. If the Numbers vs Text Ratio is too low, then the Meetings Per Decision metric is going to skyrocket. Or it’ll be low, because no one will want to invest. PowerPoints per Launch: Every product team in Silicon Valley has their own culture. Many startups learn by shipping — they launch features, see how people react, then launch more features. That’s zero powerpoints per launch. Somehow every startup turns into a big company over time, and there’s more coordination costs, and you find yourself unable to launch new initiatives due to “death by PowerPoint.” You want to pitch a new idea? OK make a powerpoint. You gotta go roadshow it though. Make more powerpoints. How about the market research that tells you this is a good idea? How about what our competitors are doing? What should be the KPIs? Make more powerpoints. The only way to reduce PPPL is to involve senior executives who want to push it forward, otherwise you’re in consensus-building hell. Dollar per IQ Point: If you hire a top 1% stanford CS student, believe it or not, you’re probably spending moderate dollars per IQ point. You might get an even cheaper Dollar per IQ Point if you hire a top 1% student from a state school — the IQ points may not be as high, but they won’t demand the king’s ransom. How do you hire the most expensive Dollar per IQ Points in the world? Probably by hiring a bottom 10% Stanford student :) Decision to Rumination Ratio: There are big decisions and small decisions. Sometimes you take a lot of time — weeks — to think through something, and sometimes you make the decision in a snap. Weirdly, they are often not correlated at all. It might take you months to pick out a new car — researching, test driving, etc., but then you might jump on a new opportunistic job with only a week of thought. Ideally, you’d like to think that you normalize this ratio so that it’s all relatively the same — so that you take a lot of time for big decisions, and only a little time for small decisions. But we all suck at this. One important prompt: What are the most important decisions in your life? Probably who you marry, what city you live in, and your career. How many of us are willing to take months or years full-time to actually figure these out? And to talk to experts? And to get expert help? Instead, we just kinda do what feels right and sometimes it works and sometimes it doesn’t. Thus — shotgun weddings, moving to a new city on a whim, ragequitting a job, impulsive breakups, getting a tattoo while on vacation, impulse buying a pet, etc., etc.
These are obviously fun/conceptual ideas, but one day I’m sure we will be able to measure them once we’re tracking all the meetings we have via our AI note taking apps! And maybe we can even have alerts automatically fire when the Lies Per Second metric exceeds a certain threshold, since we’ll be able to automatically fact check every sentence on a video conference. I’ll stop here, but if you have any examples to add, please reply in the comments! Always looking to incorporate more of these into my dinner conversations along with micromorts and so on. You're currently a free subscriber to @andrewchen. For the full experience, upgrade your subscription. Upgrade to paid | |
No comments:
Post a Comment