After 15 Years and 300 Companies, He Says It Comes Down to One Question: Notes on the LUUP Episode

Those who accomplished great things in ancient times possessed not only talent surpassing their age, but also a will that could not be bent or broken.

—— Su Shi, On Chao Cuo (Northern Song, c. 1061; my own translation)

What This Episode Is About

The 2026-08-30 episode of Investor’s Sunday is part two with venture capitalist Anri Samata. He founded his own independent VC firm in 2012 at the age of twenty-seven; fifteen years and three hundred-plus companies later, his portfolio runs from consumer internet all the way to deep tech — including LUUP, the shared electric scooter service.

The first half is oddly unfinancial: he spends a long time talking about T-shirts, socks, and how to develop an embroidered logo. Only later does it turn to the substance — how LUUP got funded, what fifteen years taught him, and what Japan’s startup sector should become next.

The interesting thing is that both halves are the same argument. What follows is my own synthesis and extension, not a transcript; the judgments are mine.

Key Points

1. He treats “would an employee actually want to wear a shirt with the company logo on it?” as a serious question.

It sounds like a joke, but he means it. Would a founder want to wear it? Would a staffer? If not, the brand carries no value in their minds. He points to fashion labels whose logos come in several variants — embroidered on a shirt, stuck on a laptop, they still look good, and that only happens when design quality and brand quality stack on top of each other. So his firm genuinely pours time into embroidery and sticker development. The host’s wife took home a pair of socks from their office and still wears them.

He also explains why the firm’s website looks nothing like a financial institution’s: he is trying to attract founders, not peers. You can make yourself infinitely more serious-looking, but the people he wants aren’t at that end of the spectrum.

2. He compresses this into one line: “irrational in the short term, rational in the long term.”

Things that look like a bad deal now and only make sense over years — that, to him, is where competitive advantage comes from, precisely because nobody else will do them. His firm can stretch its time horizon unusually far and throw itself at things that look like waste. The conclusion, he says, is: so we work hard on T-shirts.

3. LUUP was funded while the service was still illegal.

The founders originally wanted to build mobility for elder care. Then shared scooters appeared in the US; they flew over to look, came back, and decided to do it in Japan. They saw the upside (young people gain a set of wheels, places that were out of reach become reachable, neighborhoods come alive) and the downside (the abandoned-scooter mess that plagued Paris), so from day one they chose a dock-based model.

The problem was that road traffic law didn’t permit it. He says the law changed only after his third or fourth investment round. And he is blunt about it: he had no idea which way the law would go. What he was betting on was something else — this team had been in the incubation office that preceded his own since the earliest days, working astonishingly hard. And the rules themselves got shaped by the whole industry association talking to police and other stakeholders together, not by LUUP charging in alone.

4. Fifteen years, from pharmaceuticals to a YouTuber agency, and he says the conclusion is the same every time: does the founder have the toughness to keep going?

They still do technical due diligence and expert calls, of course. But he files all of that under cutoff criteria — market, technology, talent, regulation are there to eliminate the ones that genuinely shouldn’t be attempted, and most startups die right there. For the small group that clears the bar, though, he thinks the only thing worth looking at is the operator.

The reasoning is practical: his holding period is six or seven years at minimum, and the business itself often changes along the way. “We know a regulatory workaround nobody else knows,” “we have access to unique talent” — those slivers of edge get filled in almost instantly. What’s left is the accumulation of one decision after another, made daily by the founder.

5. He is unusually forgiving about why people start companies, and has a great line for it.

“Every band in the world starts because the members want to be popular. But once they make it, for some reason everyone starts singing about world peace.” A completely worldly starting point is fine, he says. Somewhere on the climb, people begin asking what impact they can have and what they owe the team that followed them, and most of them end up decent. Conversely, someone who can sustain the grind purely for money is impressive in their own way — he just doesn’t think such people really exist.

So what he actually cares about isn’t purity of motive. It’s whether the passion lasts, and whether the person can manage themselves so that it lasts. Swapping out the engine a few times along the way is fine.

6. He has defined his own role as “the person you call when the worst thing happens.”

Don’t contact him with good news; call immediately with bad. He has watched too many founders’ hearts break — running out of cash doesn’t do it, but business pain compounded by two or three personal crises at once will break anyone. And personal problems are the ones with nobody to talk to.

Here is the heaviest passage in the episode: in his twenties he lived through the Coincheck incident, forty billion yen in scale, and he handled the aftermath. Which is why he can now tell a founder that tens of millions, or hundreds of millions, or billions “lack impact — you’re still alive, that’s something.” Only someone with that history gets to say it.

At the worst of it, he was sent to beg executives at large companies for a rescue, meeting in places where they wouldn’t be recognized — and the other side sat there grinning. What they told him was: “You think you’re about to die right now, don’t you? In three years you’ll be laughing your head off.” He says it took everything not to punch them. Three years later, he was laughing his head off.

7. His view is that the problem with a rollercoaster isn’t the ride — it’s that it’s your first one.

Every crisis feels like the end of the world, and then the instant you can see it from outside it becomes “ah, so that’s what this is.” Riding a rollercoaster is awful; picturing yourself riding a rollercoaster makes you realize you’re on one. Founders’ difficulty is that this is their first ride, so they can’t recognize the shape of it. They mistake the lowest point of the process for the destination. His job is to say: in the world of ordeals, this barely counts as the entrance, and in three months something will work out.

8. On Japan’s startup sector, his verdict is: stop imitating Silicon Valley.

The whole category was imported from the Bay Area. Everyone started out infatuated with Google’s free lunches and tried to copy it. But America grew that way out of its own ideology and strategy; Japan needs to rethink its own version from the ground up.

He points to a self-driving company he backed four years ago. Its premise was: we sincerely respect Toyota — as long as humans drive combustion cars, nobody beats them. The only place they could lose is electric vehicles and autonomy, and nobody has covered that. So let’s stake our lives on it. Four years ago, the end-to-end approach was something only a corner of academia had begun discussing, and the number of people doing it in Japan was zero. He rode for thirty minutes on a test track in a car driven by two hundred lines of code, rattling the whole way, and thought: this is the future.

As for the exit, he doesn’t think Japan will swap out its corporate roster the way America did. He thinks it looks more like Europe: large firms acquire new businesses, the nameplate stays, the substance inside was built by someone else. If that’s the future, startups aren’t there to replace incumbents — they’re there to grow the incumbents’ next core business, and to mix talent in along the way. Someone from a startup joins a large company, rises to the executive floor, and naturally reaches for startups as a growth engine. The loop closes. He also notes that this route only gets forced open because activist shareholders are watching closely and a lukewarm acquisition now gets punished.

9. The talent has arrived; the compensation hasn’t.

Twelve years ago, a University of Tokyo engineering professor showed him the placement list: legacy industrial names. Today, starting a company is close to the default in computer science. But he says plainly that the industry is merely “holding these people in trust” — it has not yet delivered returns worthy of them. At Stanford last year, a classmate quoted him a big-tech engineer’s average salary that was absurd, then added, “that’s on the low side, it includes juniors.” His conclusion: trying to out-compete the Bay Area on compensation design is the wrong fight.

Going Further

1. How is “investing before the law allows it” different from gambling?

This is the easiest part of the episode to misread, and the most worth taking apart.

Most people hold something that hasn’t happened yet: the technology isn’t in commercial production, the regulation is still under review, the customer is still piloting. It’s natural to tell yourself “this is exactly the irrational-now, rational-later thing.” The problem is that a gambler can say the identical sentence.

Unpack it and his bet has two layers. The first is the one he explicitly disclaims: he had no idea how the law would move. He never pretended to have judgment there. The second is what he could actually see: this team worked to an absurd degree, they had thought through dock-based design from the start, and they went to the regulators as an industry rather than alone. All three were observable facts at the moment of investment, not forecasts.

So the test is clean: an unpredictable variable can be a condition, never a reason. Only observable behavior can be a reason. “The law will change” is a condition. “This team will still be here on the day it changes” is a reason.

In public markets the same frame applies:

• If your reason for holding is “once the rate cycle ends,” “once the subsidy lands,” “once that big customer signs” — those are all conditions. The day the condition fails, you’ll find you have no second sentence.
• If your reason is “they built capacity before the demand existed,” “they didn’t cut R&D through the two hardest years,” “they went and helped write the standard instead of waiting for it” — those are behaviors that already happened, can be verified, and can be tracked quarter by quarter.

One detail is easy to miss: he invested three or four rounds before the law changed. Staging entries isn’t about getting a cheaper price; it’s that each round gave him another look at whether these people were still there. If you’re holding a thesis that needs years to resolve, don’t fire all your ammunition at once — not because you fear a drawdown, but because you need repeated occasions to re-observe.

2. I checked the financials and studied the industry. Why do I still get it wrong?

Probably because you’re treating cutoff criteria as selection criteria.

His framing is worth copying down: market, technology, talent, and regulation exist to eliminate the ones that genuinely shouldn’t be touched — and most candidates die right there. But past that gate, those same factors lose their discriminating power, because everything still standing has cleared them.

This is exactly where retail investors get stuck. You spend a weekend on gross margin, leverage, market share; everything checks out; you conclude the research is done. But what you finished checking is the entry bar, not the reason to buy. Plenty of companies clear the bar, and their outcomes diverge wildly. The difference lives on the layer above it.

What is that layer? His answer is the accumulation of daily decisions by the operator. You can’t sit in their meeting room, but the idea translates into things you can actually check:

• How management reacts to a bad question on the earnings call. Do they name the problem and give a timeline, or re-tell the good news from a new angle? That carries more information than the guidance itself.
• How they executed their last strategic turn. Did they state plainly where the original read was wrong — or quietly swap out the metric and the disclosure format and hope nobody noticed?
• What they cut during the hard stretch. Cutting marketing and cutting R&D are entirely different acts. What gets cut tells you the real priority ordering.

And his line about small edges getting filled in instantly cuts even deeper in public markets. That thing you found online that “nobody else knows” is almost certainly not an advantage. Edges that survive years tend to be structural, visible, and known to everyone — precisely because the time commitment required is what stops people from copying them.

3. I’m underwater. Do I cut or hold?

First separate two things: did the thesis break, or did your mood break?

That’s what the rollercoaster metaphor is really about. On the ride it’s miserable, but generating an observer’s view of yourself turns the misery into something describable. Founders struggle, he says, because it’s their first ride — they can’t recognize the shape, so they mistake the lowest point of the process for the destination.

Ordinary investors are in almost the same position. At thirty percent down, your brain produces the sentence “this time is different,” and that sentence exists to give your emotions a rational-looking exit. What you need isn’t willpower; it’s a test written before you got on the ride.

He offers one more angle you rarely hear: there is a good way to quit and a bad way to quit. Some people try a bit, hit friction, stop — over and over. Those people will bail halfway even if they draw a fantastic market. So he is always watching how a person ends things.

Translated to us: the way you sell a stock determines the quality of your next one. If your exit is always “it dropped until I couldn’t stand it,” you’ll get off halfway no matter how good the thing you bought — because what made you get off was never the company. It was the drawdown.

Which means the rule shouldn’t be “stop out at minus twenty percent.” It should be “of the three things that made me buy, how many still hold?” The first measures your pain; the second measures the company. Use both if you can. If you can only afford one, take the second.

Worth a Look

• Investor’s Sunday (InterFM), broadcast 2026-08-30, part two with Anri Samata. Part one covers how he entered venture capital; the two together are more complete.
• LUUP’s dock-based service design and the timing of Japan’s road traffic law revision are both public record. Checking the sequence yourself conveys the length of that wait better than any summary can.
• If you want to extend the idea, the two-stage structure — eliminate on cutoff criteria first, then judge on the decisive one — has a large existing literature in both venture investing and hiring.

The One Thing to Take Away

In your reasons for holding, how much is condition and how much is behavior?

That’s the measuring stick this episode hands you. A condition hasn’t happened, you don’t control it, and you can’t forecast it: the law will pass, the policy will come, the market will turn. A behavior has already happened, can be verified, and can be tracked: these people kept doing this when there was no reward for doing it.

Bet on conditions and you’ll be speechless the day the condition fails, because you never had a second sentence. Bet on behavior and at least when you’re wrong, you’ll know where you misread.

Something to do today (it doesn’t have to involve investing):

Take anything you’re currently “waiting on a result” for — a promotion, a relationship improving, a child clicking into gear, a proposal getting approved. On a sheet of paper, draw two columns.

Left column: what condition am I waiting for — the thing outside your control that you can only wait to happen. Right column: what have I actually done in the past three months, specific enough that someone else could have seen it.

Then look at the ratio. If the left column is full and the right has one or two lines, you aren’t waiting on a result; you’re waiting on luck. If the right column is full, then whenever the left-column event arrives — or even if it never does — you’ve already been moving.

Rewrite the sheet every three months. The change in the right column’s length is the ruler.

Raw Markdown

This article is an educational discussion of investment method. It is not advice to buy or sell any individual security, offers no target prices, and does not analyze any current holding. Investing carries risk; make your own decisions or consult a qualified professional.