《一個70後睇高端手機殼呢門生意》A 70s Kid Looks at the High-End Phone Case Business

Last week I stood outside a shop in Mong Kok, staring at a phone case priced a bit over two hundred Hong Kong dollars.

Not because the case was beautiful. Because I know how to make this thing.

I entered the trade in 1998. I've handled more products than I can count: electronic parts, transformers, plastic housings, radios, remote-control cars, Christmas lights, senior phones, watch phones, smart bands. Phone cases — I never did one big order on them. But I've done all their neighbours: the same mould shops, the same injection machines, the same Huaqiangbei stalls, the same containers to Europe.

So when I looked at that two-hundred-dollar case, the sentence in my head was: "This thing costs less than twenty bucks to make."

I'm not trying to expose anyone. In business, if a customer is willing to pay, that's your skill. What I wanted to understand was one thing: those two hundred dollars in the middle — what exactly did they buy?

I've been asking myself that for a few weeks, and I went through a pile of public research. What follows is what I found, and what I think about it. Some parts I still haven't figured out. I'll say so.

1. Three Brands, Three Different Stories

To do this business, the first thing you have to work out is: at the top of the market, what are the highest-priced players actually selling?

I looked at the three most representative: CASETiFY, born in Hong Kong; Mous, from the UK; and Shenzhen's PITAKA. Three completely different price points, but one thing in common — none of them sells a case. They sell a story the customer can go and check.

CASETiFY: collaborations and community

It started in Hong Kong in 2011 as an app called Casetagram — "send us your photo, we print it" — at $35 a piece. The first spike came when British chef Jamie Oliver posted one. In 2014 it renamed itself CASETiFY and shifted from "customisation service" to "streetwear brand".

The numbers: about US$120 million revenue in 2021, three million units sold. By 2024, media reports put revenue at RMB 3.6 billion (that's a media figure, unaudited — I'll say that plainly). Its stated target for 2025 was US$3 billion, roughly RMB 21.5 billion.

Its own site claims: 30,000+ designs, 300,000+ SKUs, more than 100 licensed IPs. The price ladder looks like this:

What struck me wasn't the price. It was the cadence.

It doesn't bet on one hit. It uses the frequency of new releases to beat the odds. Disney, Pokémon, Marvel, One Piece, Coca-Cola, Chiikawa, DIESEL, Murakami — these aren't marketing events. They're a production scheduling tool. The IP calendar drives the release rhythm, instead of the brand waiting for demand.

Then there's the customiser (CASETiFY Studio): pick the shell type, colour, layout, add text, add a photo. That has two layers of value. One, the SKU the user generates needs no inventory. Two, it's one of a kind, so it can't be price-compared.

I'll come back to that point, because it's the core of my own thinking.

Mous: engineering proof

Mous, from the UK, founded 2014. Its first product was Musicase — an iPhone case with cord storage — which raised £12,000 on Kickstarter. Then it hit production-quality problems, and Apple killed the headphone jack, wiping out the whole selling point overnight. The team borrowed £150,000 and moved into a factory to rebuild.

In 2017, Limitless raised close to US$2.5 million from 51,000 backers on Indiegogo — the top phone-accessory project that year.

FY2026 (media figures, unaudited): revenue £36.5 million, about RMB 334 million, up 23% year on year; EBITDA £4.3 million, up 44% — its best year since founding.

But its real skill isn't the numbers. It's a public experiment that repeats every single year.

It developed its own cushioning material called AiroShock. The explanation given is that the material is full of microscopic air bubbles; on impact they compress and absorb and spread the energy. That reframes the industry default of "thicker equals tougher" as "different material equals tougher", which directly holds up its pricing.

Then it uses a series of extreme stunts to notarise that material: a 45-foot crane drop onto concrete, a throw off Waterloo Bridge, the founder throwing a brand-new iPhone outside an Apple Store on launch day in 2017, bungee jumps, helicopters, hot-air balloons, 1,200 feet, even "dropped from the edge of space". In 2026 even a suitcase got a 250-metre drop and got dragged behind a race car.

And the prices? Mostly on the UK site: the thinnest Super Thin is £44.99 for iPhone 18 Pro Max; the clear Clarity is £54.99; the flagship Limitless 6.0 in aramid fibre is £54.99 for iPhone 16 Pro, walnut is £64.99, and the US site lists aramid fibre for iPhone 18 Pro Max at $79.99. Bundles (Upgrade Kit / Complete Kit) run four tiers, £62.99 to £112.99.

Look at that — no IP premium at all. Only structure, material and function. Within one series, the material (fabric → walnut → aramid fibre) and the add-ons (camera shield, stand, card wallet) create a £10 to £60 spread. Very clean structure.

PITAKA: material science

Shenzhen's PITAKA was founded in late 2015 by Zheng Yanghui with a few designers and materials specialists; the parent is Shenzhen Lingyi Innovation Technology. The founder was previously a rock fan and partner in an indie label — no accessory-industry background. On the company's front desk are eight characters: "Rather be an outlier than follow the crowd."

The numbers: annual revenue over RMB 300 million (2025 reporting, relaying the brand's own figure), annual volume at or past one million units, cumulative users reported at 1.65 million and now 2 million+ on its own site, selling into 120-plus countries. 137 patents. 200+ inventions. Tens of millions of RMB in R&D each year.

The material it sells is aramid fibre — Kevlar. The official language is "aerospace-grade aramid fibre", "five times stronger than steel, one fifth the weight". The most powerful part of that isn't the spec. It's the analogy: aramid fibre is what fighter jets, armoured vehicles and body armour are made of. Using that material for a phone case is a story in itself.

Prices: core cases run $49.99 to $79.99 (roughly RMB 298–499 in China), five to ten times a normal case. Accessories are deliberately held at $20–60 — magnetic grip $19.99, magnetic card wallet $34.99, magnetic power bank $59.99, car mount $59.99. That pricing isn't there to make money on accessories. It's there to push "case + wallet + grip" into a $130+ basket.

But there's one number that made it, for me, the most interesting company of the three.

A third-party e-commerce intelligence tool, Grips Intelligence, estimated in August 2026 (a sampled estimate, not an official figure — I'll say that plainly): ipitaka.com saw roughly 1.2 million monthly sessions, about 25 times a typical consumer-electronics store (~48,100); average order value was a high $150–175; but monthly transactions were only 1,868 — a conversion rate of 0.0% to 0.5%. Its direct competitor Rokform converts at 2.5% to 3.0%. The same estimate put its 2025 direct-site GMV at about US$3.5 million, down more than 50% year on year.

How do I read that?

My reading: traffic isn't the problem — the ability to close is. The material narrative successfully pulled in 1.2 million people, and only about a thousand of them paid. Where did the other 1.19 million go? They looked, nodded, and bought elsewhere.

That means two things to me.

First, telling a story isn't the same as collecting money. The story brings people in; something else turns them into revenue — a sensible price, a simple decision, no hesitation, someone willing to wait. If I'm doing handmade work and asking customers to wait three weeks, I have to solve "why is it worth waiting" even harder, or my conversion will be worse than 0.5%.

Second, my own numbers will be far smaller. I won't expect 1.2 million monthly sessions. If I get a few thousand sessions a month at 2% conversion, I'd call that a healthy business. Those are two completely different objective functions — one chases scale, the other chases density. I want the second.

The cleverest thing it does is name its materials. Fusion Weaving, Mammoth Resin, PitaTap, Lumintex, Cadmyra, SlimBoard. All trademarked, all with their own explainer pages. After doing this research I wrote one line down: a named process can be told over and over; an unnamed spec can only be told once.

What the three have in common

After reading all three, I picked up a pen and wrote one line:

A high-end phone case was never selling a case. CASETiFY sells collaborations and community, Mous sells engineering proof, PITAKA sells material science. The premium doesn't come from the plastic — it comes from a story the customer can check.

Those two words — "can check" — matter enormously.

If you say your case survives a drop, the customer can drop it once and see. If you say you use aramid fibre, they can go look up aramid fibre. If you say you collaborated with Murakami, they can go and verify it. All of that is verifiable. The plastic itself is not.

That's the first truth I found: the source of the premium is verifiability, not quality.

Put it more bluntly. Two cases might differ by two dollars in cost. One has a story and one doesn't, and the price difference can be two hundred. That remaining one hundred and ninety-eight dollars buys "I know what I bought, and I can prove what I bought."

But that story has a fragile spot, and I have to name it.

In April 2024, the UK Advertising Standards Authority ruled against a Mous TV ad. The ad showed 50 people throwing phones in Mous cases into the air. The ruling noted that apart from three paid actors, the participants were mostly Mous employees, friends and family, so it did not constitute an independent test; the phones weren't dismantled to check for internal damage; and the internal drop report the brand supplied was 26 drops from 120cm — nowhere near the heights shown in the ad. The ASA found the ad exaggerated and misled, breaching BCAP 3.1, 3.9 and 3.12, and barred it from being broadcast again in that form.

That's an expensive lesson for anyone building a brand: the bigger the story, the more it has to hold. If you claim 250 metres, someone has to believe you really threw it 250 metres. When it doesn't hold, everything you told before devalues at once.

2. Ninety Percent Compete on Price, and Die There

Those three are the top of the pyramid. Now let me talk about below.

How big is this industry? The global phone case market was about US$16.31 billion in 2025, about US$17.26 billion in 2026, growing at a 5.8% CAGR, projected to reach US$21.6 billion by 2030 (The Business Research Company). Asia-Pacific was the largest region in 2025. Another report (Bozhesi Consulting) puts 2028 at RMB 129 billion globally at a 4.36% CAGR — different methodology, so don't mix the two growth rates.

The pie is not small. So a lot of people rush in.

But in the same pile of public material there's one number that made me sit down:

Data from 1688 shows that among standardised goods like phone cases and cables, more than 68% of products share common moulds, and the industry's average gross margin has fallen below 15%.

Sixty-eight percent on shared moulds. Margins below fifteen percent.

I've done twenty-six years of foreign trade. I know what 15% gross margin means. It means that after freight, FX, returns, interest on payment terms and the customer squeezing you, what's left in your hand may be negative. It means you can't hire a good designer, because one designer's salary eats a whole shipment's profit. It means you can't use good material, because good material costs more and the customer won't pay extra.

And then the death loop: no money to do good work → compete on price only → even less money to do good work.

I've watched this loop happen in many industries. Christmas lights, USB fans, Bluetooth earphones — all walked the same road. At the start everyone makes money, then someone cuts price to win an order, then everyone cuts, then the whole trade drops to single digits, and in the end only a few scale players and a few brand players are left.

This industry also has a very cruel feature: brutal seasonality and an extremely short window.

A few numbers:

• In the first week after the iPhone 18 Pro / iPhone Duo launch, accessory searches on Tmall jumped 600%, and week-on-week phone case transaction value rose more than 300%;
• On AliExpress, searches for iPhone 18 accessories were up 182% in August versus July;
• By launch day, AliExpress had added over 300,000 new listings in just phone cases and screen protectors;
• On TikTok Shop, posts under #iPhoneduo passed 42,900.

Put those four together and the picture is clear: demand explodes within one or two weeks — and 300,000 new listings charge in at the same time.

If you're not fast enough, you miss it. If you're fast but undifferentiated, you can only cut price. You cut price, you become one of the 68%, your margin drops below 15%, and you're the group that "competed on price and died there".

Interestingly, the same German Google Trends data shows a reverse pattern: when the index for "iPhone" (phone replacement) fell from a 12-month average of 48.73 to 42.0 (0.86 times the annual average), the accessory term "Handyhülle" (phone case) sat at 1.0 — higher than its December average of 0.88 (a ratio of 1.13), with 30-day momentum positive.

That says one thing: the urge to replace your phone fades, but the urge to make the phone you already own nicer does not. The same data also shows phone-case searchers skew female (male-to-female 0.80) while iPhone searchers skew male (1.02), both with a core age band of 35–44.

That finding is useful to me, because it means this business doesn't have to be bolted to launch day. New phones are quick money; the installed base is the long tail. And what I intend to build is, at heart, a long-tail business.

3. My Read: A Different Way to Build It

What follows is my own thinking, not market consensus. I'll try to be candid, including about what I'm unsure of.

If the premium comes from a verifiable story, the next question is: can I tell a different story?

The three brands above tell stories about collaborations, engineering, and materials. All three are "mine is better" stories — better IP, harsher tests, higher-grade material.

What's in my head is a fourth kind: hand craft.

(1) Material is the product, not the cost

The longest-standing habit I have from foreign trade is reading cost sheets. On an injection-moulded case, material might be under 5% of the retail price, so most manufacturers work backwards: set the price first, then figure out what material fits.

What I want to do is invert it: set the material first, then the product.

In the short term that's stupid. Good material raises cost, and customers may not pay more immediately. But PITAKA proves a material narrative can hold a five-to-ten-times price gap. It made the world's first magnetic aramid-fibre phone case in 2016 and built the MagEZ magnetic line in 2017 — three years before Apple shipped MagSafe. What a first mover gets is the naming right. Win the naming right and you're no longer "a case maker"; you're "the one who sets the standard".

HANDSOME works in three materials today: aramid fibre, full-grain leather and engineered drop protection. Aramid fibre is the road PITAKA has already run furthest down; I'm not going to fight them on their own ground. What I have in mind is a different set: full-grain leather, wood, bamboo, recycled fibre, natural rubber. What they share is this — every piece is different, and they age.

Aging is a very important property. A piece of plastic is still the same piece of plastic in ten years. A piece of leather has, after a year, a patina you raised yourself. That's time telling the story for you, so you don't have to.

(2) One at a time

This is the point I'm most confident about and most nervous about.

The nervous part is obvious: if you make one at a time, how do you ever reach volume?

I thought about that for several days. My answer has three layers.

Layer one: what if I don't need "volume"?

As I said, 1688 data shows 68% of products share common moulds and average margins are below 15%. In that structure, volume is a trap — the more you make, the more you have to compete on price; the more you cut price, the less you can afford to do good work. Scale is a shield, but it's also a shackle.

Handmade orders are a different economy: the customer orders, then waits. While they wait, you don't hold inventory, you don't tie up capital, you don't worry about unsold stock. They're willing to wait because they want the one-of-a-kind. It's the same logic as CASETiFY's customiser — the SKU the user generates cannot be price-compared.

Layer two: I can reach volume "one after another", just not "one hundred thousand at once".

That's the nature of hand craft. One craftsman makes five a day — twelve hundred a year. A hundred craftsmen make one hundred and twenty thousand a year. Against CASETiFY selling 15 million units a year (2022 figure), that's a rounding error.

But flip it: 120,000 units at a good price, versus 15 million units at a price squeezed to nothing — which one actually makes money? I can't give you an exact answer, because I don't have their cost sheets. But I've done too many thin-margin, high-volume orders to believe the answer is usually "the bigger one".

Layer three: maybe I genuinely don't need to reach "volume". And maybe that is the point.

If the goal is "become a big company", then the handmade model is a dead end. But if the goal is "a small business that supports a group of craftspeople for thirty years", then scale isn't the objective — it's a side effect.

I'm in my fifties. I don't want to build another factory that survives on cutting prices. I want to build something that's still running when I've stepped back. What that needs might not be volume. It might be time.

(3) One studio per country

This is my layout idea. Say it out loud and most people's first reaction is "you're dreaming".

Roughly, it goes like this:

• Raw materials supplied from Hong Kong. Not produced in Hong Kong — sourced, inspected, consolidated and settled there.
• Roughly one studio per country. Europe, the US, Southeast Asia, Africa — landing one place at a time.
• Craftsmen teach apprentices, hand to hand. Not paid classes — learning inside the studio while working.
• One at a time, small batch, made to order.
• Light, thin, good-looking. Not heavy armoured drop-proof boxes.
• Build our own brand, and run OEM/ODM studios for others. The brand builds reputation; the contract work builds cash flow.
• Funding support to get local studios started. Not a loan — an upfront investment.

The real moat of this model is not technique.

I need to be very clear about that. There's nothing earth-shattering in this craft. Teach an apprentice for three months and a bright one has the basics. Technique was never a moat.

The real moat is the cost of trust.

Open a studio in Thailand, and the problem isn't "how do we make it". The problems are:

• Which local craftsman will trust you enough to teach your people?
• Which local customer will trust you enough to wait three weeks?
• Will the local material supplier give you 30 days' credit?
• Can you handle the landlord, the tax office, the customs people?
• What happens to you when the exchange rate jumps?

Every one of those is a relationship problem, not a technical one.

And the sum of those five questions is what I mean by "the cost of trust". I want to be precise about that term, because it's the core idea of this whole piece.

Picture opening a studio in a new market. Day one, you've hired three craftsmen. You don't know them; they don't know you. You place an order and they deliver — will you inspect every piece? Yes. They quote a price — will you shop it around? Yes. They say they're short on material — will you suspect they pocketed some? Yes.

All of that is cost. And all of that cost exists because there's no trust.

When trust is low, quality is an expensive thing to produce: you inspect, you spot-check, you write contracts, you withhold payment, you station someone to watch, you hire a third party to audit the factory. Every step is defensive.

When trust is high, the cost of the same work collapses: you don't inspect every time, because they know that you know they won't cheat you. That gap is what I mean by the cost of trust — it isn't a moral term. It's a real line item. It can be measured, and it can be saved.

And relationships are the one thing twenty-six years of foreign trade has actually taught me.

I started in 1998 at a trading company in Guangzhou. In 2009, the German client KTL stationed a man in our Guangzhou office for two to three months, all for a quad-band senior-phone project. In the early 2000s in Xiaobei, an African client's phones were stolen at her hotel, and I took the loss and compensated her — later, her family produced a president's relative in an African country. In 2008, the New York designer Stephanie flew to Guangzhou and asked me to set 400 zircons on a watch phone. In 2005, a foreign client brought me a watch-phone sample and offered 5,000 Hong Kong dollars to make it real.

Not one of those orders was won on technique. Every one was won on three words: "I trust you".

So "one studio per country" sounds like a supply-chain design. It's actually a design for a network of trust. And building networks of trust is the only thing I genuinely know how to do.

(4) Why Hong Kong supplies the material

I first assumed this was a logistics question. Then I realised it isn't.

In July 2026, Hong Kong's merchandise exports rose 50.7% year on year, with AI-related electronics again posting a notable increase; total goods exports for January–July were up 40.9%. I wrote a piece about that at the time ("Hong Kong Exports Up 50.7% — How Our Generation of Traders Keeps Up"), and my reading then was: the chain has a new engine, but the engine is still Hong Kong.

Looking at phone cases now, I see one more layer.

Hong Kong is not a manufacturing origin. Rents and wages there rule out production. But Hong Kong has something Shenzhen, Guangzhou, Bangkok and Lagos don't:

Hong Kong is a credit intermediary.

What does that mean?

Suppose I open a studio in Kenya and need a batch of full-grain leather. The leather might come from Italy, or from Vietnam. If the Kenyan studio buys direct from an Italian tannery, the Italian asks: who are you? How do I know you'll pay? If you're three months late, how do I chase you?

But if that batch moves through Hong Kong — sourced there, inspected there, consolidated there, settled there with a letter of credit from a Hong Kong bank — then what the Italian sees is a Hong Kong company that knows how to do this trade, knows how to read a bill of lading, knows how to open an L/C.

Hong Kong isn't selling goods. It's selling "I guarantee this deal will complete."

In my trading years, the three essentials were the letter of credit, the bill of lading and the invoice. I can still rattle off the abbreviations. And Hong Kong's density of expertise in those three things is still world-class in 2026.

So "materials supplied from Hong Kong" isn't about saving freight. It's about standardising trust. A small studio in Kenya cannot build an international credit system on its own. But it can plug into the one Hong Kong already has.

Which materials, concretely? My list: full-grain leather from Italy and Vietnam, natural rubber and rattan from Southeast Asia, hardwoods from Africa, recycled fibre and bamboo from the mainland. They share one property — none of them has a standard price. The same batch of hide, depending on how you grade it, cut it and work around the flaws, can yield twenty or thirty percent more or less usable material. A business like that doesn't need cheap purchasing. It needs a middleman who can look at goods and know. And looking at goods and knowing is what I've done for twenty-six years.

So Hong Kong's job in this model is four things: sourcing, inspection, consolidation, settlement. None of them needs a factory floor. All of them need credit and expertise — two things Hong Kong still has in 2026 and won't lose quickly.

That, as I see it, is Hong Kong's position in 2026: not the factory — the issuing bank.

(5) Teaching the craft in Africa and Southeast Asia

I need to tread carefully here.

Because it's very easy to frame this as charity, or as ESG marketing. I want neither.

What I mean is something very practical: if you want to do business in a place for thirty years, you have to leave behind a group of people there who became more capable because of you.

Not because you're noble. Because without it, thirty years later you're still an outsider, your costs are always higher than a local's, you can never hire, and your goods always have to fly in and out.

In 2011 I wrote a piece called "My Friend Fank". Fank is Ugandan. He came looking for work during Guangzhou's busiest trade years, 2008 and 2009. He wasn't cut out for trade — too honest, couldn't haggle, couldn't bluff. But he was a good teacher. Later he became a professor at a Guangzhou university, and his family settled down.

That taught me something: some people aren't your business partners, but they are your local roots.

So with "craftsmen teach apprentices, hand to hand", what I see is:

• For local people: a skill, which is an asset nobody can take away.
• For my business: a group of people who can do the work, will do the work, and can be trusted.
• For the brand: every craftsman in every studio is part of the story — and it's a true story.

Those three point the same way. I won't pretend I'm doing it to help people — I'm doing it to do business. But on this particular thing, doing business and helping people happen to point the same way. So I don't have to choose.

4. The Hardest Part Isn't the Business — It's Choosing People

I often tell people that business isn't hard.

Most people hear that as arrogance. But I mean it. Products, moulds, supply chains, customs, letters of credit, exchange rates — I know those. Learning them takes ten years, but once you've learned them they stop changing. All of it is learnable.

Choosing people is different. Choosing people is the thing you never finish learning, and the thing you have to pay for when you get it wrong.

What twenty-six years of trade actually left me with

I entered the trade in 1998, at a trading company in Guangzhou, and I'm still here.

Over those twenty-six years my clients have spanned Europe, the Middle East, Africa, Southeast Asia and South America. Some I've worked with for over a decade. Some I know well enough to know when their kid is getting married. In 2009, Frank at Germany's KTL stationed a man in our Guangzhou office for two to three months, all for one quad-band senior-phone project. Just before Mid-Autumn 2026, after his son Lukas took over the company, they ordered a hundred boxes of mooncakes from us. Frank is 72 this year.

What are those relationships?

Most people call it a network, a resource, the key I can use to open markets. That's not wrong, but it isn't precise either.

The precise version is this: I know which person is honest when there is no contract on the table.

That sounds simple. It is the most expensive thing twenty-six years of business has given me.

Because "no contract on the table" is the most dangerous and the most truthful state in business. Once a contract is signed, everyone is honest — dishonesty costs money. But before that? In a verbal promise, a voice message, a quick "yeah, I'll sort it" — how do you know the person on the other end will stand by it?

I did one deal in the early 2000s, in Xiaobei, Guangzhou. An African client's phones were stolen at her hotel. Strictly speaking it wasn't my problem — the goods were delivered, the paperwork signed. I compensated her anyway. Plenty of people told me I was a fool. Later, her family produced a president's relative in an African country. I've written about that story. But what I want to say today isn't "good deeds come back". It's this: I paid because I judged that person was worth it — not because I'd calculated a return.

That kind of judgement, built over twenty-six years, has given me samples from a dozen-plus countries.

So when I say "one studio per country", people hear a supply-chain plan. What I'm actually talking about is a shortlist of names. I don't need to start from zero finding people in Bangkok or Nairobi. I open my phone book and I already know which clients are worth a plane ticket and a sit-down.

This plan isn't a daydream. It already has a concrete landing place: HANDSOME, the bespoke-luxury brand I invested in, now selling direct overseas. And my role isn't the one on the workshop floor. I'm the one who recognises people, who can back them with credit, and who is still there when something goes wrong.

That is my real asset. Not the brand, not the product — the list.

How I judge a person

I don't know how to use psychometrics, I don't read CVs, I don't run background checks. My method is crude. I've used it for twenty-six years.

One: watch what they do on a small order.

This is the one I trust most. Nobody sells their reputation for a five-thousand-dollar order. Which is exactly why a small order shows you who someone really is. Will they swap in worse material to make an extra two hundred bucks? Will they deliver three days late and not tell you? Will they say yes to something they can't do?

A person's choices on a small order are their choices on a big one — it's just that the consequences on a small order are small enough that they don't bother to hide.

Two: do they volunteer bad news?

This is the signal I weight most heavily — heavily enough that I'll walk away from a deal over it.

In business, everyone is willing to deliver good news. Bad news is the dividing line. One person finds a problem and calls you straight away: "Hany, we've got a problem, here's what I've done." Another sits on it for three weeks and only admits it when you chase. Those are two kinds of people, two kinds of company, two endings.

In all my years of trade, nine out of ten blow-ups weren't caused by the problem itself. They were caused by the problem being hidden. The problem itself is usually fixable. A problem that's been buried for three weeks is not.

So when I hire now, I don't look at how brilliant someone is. I look at how fast they tell me the first time they get something wrong.

Three: when do they pay?

This sounds mercenary. It's the most honest test of the three.

Timing of payment isn't about money. It's about where you sit on their list. A client who is always punctual with their big suppliers and always three months late with their small ones — you've just learned how they see people. How they treat a supplier smaller than themselves today is how they will treat me tomorrow.

I've been a supplier and I've been a buyer. Having been both is what taught me: people who pay on time aren't doing it because they have money. They're doing it because they treat you as a real thing.

These three tests sound simple, even crude. But I'll tell you: twenty-six years, and they've been enough. Because they don't measure capability. They measure character. Capability can be learned. Character can't.

Young people run it, I stand behind

This is the part of the plan I'm most confident about and have said out loud the least.

I'm in my fifties. The times have changed. I'm not going to live in Africa for three years again, or chase suppliers in Bangkok, or learn how TikTok works. So my role is not the operator's role.

The operating is for young people. Every studio has one young person on the ground: they find the space, hire, chase orders, talk to local customers, handle shipping and after-sales. They are the boss.

My role is behind: putting up the capital, supplying the first batch of material, lending judgement, lending the human network, and lending a promise that says "when it goes wrong, I'm still here".

I want to be precise about that word "behind". It isn't "boss", and it isn't "investor". It's an older role than either — guarantor.

What do I owe these young people? I've thought about it a lot. Not a salary. Four things:

• A backer who won't vanish halfway. When it goes wrong, they don't have to carry it alone.
• A material line that won't break. They don't have to sort out Italian leather and letters of credit themselves. That's my job.
• A profit split that's written down. Don't wait until there's money to discuss it, because by then you'll never agree.
• A way out. If they want to leave, they can leave cleanly, without it turning into a falling-out with me.

The fourth is the one nobody talks about, and I think it matters most. Because if a person knows they can't leave, they won't go all in. Tied-down people don't make good work — I said that earlier and it holds here too.

And what I get out of this structure isn't a salary. It's somewhere to finally use a judgement I spent twenty-six years building. That is the only thing I'm genuinely interested in doing now.

The one thing I'm most afraid of

Everything above is the good side. Now the bad side, because it's what I think about most.

Choose the wrong person and you lose two things.

The first is money. I'm used to that. Twenty-six years of business, I've lost plenty. Money can be earned again.

The second is a friendship. That one I'm not used to.

This is the most insidious part of the model: my candidates are precisely my friends.

The dozen or so names on my list are all clients I've worked with for over a decade and trust. Asking them, or their children, to open a studio is the most natural move in the world.

But what if they can't make it work? What if the money I gave them genuinely doesn't work out?

At that point I can't chase them, because what I said was "investment", not "loan". And I can't pretend nothing happened, because I keep score inside.

The worst case isn't losing the money. The worst case is this: next time we meet, we're not friends any more. We're an unsettled account.

In all my years of trade I've seen this many times. Two families at dinner, perfectly polite, talking about the weather and the kids — but there's an open account on the table, everyone knows it, and nobody mentions it. Then everyone goes home, and you stop arranging dinner.

That's what frightens me. And it isn't a risk-management problem that a contract can solve. Because if you need a contract for it, you shouldn't have gone to a friend in the first place.

The only rule I've come up with is an unappealing one: treat money you lend as money you gave away. If they repay it, I'm delighted. If they don't, I haven't lost a friend, because I never treated them as a debtor.

The cost of that rule is that I have to control the scale from the start. I can't "invest" the whole of a studio. I need them to have their own skin in the game — not because I want them to suffer, but because someone with nothing at stake won't push.

It's a hard balance to strike. I haven't struck it yet.

5. How Do Other Countries Run This Business? I Went and Studied It

I didn't want to make this decision on gut feel. So for the past few weeks I went and studied how different places run "hand craft" as a business — what they do well, what it costs them, and what I can steal.

I looked at five: Italy, France, Germany, Japan and Korea — plus China itself. Here's what I found. Some figures are official or industry-report numbers, some are press figures. I'll say which where I can.

Italy: one town, one trade

Italy has a concept called distretti industriali — industrial districts. One town does one thing, and does it so well the whole world has to buy from there.

The best-known example is eyewear. In the Belluno and Cadore area there are 814 eyewear companies and 19,000 employees (2024 figures). According to Italian official material, that district produces about 70% of the country's eyewear. In the first half of 2025, Italian eyewear exports were €2.8 billion, down 3.7% on an exceptional 2024; around 90% of production is exported. Italy is Europe's largest producer of frames and sunglasses, and the world's largest producer and exporter in the luxury tier. The district's story starts in 1878 in Calalzo di Cadore.

What the model does best is this: the whole chain lives in one town. Big groups and small and mid-sized firms coexist; the mould-makers, the hardware people, the surface-treatment shops are all specialists, each doing one small segment extremely well. You don't need to run your own plating line to make eyewear. You walk two streets and someone does it — and their family has done it for three generations.

And the cost?

First, it's dangerously concentrated. One town doing one thing means when global demand dips, the whole town dips together. The numbers I saw show exactly that: exports down 3.7% in H1 2025, sunglasses down 5.5% — while Europe grew 8%, China 29.5% and Mexico 47%. You survive single-region swings only by spreading your markets.

Second, being famous isn't the same as being profitable. An eyewear district sells the "Made in Italy" label, and the biggest beneficiary of that label is usually the brand — not the small workshops doing the machining.

What I'll steal: not "open a big factory", but "one place, a cluster of specialists who supply each other". A studio shouldn't try to do every process itself. It should sit among a group of related craftspeople.

France: five years on the road

France has an organisation called the Compagnons du Devoir, founded in 1941. Its system is the closest existing version of what I mean by "craftsmen teaching apprentices, hand to hand".

How it works: apprentices join between 15 and 25, with around 30 trades to choose from — construction (tiling, carpentry, stone carving), soft materials (leatherwork, shoemaking, upholstery), industry (boilermaking, mechanics), the trades of taste (baking, pastry, charcuterie, cheese), plus farriery and barrel-making.

Then comes the crucial step: the Tour de France. After their first qualification, a young person goes to work under masters in different cities for an average of five years, required to change city once a year, and required to spend a stage abroad — in a non-French-speaking country, with 65 countries offered. There are around 90 Maisons (houses) across France, training close to 15,000 people a year.

The cleverest thing about it: it doesn't teach you technique, it forces you to change environment. What you learned from this master this year may be wrong in another city next year. After five years you don't know one way of doing things. You know why things are done that way.

The cost? Obvious. Slow, and expensive. Five years on the road isn't for everyone. And the system only stands up because of a large association, sponsorship and government support — in 2024 the French apprenticeship subsidy to employers was around €6,000 per person, and cuts are now under discussion.

What I'll steal: I can't run a five-year tour. But I can steal a smaller version — every master in every studio must have spent at least three months working in a studio somewhere else. It isn't expensive, and it turns "standards" into something shared instead of everyone doing their own thing.

Germany: the master's certificate isn't an honour, it's a gate

The German Handwerk system is the one I admire most.

The path: you enter the Dual System at 15–18, spending a few days a week doing practical work in a workshop and the rest at vocational school, for about three years. Pass the journeyman exam (Gesellenprüfung) and you're a Geselle. To go higher you work as a journeyman for several years, then attend a master school (Meisterschule) for one to two years and sit the master exam (Meisterprüfung).

The key point: the master's certificate isn't an honour. It's a gate. In 2020 Germany amended its crafts code and restored the master requirement (Rückvermeisterung) in 12 trades — without a master's certificate you cannot open a business or take on apprentices. Today 53 trades work that way. The master is simultaneously educator, business owner and technical authority, all bound into one certificate.

How big is the system? Between July 2023 and June 2024, German Handwerk had roughly 2.6 million qualified workers: about 2.3 million journeymen and around 167,000 masters. At the same time the sector had around 107,000 unfilled vacancies. Over the decade from 2013 to 2023/24 the number of journeymen fell while the number of masters rose — because the rules changed and because employers began funding staff through the master's qualification.

The cost? Slow, and bureaucratic. Becoming a master can take close to ten years. And tightening the master requirement always draws accusations of protectionism and keeping people out — the 2020 change is still argued about.

What I'll steal: one very specific idea. I don't need the most skilled master. I need a master who can teach. Germany taught me that if teaching isn't part of what a master is, masters won't teach — because teaching an apprentice dilutes your own position. So my thinking now is: a studio master's income should be tied to how many apprentices they've brought to the point of being able to make goods independently. That's the only design change this research actually made me commit to.

Japan: Tsubame, saved by design

Tsubame, in Niigata Prefecture, has been working metal for 400 years.

One figure stuck with me: over 95% of the spoons and forks made in Japan are made in Tsubame. It used to be in social studies textbooks. Today it isn't just cutlery — pots, kettles, frying pans, strainers, knives, even microwaves, fridges and coffee makers. The world's first IH-compatible commercial cooking pot was developed there in 1985 by a Tsubame stainless-steel processor.

But what's worth learning isn't the "400 years". It's what they did in the last 30.

Tsubame-Sanjo faced what every old industrial district faces: high labour costs, orders moving to China and Southeast Asia, contract manufacturing getting thinner every year. Their answer wasn't cutting prices. It was changing the use. They moved from mass-market cutlery to design-led household goods: the same precision sheet-metal skills applied to a beautifully designed waste bin, a one-slice hot sandwich maker, a coffee machine supervised by a legend of the coffee world. Their own municipal site's tagline is literally "design is functionality".

The result is that "Made in Tsubame" became a brand, not just a place of origin.

The cost? The shift took decades, and a lot of factories died in between. And in the end they still depend partly on orders from large manufacturers — being a supplier to an appliance brand and being a brand yourself are two different fates.

What I'll steal: stop telling people "we make phone cases". Tell them "we use a certain craft to make a thing you carry". Tsubame's craft was metalwork; mine can be leather, weaving or wood. The same pair of hands can make many kinds of product. That's the only way out for an old craft.

Korea: cosmetics ODM — the exact opposite road

This is the one that made me sit down.

Korea's cosmetics dominance over the past decade wasn't the brands' achievement. It was contract manufacturing. Cosmax is one of the world's largest cosmetics ODMs: 2024 consolidated revenue of KRW 2.1661 trillion, up 21.9% year on year; operating profit KRW 175.4 billion, up 51.6%; net profit KRW 88.4 billion, up 133.9%. Cosmecca runs the same play.

What do they do? They are a plug-and-play back end for small brands. Want to launch a skincare brand? You don't need a factory, you don't need to hire chemists, you don't need to sort out regulation. You bring an idea and a packaging design to Cosmax and they handle formulation, sampling, testing, regulatory paperwork, filling and packing. You can order a thousand units.

How strong is the model? Korea's indie brands — many of the names you've heard of — used this back end to go from zero to global shipping within a year or two. Cosmax's own growth was driven by those small clients, and its Southeast Asian entities (Indonesia, Thailand) grew even faster than its Korean base.

This is the exact opposite of my plan.

Mine is one studio per country — expensive, slow, dispersed, each one needing people.

Theirs is one back end serving a thousand brands — cheap, fast, centralised, scalable.

I asked myself seriously: have I got this wrong?

The Korean model has costs too, and they aren't small: the brands never learn to make products, the factory never touches the consumer, and the back end itself never becomes a brand. And when every brand uses the same back end, their products converge — which is a real problem Korean indie brands face today. But none of that changes the fact that it makes money.

So I won't say the Korean model is bad. I'll only say: it's a different business.

China: the part I'm most qualified to talk about

Finally, home. I'm from Guangdong, I've done twenty-six years of trade, and I watched this road get built.

Chinese contract manufacturing starts with 1978's "three supplies and one compensation" model. In 1981, processing trade was just 6% of China's total imports and exports. By 1998 it was 53.4%. By 2008 processing trade volume had passed US$1.05 trillion.

What did Chinese factories actually learn in those thirty years? One detail. A manager at a Shaoxing home-textiles firm described taking contract work from a brand client, where the specification went down to "a needle, a loose thread": the character's shape, its eyes, the stitch spacing, the stains — every item had a rule. "Flaws we'd let through on an ordinary order had to be pulled out and redone on this line; the offcuts after cutting piled up into a mountain by the end of the day."

That's what contract manufacturing really taught Chinese factories: factory-audit standards, process discipline, delivery specifications. Not technique — discipline. Through repeated sampling and rework they learned to control colour variance, waste and yield.

And the cost? Four words: you have a product but no name.

A vacuum-flask and cup maker in Yiwu, Wu Xianmin, put it most bluntly: "The factory has the right to produce, but not the right to sell." Back in 2016 he ran five tyre shops in Shenzhen with "Hua Xin" clearly written on the fascia. Customers walked in and asked only: do you have Continental? Michelin? Nobody remembered the name of his shop.

So what is Zhejiang's industrial belt doing now? Two roads.

Road one: borrow IP. The same factory's fan sells for a bit over ten yuan unbranded, and can be priced above 40 yuan with a well-known cartoon character on it. Disneys-licensed cold-drink cups retail at two to three times an ordinary product. How big is that market? An industry report from the China Toy and Juvenile Products Association puts 2025 licensed merchandise retail at RMB 179.85 billion, up 15.9%, with RMB 6.83 billion in licensing fees, up 12.3%.

But a borrowed name is never yours. When the licence expires, the story and the traffic leave together. Consumers remember the character. They may not remember the factory.

Road two: spend real money on your own name. One Zhejiang cup maker began building its own brand IP, "Tang Zui Ge" (Burns-Your-Mouth Brother), in 2023: first packaging and product marks, then dealer stores, putting the character on shelves and shopfronts, then advertising on high-speed rail and at airports. Last year's brand promotion spend was around RMB 30 million, of which the character — developed with an outside consultancy on a contract of three years or more — costs over RMB 6 million a year.

That's the price of pricing power. Under contract manufacturing you count steel, labour and depreciation. When you build your own brand, design fees, licensing, packaging, content, advertising, after-sales and inventory all land on the books. Market risk that used to sit with the brand shifts forward onto the factory.

Then there's a third thing, and it was the one that excited me most: shared factories.

At the XbotPark robotics base in Songshan Lake, Dongguan, there's a shared factory. It doesn't rent equipment — it does supply-chain management. Public material from May 2025 shows it has worked with over 100 companies or teams, partnered with more than 100 supply-chain service providers, and helped startups ship 300,000 units. Its general manager put it precisely: "We're not a shared space renting equipment. We're a manufacturing brain providing supply-chain management, process optimisation and technical iteration support."

The mechanics: common components — injection moulds, cameras, sensors, motors, batteries — are built into a resource pool first. Roughly 70–80% of the supply-chain development work is done up front; only the remaining 20–30% is developed per project. And it does one more thing: it aggregates similar demand from several small companies and negotiates with suppliers as a bulk buyer — one outlet called it an "industrial Pinduoduo".

Two thoughts arrived at once when I read that.

The first: that's what I should be doing — not making products, but building a back end that lets many small makers survive.

The second: my version of that back end would have to be replicated across a dozen countries. Theirs operates in one Dongguan.

What China's industrial belts do best is three things: speed, density, cost. Around Shenzhen and Dongguan, if you need a new mould, two streets away someone will do it; if you need to change a structure, there's a sample the next day. That density is the strongest anywhere, and no country has caught up. What they don't do well is two things: brand and pricing power.

After finishing this homework I wrote down one line: craft can't be mass-produced, but credit can.

Six models side by side

Six places, one table:

Two conclusions from that table.

First: five of these six models suggest I should not open a dozen studios at once. Italy and Japan are the exceptions — and their price was decades, or a century.

Second: my "one studio per country" is really the Italian model plus the French model, minus the institutional scaffolding that holds both of them up. That subtraction is where the risk lives.

A question I don't want to answer

So, at the end of it, I have to ask myself one question.

Put the six models side by side: does "one studio per country" actually beat the Korean ODM model?

I genuinely tried to answer it.

Speed: Korea wins, by a lot. A new brand from concept to shipping might take three months. In three months I might not have finished signing the lease on one studio.

Cost: Korea wins. One back end serves a thousand brands; marginal cost is close to zero. Every studio I open, I re-count the cost from scratch.

Risk: Korea wins. If a brand doesn't sell, that's the brand's problem; the back end still gets paid. I'd be eating it myself.

Scale: Korea wins.

The one thing I win on is this: if I can pull it off, my brand is one nobody can copy. Because its foundation isn't capacity — it's trust, and trust is made of time, people and relationships. It's a variable money can't accelerate.

So honestly: I am not choosing the easiest road.

Half of my reason is business logic — I believe the cost of trust can become a moat. But I have to admit the other half might simply be that I don't want to be the back end. I've been the back end for twenty-six years. I've shipped for a lot of people. Their name was on the box; mine was on the customs paperwork. I don't want to play that part again.

That isn't a rational reason. But it is a true one. And I think a man in his fifties, making a decision, is better off admitting he has a reason like that than pretending he's purely rational.

6. The Thing Nobody Will Say: This Doesn't Look Like Luxury

Having done all that research, I know which road I want. But then I have to face something harder — not the market, not the competition. Myself.

This isn't a daydream. It already has a concrete home: HANDSOME, the brand I invested in, now selling direct overseas. Its own words for itself are bespoke luxury.

Then I open my own site and look at the numbers:

• Aramid Fibre: $39–$50. 0.7 mm, 12 g, MagSafe.
• Leather: $29–$39. Full-grain leather over a shock-absorbing frame.
• Drop Protection: $15–$29. Raised camera and screen lips, air-cushion corners.
• Free shipping over $60.

And the words on the site are these:

"We state the actual material and thickness. No 'military-grade' hand-waving."
"Every claim on this site is a number you can measure with a caliper and a scale."
"Aramid is thin, not bombproof. Leather patinas, it does not stay new. We say so."
"$15–$50, with the material cost difference visible in the price."

I read those sentences back, then read the words "bespoke luxury" again. I sat there a long time.

Because those two things contradict each other.

Luxury does not talk like that.

The rules of luxury are: price it high enough that most people can't afford it, produce little enough that most people can't get it, tell a story most people can't follow. "We publish our limits" is what an honest company says, not what a luxury brand says. Luxury never volunteers its own weaknesses. It talks about heritage, about craft, about the fact that what you own isn't a product but a position.

I can go on saying "we only work in three materials", "we publish the actual thickness", "we state the limits". All of that is true, and I believe it's good. But a brand selling at $15–$50 that calls itself luxury is making a claim that doesn't hold.

So there are two roads, and both have a price.

Road one: push the price up.

From $50 to $200, $300, higher. The conditions for making that work are everything I described earlier: hand craft, one craftsman spending three days on one piece, named makers, traceable materials, a story not everybody can buy into. PITAKA did $49.99–$79.99 on a material narrative; Mous does £44.99–£64.99 on engineering proof; CASETiFY goes past ¥1,000 on collaborations. Those are precedents. Getting there is possible.

But what's the cost? The cost is that I'd have to give up the most valuable thing I currently have — plainness. Because once you're at $300, the job becomes making the customer feel it's worth $300. And you will slowly discover that the fastest way to make a customer feel something is worth $300 is to mention fewer of its flaws.

That isn't the road I want. But I won't claim I'll never take it.

Road two: put the word "luxury" down.

Keep selling at $15–$50. Keep stating the limits. Build an honest, good company. The upside is that I don't have to say anything I don't believe, I don't have to learn a luxury accent, I don't have to hire a PR firm to write me the word "heritage".

But the cost is this: the ceiling on that business is "an honest, good company". I still remember the number from earlier — industry average gross margin below 15%, 68% on shared moulds. In the $15–$50 band, without a story that supports a premium, you eventually slide back into price competition — and then you die there.

So here's where I am: I either give up the word "luxury", or I give up the tone of plainness. And one of those is my ambition, and the other is my character.

My provisional answer is an uncomfortable one:

I don't think "luxury" is a status you can declare. You can't announce that you're luxury. You can only do the work until one day customers decide you are. And until that day, the only thing I can do is keep telling the truth while raising the share of material, craft and handwork in what I make — until the price rises because of cost and craft, not because I want to charge more.

That may be a slow process. Possibly so slow it never catches up with the word.

So my honest answer to the question: I don't yet know whether this is a luxury business or an honest, good one. What I do know is that I'm not willing to learn to say things I don't believe in just to earn the word.

If the answer turns out to be "an honest, good business", I'll take it. But I'll keep trying — trying once, with real craft, to push that door open. If it doesn't open, I won't blame anyone. I'll go on doing what I know how to do.

7. The Risks: I Won't Sell You This Dream

Everything above, told by a marketing person, would come out as a beautiful story. But twenty-six years have taught me: a plan that won't name its risks is usually a plan that hasn't been thought through.

I see four real risks. I've already named the first above — choosing people. Here are the other three, and I'll say which one keeps me awake.

This is the least romantic point, and probably the most practical.

The problem with building a brand is slow payback. Start a brand today and it might be three years before anyone knows it. What do you eat for those three years?

So I want to do OEM/ODM at the same time. Use the studios' capacity and the craftsmen's hands to take other people's orders — designer labels, concept stores, independent brands. Those orders carry a lower price than our own brand, but the cash comes faster and the risk is lower.

There's precedent. Every DTC brand I mentioned above has its supply chain concentrated in Shenzhen and Dongguan. They don't run factories; they contract. And in that chain, the contract manufacturer earns the thinnest slice. But what if the contract manufacturer also had its own brand?

I don't know the answer. But it's a better structure than "pure contract" or "pure brand". At minimum, it means I don't have to survive on brand revenue in year one.

Risk one: management radius

One studio per country sounds lovely. Ten countries means ten tax systems, ten labour codes, ten languages, ten time zones and ten sets of unwritten local rules you don't understand.

The biggest team I ever ran was about twenty-odd people at a Guangzhou trading company. I know what managing people is: you think you're managing work, but you're managing hearts.

And this model has an insidious property: it won't die all at once. It will die slowly. One studio in trouble and you can carry it. Three at once and you start robbing Peter to pay Paul. Five and you start lying. Eight and you no longer know which one is actually profitable.

Honestly, I don't have the answer yet. The only thing I've got is: don't open ten at once. Get one working, then open the second. Which is the same logic as the craft — don't make a hundred thousand at once; get one right, then the next.

Risk two: how do you keep handmade quality consistent?

This is the hardest one in my view, and there's no way around it.

What is industrialisation, at bottom? Eliminating variation. One mould, one machine, one batch of material, ten thousand identical cases. Consistency is a by-product.

What is hand craft, at bottom? Preserving variation. One craftsman, one design, ten pieces — all ten different. That's the value of handmade, and also its disaster.

Because customer tolerance differs. A customer buying art accepts that every piece differs. A customer buying a phone case does not. You say "that's the character of handmade", and they say "then show me the second one".

I looked at PITAKA's page: the PDP states "2 months warranty". Against a $60–80 price, a two-month warranty is an obvious weak point. And BGR's testing found its UltraGuard Orca Resin coating "liberally" collects fingerprints and oil. It can get away with that because it sells a material narrative; customers are buying the material, not the durability.

But if I sell hand craft, I can't do that. Because the promise of hand craft is "this was made well". The opposite of made well is made badly, and customers can tell.

My current thinking: give the consistent parts to machines, and the inconsistent parts to people. Structure, dimensions, cut-outs — machine, must be identical. Surface, feel, colour matching, stitching — human, variation allowed. That might be a way through. But I haven't tested it, and I won't claim it works.

Risk three: the cash-recovery cycle

This is the one that worries me most.

The model I'm describing has three cash outflows:

• Funding to start local studios. Once that money goes out, nobody knows when it comes back. And since it's "not a loan", if the studio fails, it's genuinely gone.
• Material purchasing. Supplied from Hong Kong means I finance it: leather, timber, freight, insurance, letters of credit — all paid first, collected later.
• Training people. A craftsman teaching an apprentice; for the first six months the apprentice is a liability — they ruin material, slow things down, and leave.

Three outflows open, one inflow: selling goods.

And handmade, made-to-order, one at a time — the essence of this model is slowness. If you ask customers to wait, you can't take a brutal deposit (take too much and they walk). And if you don't take a deposit, you're financing the customer's goods with your own money.

This is the thing I've seen kill the most people in twenty-six years of trade: not dying from no orders — dying from orders you can't collect on.

From 1998 to today, every trading company I've watched fold went under not for lack of business, but because business was too good and the money didn't come back.

This is the one that keeps me up. Management radius can be solved with slowness. Quality can be solved by splitting the process. But the cash-recovery cycle is structural — the more successful you are, the more money you have out. It's a deeply counter-intuitive trap: the better business gets, the tighter cash becomes.

My current answer: let the fast money from OEM/ODM feed the slow money of our own brand. It isn't an elegant answer, but it's a survivable one.

8. Three Plain Conclusions from Hany

One: ninety percent of this industry competes on price, and dies there. That 1688 number — 68% on shared moulds, average margin below 15% — isn't a statistic. It's an epitaph.

If you're thinking of entering, the first thing to decide isn't your supplier. It's whether you're going to fight a price war. Fight, and you're one of the 68%. Don't fight, and you need a story the customer can check. There is no third road.

Two: the premium doesn't come from the plastic; it comes from a story the customer can check. CASETiFY sells collaborations and community, Mous sells engineering proof, PITAKA sells material science — all three are verifiable. Hand craft is the fourth, and its moat isn't technique. It's the cost of trust.

Technique can be bought. Three months with a craftsman and you have it. Trust can't be bought. Building trust in a place takes one year, two years, five years. That's slow — but it's a slowness no one can catch up to.

Three: making one at a time isn't a production method. It's a pricing method. I don't know whether I can reach volume. But I know that selling 120,000 at a price that doesn't require a price war beats selling 15 million at a price that's already through the floor.

I hesitated writing that. Because "beats" is a word I have no data for. All I have is twenty-six years of feel. Feel isn't evidence, but feel is what I make decisions with. I'll admit that.

9. What I Still Haven't Figured Out

I'm not going to close with a sales pitch. I have nothing to sell. This whole piece is one man taking a business apart.

Here's what I haven't worked out. I'll list it. If you know the answers, tell me.

1. Will the two lines trip over each other? Same studio, own brand by day, OEM by night. Sounds reasonable. But the nature of a brand is scarcity, and the nature of OEM is capacity. The same people, in the same place, doing two opposite things — will both end up done badly? I've seen plenty of factories fail at their own brand precisely because their DNA was contract manufacturing: they know how to make goods, not how to make a brand. My current structure is another version of the same problem.

2. Will the craftsmen leave? I train an apprentice, they work three years, they know everything, then they open their own place. Is that good or bad? If they stay in my system, they're mine. If they leave, they're a competitor. But I can't tie them down — tied-down people don't make good work. My provisional view: if they can leave, that means my system isn't good enough. But that's a nice-sounding line, untested.

3. How long can Hong Kong's credit-intermediary role last? In July 2026 Hong Kong's exports rose 50.7%, and I wrote a piece about it at the time. But behind that number is AI electronics. If that wave recedes, does Hong Kong's credit-intermediary role shrink with it? My studio model sits on top of Hong Kong's credit system. That's a foundation question, not a decoration question.

4. Will the market pay handmade prices? This is the most basic question and the one I'm least sure about. In every business I've done, the market sets the price. I can tell a beautiful handmade story, but if the customer won't pay, then I'm a storyteller, not a businessman.

5. When will I know to stop? In twenty-six years I've done many trades. Some worked, some didn't. Looking back, the ones that didn't shared one thing: I admitted it too late. I don't know whether I'll be too late again this time. That question has no answer; it's just a reminder to myself.

Back to that two-hundred-dollar case in Mong Kok.

I didn't buy it in the end. Not because it was expensive. Because after going through all this, I knew those two hundred dollars were buying a story — and I didn't think the story was worth it.

But if someday someone makes a case where the material is real, the work is genuinely by hand, the craftsman has a name, and I can find out where it was made and by whom —

I might buy it.

And I might pay more than two hundred.

That's my last line on this business: expensive is fine. Not being able to explain it clearly is the problem.

#HighEndPhoneCases #CASETiFY #Mous #PITAKA #ForeignTrade #Premium #Materials #Handcraft #BusinessModel