The Non-Verbal Expert: These Behaviors Tell You Everything You Need to Know About Someone

Guest
Blake Eastman
Read time
~5 min
Uploaded
9 Jan 2024
Duration
1:37:37
Views
469.4K
OPEN ON YOUTUBE ↗

KEYPOINTS

  • 01

    Reading people starts with mapping how wrong you are.

    Blake Eastman says everyone carries a mental model of behaviour built from past interactions, pop science and cultural biases. Before you can get better, you have to see where that model fails. The next step is behavioural awareness: noticing shifts rather than rushing to assign meaning. A head nod, for example, is usually just social coordination — 'I'm listening' — not proof of interest. Only by watching someone's nods over months could you tell genuine interest from politeness. He says the real work is video work: breaking a 20-second clip down frame by frame until people see where their snap judgments came from.

  • 02

    Know whether you're a literal or a contextual reader.

    'Shane, how are you?' can be said three ways. Some people hear no difference; others over-read every tonal shift. Eastman is the second type: his team isn't allowed to send him a one-word reply, because a bare 'yes' reads to him as anger. He needs a 'yes' plus an emoji or GIF. Clashes in organisations often come down to hyper-literal and hyper-contextual people misreading each other. That's why he likes personal 'operating manuals' that lay out your quirks up front.

  • 03

    Trust lives on a bell curve of expected behaviour.

    Behaviour far outside what society expects reads as untrustworthy: eyes darting constantly, or staring dead-on while recalling an emotional memory, when glancing away first would look more genuine. Eastman aims for conversations three standard deviations to the right — topics people rarely discuss — with behaviour that matches the interest. Trust triggers are also personal. One executive never trusted anyone who didn't shake his hand first, because his father taught him that at 12.

  • 04

    Most people only update when shown they were wrong.

    In training, he has executives predict someone's behaviour, then reveals they were completely off and why. After enough of those misses, they start questioning their lens. Leaders often don't see their own impact. One loves a direct report but delivers feedback that rips them apart. His fix is the funeral exercise: imagine everyone you've worked with at your funeral. What stories do you want told? Then he holds you to behaviour that matches.

  • 05

    Record yourself — and record your audience.

    Arguing couples remember the same event wildly differently, and video lands 'somewhere in the middle.' He says the observer effect fades in about ten minutes. For three years he recorded the audience at every third presentation he gave. 'How did my presentation go? Let's see the audience.' His coaching starts by getting people as comfortable as possible, because people are reliably at their best when most at ease. Only then does he layer in technique. Without work video, he suggests recording yourself with a friend who shows unconditional positive regard — that's how you should be at work.

  • 06

    On dates, tell stories rather than ask questions.

    In 2008, Eastman paid Craigslist volunteers $50 each for 30-minute recorded dates in a New York restaurant and got about 300 applications. The dates that looked best to viewers weren't the best. The slightly awkward ones with more depth produced deeper connection. Viewers also projected their biases: when the woman was more conventionally attractive, they assumed she wasn't interested. A story like 'the craziest thing just happened' creates far more threads of connection than question-and-answer. He warns against 'performance storytelling,' as opposed to telling a story the way you would to friends.

  • 07

    Expose bluffers with fake information.

    Deception is the hardest non-verbal problem, and he doubts any system will reliably detect it. With academics he suspects are bluffing, he invents a study — say, a 'Dillington study' from 2014 — and asks about it. When someone claims to have read it, he doesn't call them a liar. He notes that they'll trade accuracy for looking informed. Real experts say 'I don't know,' and volunteer, unprompted, when they've been wrong.

  • 08

    Machines will read behaviour, judged by perception, not meaning.

    His first big study involved 70 people manually coding 550,000 blinks, at a cost of maybe a quarter of a million dollars. Now his systems do it automatically. His company Behavioral Robotics maps facial wrinkles and movement against every spoken word. It can flag when someone shows little facial reaction at a moment that called for it. He's noticed exactly that since his father died of ALS two weeks ago, with people who don't react when told. The goal is to reveal blind spots, not to push people toward normal.

EXAMPLES

  • 01

    The cheater on the double date.

    Within minutes Eastman was '100% confident' a new acquaintance was cheating: a predatory, lingering gaze at passing women and hints testing whether Eastman was 'one of us' — he knew about a certain poker game.

  • 02

    His uncle's divorce test.

    Couples who talk only in transactions — groceries, diapers, errands — and never ask each other questions are a huge red flag. Eastman says it has predicted every divorce, and it's let him check in with friends early.

  • 03

    The J.Crew pants.

    Teaching at CUNY, he wore the same pants daily and worried. A month in, he asked 140 students what hadn't changed about him. One noticed — a J.Crew employee. No one cares as much as you think.

  • 04

    'I need a nose job.'

    After one of the flattest presentations he'd ever seen, he played a client her video. Her first comment was about her nose, not her delivery — that's where our attention goes.

  • 05

    The Rockefeller method.

    From Titan: Rockefeller asked why workers used nine or ten pieces of tar on each oil barrel and had them test how few would do. Each quarter Eastman asks where he's using more than he needs, such as SaaS bills that can be renegotiated.

OTHER RELEVANT

  • 01

    Ask predictive questions.

    'Would you invite this person to dinner with your closest friends?' gets honest answers where 'rate them 1–10' doesn't.

  • 02

    Five passages per book.

    A professor's only assignment was five highlights and why they matter. Eastman still uses colour-coded Kindle highlights.

  • 03

    Make it about others.

    Like carrying the crew's stands in unasked: you can never measure the return on a social interaction — just don't expect one.