Micha Kaufman

Founder & CEO @ Fiverr (NYSE: FVRR)

[Micha Kaufman @michakaufman]

Almost ten years ago, we ran a campaign that pissed off the @NewYorker.
In Doers We Trust.

The idea was simple: success belongs to people who do. Not the people who talk, plan, or wait for permission. We didn't invent it in a boardroom. We came up with it by looking at how we actually worked.

The backlash came fast. The New Yorker ran a piece titled "The Gig Economy Celebrates Working Yourself to Death." The charge was that we were romanticizing the grind and dressing up a world with no safety net as an aesthetic.

I understood the criticism then, and I still do. But here's what almost a decade has clarified for me: The objection was never really about effort. It was about how much effort it used to take to get anywhere, and how little protection people had while they spent it.
That's the part that changed. AI didn't replace doers. It collapsed the cost of being one. People are shipping products while they sleep, one finger holding a laptop open so an agent can finish the job overnight. The grind that critics worried would consume people is being handed, piece by piece, to machines.

What's left is the part that was always the point: deciding to build something, and starting.

In Doers We Trust. It was true then. It's more true now, and for a better reason than we could have argued in 2017.

[Micha Kaufman @michakaufman]

A/B testing is really a shitty tool for understanding your users.

An A/B test is an averaged result across all the cohorts that ran on A and B, but those cohorts aren’t uniform, they’re full of nuances. An average can’t surface insights, can’t tell you what the better investment is, can’t make critical distinctions. The moment you break that average into its cohorts you might discover it was hiding the fact that one audience saw a moderate lift while another saw a drop that’s nothing short of a disaster. But on average it looked better, so you saw a green number and moved on.

And going just for the average is the choice to stop at the number that’s easy instead of the stories that are true.

[Micha Kaufman @michakaufman]

/4
Virgil proved that you can change the world by changing just 3%. The trick, of course, is knowing which 3%. I never got to meet him. I wish I had. Close to his passing, and unaware that he was sick, I tried to reach out to discuss a donation to his scholarship fund supporting Black students in fashion. It was too late, unfortunately.

41 years wasn't nearly enough for someone with that kind of mind and soul. But what he left behind is still shaping how people think and build.

RIP Virgil

[Micha Kaufman @michakaufman]

/1
A fashion designer changed how I think about building companies. His name was Virgil Abloh, and if you don't know who he is yet, keep reading, because his 3% rule applies to way more than fashion.

Virgil was an architect, a DJ, a fashion designer and entrepreneur. He passed away from cancer in 2021, at 41. He was one of the most original thinkers of our generation.

(📷 Photo by Sebastian Kim for Vogue)

[Micha Kaufman @michakaufman]

/3
We romanticize creative freedom. Give people space, remove the limits, let them dream big. But the most interesting breakthroughs in human history, in art, in science, in business, didn't come from total freedom. They came from scarcity and boundaries. Give someone an impossible deadline and a tight brief, and you'll often get better work than giving them all the time and space in the world. Constraints don't limit creativity, they focus it. I think about this when I watch teams flail with too many options and too much room to wander. The answer is usually not more freedom, it's a sharper frame.

[Micha Kaufman @michakaufman]

Data made us lazy. Someone has a chance to talk to a customer and they say "I don't need to, I have it in the data." But dashboards only tell you the end of the story. If you built an app, sit on the App Store, read the reviews, pick up the phone, get on a Zoom, because there is no greater gift than a real conversation with someone who uses your product.

[Micha Kaufman @michakaufman]

The story of how I almost founded OnlyFans - only to shut it down!

When we started Fiverr in 2009, we had no money. It was a bootstrap business, which meant it had to have a business model from day one, and it had to have stickiness and virality baked into it, because we needed grassroot growth. So we came up with this idea: a single price point, five bucks, and a syntax inspired by a site called FML (if you remember it, you remember it). You land on the homepage, you see "I will _ for five bucks," you type something in the box, click submit. It seemed too easy, so most people tried it. Two minutes later, you had a live offering. It was irresistible. People had to engage with it.

And it worked. The tagline we landed on was "the things people do for five bucks." Turns out, people were willing to do the quirkiest things for five bucks: a professional juggler juggling working chainsaws, people recording happy birthday videos for your loved ones, and then there were offerings that were, let's say, very similar to what we now know today as OnlyFans.

I'm not gonna lie, from a business perspective, we had a dilemma. But from a moral standpoint it was clear to us that this is not our path, and we cleaned it up.

That was a long time ago. After generating billions of dollars for millions of people around the world, helping businesses start and thrive, I can't say for sure that this was a good business decision, but it was definitely one that helps us sleep better at night

[Micha Kaufman @michakaufman]

Studies show: 90% of people who start a conversation with ‘let’s think outside the box’ are about to suggest something very inside the box

[Micha Kaufman @michakaufman]

"Data driven" is the most overrated phrase in tech and I say that as someone who pushed it for years. I used to tell people at Fiverr "if you don't have data, don't talk to me."
It trained people to stop asking why. You win an A/B test, the numbers go up, everyone moves on, but nobody stops to understand what actually made it work and whether they could replicate it ten more times.
People don't ask "why" enough, and that curiosity is exactly what lets you take one insight and build something bigger from it.

Data should be the beginning of a question, not the end of one.

[Everything Marketplaces (EM) @marketplaceshq]

.@michakaufman shares how @fiverr has evolved and scaled.

[Micha Kaufman @michakaufman]

A meeting is where conviction turns into action. When it turns into people saying what they ״think״ and ״believe״ and ״need״ with nothing attached, that’s not a meeting, that’s a support group.

[Micha Kaufman @michakaufman]

As the CEO of a public company, I use AI in every step I take each day. Let me show you what that actually looks like.

I built what I call the Micha agent.

The idea behind it is simple. Real AI usage isn’t about sprinkling a chatbot on top of your day. It’s about building a system that runs as autonomously as possible, pulling me back in only where I actually matter. A lot of what I do every day doesn’t require my judgment at all. It has a logic to it, which means it can be taught, and it can be repeated. The whole point is to clear out the autopilot work and protect the hours for the things that genuinely need me.

Email triage is the obvious one. Who wrote this, does it even need a response, and if it does, what would I say. I get all of that as a summary, sometimes with a draft already sitting there in my voice. Before a product review, it knows the questions I always end up asking, so it preps the ground before I walk into the room. Before a call, I can ask it the questions I’d want someone smart to ask me: what do I actually think about this, what am I missing, what haven’t I considered.

But the tasks aren’t the interesting part. The interesting part is that I teach it to become a slightly better version of itself after every single one. Every time we finish something together, I want it to walk away having learned something that means I never have to do that thing again.

The whole thing is in the feedback loop. When we’re done, I ask it to tell me what we learned, not about the task itself, but about how the two of us work together. Then I ask it to turn that into a framework, a small set of principles, and to rewrite those principles as instructions it saves to memory. Over time it gets sharper, more specific, more me.

Here’s the part that humbles me. If an agent trained on me can do something well, that’s a strong signal I probably shouldn’t be the one doing it. It forces an uncomfortable clarity about where my judgment is actually irreplaceable and where I’ve just been telling myself it is. That clarity is the real product. It’s how I make sure that when something truly needs me, I’m fully there for it.

[Micha Kaufman @michakaufman]

That's the gap between getting 80% out of AI and getting 99%. Nine people are operating a tool. One is architecting a system.

The 100x performers right now are not the ones writing the code or drafting the spec. They're the ones orchestrating agents, choosing the right tools, defining what good looks like. Their entire job has collapsed into judgment.

We're all typing into the same box. What separates people pulling extraordinary work from everyone else is how they think before they hit send.

I'm not hiring people who can prompt. I'm hiring people who can architect.

[Micha Kaufman @michakaufman]

I've added one task to interviews lately, inspired by a conversation with my brother, and the answers tell me almost everything I need to know about whether someone is genuinely AI-native or just performing it.

The task: "Produce the most photorealistic image of exactly what your eyes see right now. One constraint: verbal methods only, no photos. Go."

I've given this to ten people. Nine did some version of the same thing. They started describing the room, the wall color, the furniture, the light through the window, who's sitting where. A few skipped straight to naming a model and prompting. They differed in approach but had one thing in common: they started executing immediately. They treated AI as a tool to operate.

One person did something completely different.

[Micha Kaufman @michakaufman]

Before prompting anything, they built a system.

"Before I generate, I need to build a scene specification. Tell me the camera position first, everything else is defined relative to it. Then room geometry in measurements, not adjectives. Light sources by type, direction, and color temperature. Each object as a coordinate, not a vibe. If there are people in the frame, I need posture, facial expression, what they're wearing, described precisely. Before I start, tell me what matters most so I know where to spend my specification budget. And let's agree now on what done looks like, so we're not iterating on feel."

[Micha Kaufman @michakaufman]

Goodbye, Boss!!

That’s what Jeff and I used to call each other. I got to say goodbye to him on his last day at Fiverr before he headed back home to South Africa, where his wife and three kids were waiting for him.

Jeff was part of our cleaning crew this past year. With a positive spirit and a work ethic I’d take on any role on any team I’ve ever built, Jeff’s smile reached me in more than a few moments when I really needed it, and he probably had no idea.

Thank you, Jeff, and the entire crew, for the work you do that too often goes unthanked. You’ll always have friends at Fiverr, Boss.

[Micha Kaufman @michakaufman]

@yoavrimon Not really, because the point isn’t the task, it’s the deadline. “Next week” is not a deadline. But “Wednesday by noon” is. Now we have a touchpoint, it’s on your goals, not mine, and there’s actual ownership. That’s the difference

[Micha Kaufman @michakaufman]

Someone told me yesterday they’d “circle back next week.” That’s one of the things I hear that’s driving me crazy. Next week from what? Monday? Friday at 4:59 PM?

I hate it because it’s a form of corporate patience I fucking can’t bear. Why wait three days for an email reply from someone sitting on the same floor? Why write a small task on your to-do list when the same energy gets it done?

Impatience gets a bad rap. People confuse it with being harsh, aggressive, or difficult to work with. It’s none of those things. Impatience just means you believe both you and the people around you are capable of moving faster. That’s not being rude. That’s how things actually get done.

Most companies don’t slow down because they got the strategy wrong. They slow down because they’re patient with everything, including what should’ve moved yesterday.

[Micha Kaufman @michakaufman]

@jordanlwalker Not this time, but we talked about taking some fishing gear with us next time

[Micha Kaufman @michakaufman]

This trip lived in my head for years. One of those things you keep saying you’ll do when things slow down. They never slow down. So I just stopped waiting.

A few close friends, a light bush plane, a tent, a sleeping bag. Ten days flying from Nevada all the way to Washington. Landing on snow, gravel, and patches of grass next to lakes your eyes refuse to believe are real. My dream trip.

It came true! And I’m grateful for every moment of it.