PreviewThis page isn’t published yet — you’re viewing a shareable preview link.

Making the squishy legible

The things you most want (and don't want) in teammates are the hardest to name. Turn gut feelings into criteria you can actually interview for.

What is making the squishy legible?

A lot of what people do at work is easy to read. You can see whether someone writes a clean email, builds a spreadsheet you can follow, or holds a room on a sales call. That stuff sits on the surface.

The hard part is drawing a fence around who belongs on the team and who does not. Doing so is deeply uncomfortable, because it is judgmental, it is exclusionary, and it runs on preference. So we reach for objective criteria over subjective ones, hoping to be fair. But much of what actually matters about working with people is hard to put your finger on. Those are the squishy things.

The squishiest of all is whether it is okay to simply want to like the people you hire. Positive preference feels loaded with bias, because it is, and ethical leaders stay on alert for their own. Too often, though, the effort to be bias-free mutates into something counterproductive: the leader overrides a real read and defers to a borrowed set of values that sounds more acceptable, and builds an ineffective team as a result. That reflexive defense against bias is what keeps most people from ever making the squishy legible.

To make the squishy legible is to take one of those fuzzy reactions, “I don’t like how that person talks to the team,” and finish the job. You name the behavior underneath it, you make the case for why it helps or hurts the team, and you turn it into something you can say out loud, filter for, interview for, and call out.

But squishy things aren’t all bad. Dr. Jenna LeJeune and Dr. Jason Luoma call a value an appetitive, something you are hungry for, and an anti-value an aversive, something you want to push away [1]. Both deserve to be made legible, and most of us are lazier about the appetitive: naming what repels us comes much easier than naming what draws us in.

Fear of bias freezes many leaders in their tracks. Here’s an antidote: get specific. Interrogate “I don’t like this person” and you almost never land on “they went to a different school” or “they don’t look like me.” You land on a behavior that genuinely corrodes a team: the person is insubordinate, or mean, or unreliable, or bad at the job. Don’t let the fear of bias stop you from getting specific about what builds a great team and what wrecks one. More often than not, that fear is what does the damage.

This is why “culture fit” is not quite the cop-out it is accused of being. It is the first step, un-followed-up-on. “I don’t think they’re a culture fit” is so general it tells your team nothing. Finish the job. Why not? Name the behavior you can get agreement on and defend as genuinely counterproductive, and you can build a filter and run it as a test. You might be wrong about which trait predicts success, and that is fine, because having an opinion about it is the job.

Skip the work and be prepared for two things to happen. First, you’ll hire the wrong people. It will feel unpleasant, and then their output won’t be up to snuff. And so, second, the day of reckoning will arrive and they’ll either quit or you’ll have to fire them. Here, you have a choice: leave it illegible or solve the root cause once and for all, get clear on what you failed to codify, and make a change to how you filter people in and out.

Underneath all of this sits the idea of permission. In “Managing Oneself,” Peter Drucker argues that bosses and coworkers alike are individuals entitled to work their own way, and that what matters is whether they perform and what their values are [2].

One of my favorite things to tell leaders is that the one altar they should worship is that of effectiveness. Effectiveness is partly a matter of preference. If you are at your best around concise communicators, then hiring concise communicators is just you taking responsibility for the team you have to build.

But another way of thinking about it: maybe all the talk of effectiveness is just a permission structure. To put it as Claire Vo does, “sorry homies, this is how I roll and I’m king of the castle.” [3] No apology required.

Someone great at this catches the small signals instead of punting them into the background of our noisy mental chatter. Michael Lopp calls the sense that something is off a “Twinge,” arriving in the unstructured way experience speaks to you [4]. His line belongs on the wall: “your failure to heed your Twinge is a management failure.” Lopp means it about the whole organization; the version that matters here is the twinge about a person, especially the unpleasant one.

Noticing is only the beginning. You feel the reaction, you act on it, you put your finger on the specific thing, and then you make it legible to everyone else with a label made of words other people hear and repeat, saying, “Yes, my god, that’s it!” You circulate it, you teach the team why it matters, you win their buy-in, and you codify it into a filter the team actually uses. Private clarity does not get you there; the team has to be able to run the filter without you.

And the great ones own their preferences without laundering them and without apology, interrogating them for fairness as well, yes. That is the golden mean of confidence: you own your read without apologizing for it and without going rotten into an arrogance that assumes you have nothing left to learn and no biases at all.

At Management Craft, we chart each competency on an abridged version of the Dreyfus Model, a system developed in the 1980s to chart how humans develop skills. The original model has six rungs, going from novice to master, but for simplicity’s sake we’ve simplified it to three: novice, competent, and expert (corresponding to 1, 3, and 5 on the Dreyfus Model). Read the proficiency rubric below for this competency and try your best to honestly assess where you sit on the spectrum from novice to expert.

Figure: The Proficiency Rubric

Novice
  • Has a low-level sense that something is off (or something is good), but never labels it, and tolerates squishy behavior for a long time before acting. The lag itself costs their effectiveness and their career.
  • Hides behind "culture fit": reaches for the catch-all and stops, never finishing the job of naming what is actually wrong.
  • Lets the fear of bias, and low self-trust ("I've only been a founder a year, how would I know?"), stop them from ever interrogating the feeling.
  • Has no real opinion on what makes a teammate great; asked directly, draws a blank or repeats something they heard on a podcast.
Competent
  • Heeds the twinge, catching the strong reaction instead of letting it evaporate.
  • Finishes the job on the aversive: names the specific counterproductive behavior underneath it, whether it is insubordination, meanness, unreliability, or being bad at the job.
  • Turns the named thing into something the team can share and interview or filter for.
  • Still mostly naming what bothers them; the appetitive side, what they are hungry for, lags behind.
Expert
  • Names both directions fluently, what they are hungry for as readily as what repels them.
  • Anchors every preference in effectiveness and owns it without laundering it, while interrogating it for fairness in the same breath.
  • Runs the full social arc: labels it, circulates it, teaches the team why it matters, gets buy-in, and codifies it into a filter the team actually uses.
  • Knows how to interview for each trait, has a test per trait, runs the filter as a hypothesis, and closes the loop after a mis-hire by upgrading it.
Based on The Dreyfus Model of skill acquisition, six stages compressed here into three rungs.

The rungs above are based on the Dreyfus model.

You do not need a decade of scar tissue to start, though most first-time founders think they do: “I’ve only been doing this a year, how would I know what works?” My friend, you have been a human far longer than you have been a founder. You have built friendships, you have talked people into joining you, and you already carry a great deal of hard-won data about who lifts you up and who lands you in trouble.

Author Alan Weiss put it well: “Confidence is the honest-to-God belief that you can help others. Arrogance is the belief that you have nothing left to learn. And smugness is arrogance without the talent.” [5] Steve Magness adds that real confidence is self-knowledge rather than delusion, “being vulnerable, not delusional” [6]. Confidence here is just self-trust, and you have more grounds for it than you think. As Alain de Botton observes, “it might be brutishness, not confidence, that we hate.” [7]

So, go forth and heed the Twinge. Treat the strong reactions, to the things that make you say, “I fucking love that guy,” and the ones that cause you to wrestle with intrusive thoughts of throwing someone out a window, as legitimate signals rather than noises to suppress.

Put your finger on it. Move from “I don’t like this person” to the specific behavior. Someone sucks the air out of every room; you dig, and the real thing is that they never take responsibility and always blame someone else.

Name the positive version. Flip the aversive into the appetitive you actually want. “Blames everyone” becomes “high agency, takes responsibility.” Do this especially for the traits you admire, since those are the ones you are least practiced at naming.

Mine your network for the test. Use the Calibration Call: ask people who have hired for this how they interview for it. When I was trying to screen for high agency, an investor of mine gave me a question I still use: tell me about a time you felt slighted or treated unfairly, and what you did about it. Most candidates tell on themselves by blaming whoever wronged them. We all eventually find ourselves in unfair situations. The difference between people is how they respond.

The best answer I ever got came from a guy who shared a quota with a teammate who went out on parental leave. He asked his boss whether his number would come down, and was told no, so suddenly he had to hit double. I asked him what he did. He said, “I hit the fucking quota.” He was not going to let an unfair situation keep him from performing.

Build it into the filter, and run it as a hypothesis. Interview for the trait, reward it once people are inside, and stay open to being wrong and re-tuning.

Most importantly, close the loop. You will not be perfect at this. When a hire goes wrong, or someone quits, stop and ask what squishy thing stayed illegible, the thing you never interviewed for, filtered for, or rewarded. Then fix the system so it does not happen again.

The main way founders get this wrong is simple: they don’t do it. They sit at the novice level and decide that is fine. I have watched founders let themselves be treated like shit by people they hired and pay. The failure to make the squishy legible always shows up as underperformance eventually, because if you never filter for what actually works, you’ll hire ineffective people. Q.E.D.

Part of what makes it hard is that nobody warns you how becoming a founder is tantamount to unwittingly enrolling yourself in a crash course in knowing yourself. The whole premise of a lot of psychotherapy is that there’s a boatload of self we don’t see, sitting beneath the waterline of our own self-awareness, and plenty of people go a lifetime skimming the surface. Build a team and you lose that option. You have to work out what your preferences are, trust them, and learn to filter for them, or else the team pays for the parts of you that went unexamined, at the brutal cost of failure, sadness, and regret.

Another failure mode is darkly ironic. A good leader knows exactly how explicit you have to be when you communicate, and then forgets the rule applies to them on the one message that matters most: who belongs here and who does not.

And the darkest pitfall has to do with tolerance. When the team is not working, founders tend to conclude they need to become more tolerant of a wider range of people. Usually the opposite is true. They need to become more intolerant, which is a deeply uncomfortable idea to even propose.

Molly Graham, writing about founder mode, is blunt about how badly most of us start out. “Most new founders can’t identify the right kind of hire,” she writes, “let alone onboard them in a way that sets clear expectations or delegate in a way that makes their strong preferences or desires clear. This is why ‘hire great people and give them room’ is such overly optimistic, toxic advice.” [8] She tells a story from a conversation with Mark Zuckerberg, who said everyone thought he was a terrible manager. Her answer was that he simply had to find people who thrive working for him, because every manager has a type. Working for Mark, she writes, meant no performance reviews but “robust, sometimes bash your head against the wall debate about ideas” and a lot of direct, sometimes searing feedback. It was not for everyone, and he built a team of people who thrived in exactly that. Ed Batista wrote a whole essay whose title is the rest of the advice: don’t wait [9].

Figure: The Golden Mean

Under-doing
Apologetic tolerance: too flexible to hold a line, so squishy behavior corrodes the team
Just right
The virtue: name your preference, own it, test it for bias, and forgive the human reality that nobody is perfect
Over-doing
Rigid perfectionism: so fluent in your preferences that you leave no room for anyone to fall short
Based on the Golden Mean, originally developed by Aristotle, named by Horace, and immortalized in fairy tales by Robert Southey.

The mean here is its own kind of golden mean, with a vice on each side. Under-do it and you are too flexible: you hold no line, you tolerate what quietly corrodes the team, and you apologize for preferences you should own. Over-do it and you are too rigid: so fluent in your own preferences that you leave no room for anyone to fall short, forgetting that every hire is a trade, and you take some weaknesses to get the strengths you were really after.

The thread through all of it is a little tragic and a little funny. A CEO holds every lever of power in the company and is somehow still a tortured soul, and the torturer is them. They are the warden of their own prison yet again. The way out is inner work: figuring out what you like and what you don’t, and holding it with enough confidence to draw the line and enough grace not to demand perfection. People say founders should go to therapy to find this. Maybe. But being a founder is one of the few roles that forces the door to that self-knowledge open. All you have to do is have the courage to walk through it.

1
Jenna LeJeune and Jason B. Luoma, Values in Therapy: A Clinician's Guide to Helping Clients Explore Values, Increase Psychological Flexibility, and Live a More Meaningful Life (Oakland: New Harbinger, 2019).
https://www.newharbinger.com/9781684033225/values-in-therapy/
2
Peter F. Drucker, "Managing Oneself," Harvard Business Review, January 2005 (originally 1999).
https://hbr.org/2005/01/managing-oneself
3
Claire Vo (@clairevo), X, June 22, 2026.
https://x.com/clairevo/status/2069159292217995322
4
Michael Lopp, Managing Humans: Biting and Humorous Tales of a Software Engineering Manager, 3rd ed. (New York: Apress, 2016), ch. 6, "The Twinge," 39-43.
https://www.oreilly.com/library/view/managing-humans-biting/9781430243144/
5
Alan Weiss, Million Dollar Consulting: The Professional's Guide to Growing a Practice, 5th ed. (New York: McGraw-Hill, 2016), 159.
https://alanweiss.com/alanism-the-evolving-collection-of-alans-apothegms/
6
Steve Magness (@stevemagness), X, June 22, 2022.
https://x.com/stevemagness/status/1539593257927294978
7
Alain de Botton, On Confidence (London: The School of Life, 2017), 93.
https://www.theschooloflife.com/shop/on-confidence/
8
Molly Graham, "Founder Mode: A Liar's Perspective," Substack, April 14, 2026.
https://mollyg.substack.com/p/founder-mode-a-liars-perspective
9
Ed Batista, "Don't Wait," edbatista.com, May 8, 2018.
https://www.edbatista.com/2018/05/dont-wait.html