Whatever You Do as a Brand Compounds. For Better or Worse.

Fourteen stores in Henan. No advertising. Revenue more than doubled in three years. Trust, printed on the price tag.

There’s been over a million people who have travelled to a fourth-tier Chinese city this year to visit a supermarket, and not one of them was sent by an ad.

Fourteen stores in two cities in Henan province, and last year they turned over 23.5 billion yuan, about A$4.9 billion. Three years ago it was 10.7 billion yuan. Exactly the same company, same two cities, roughly the same number of stores, and revenue more than doubled while their competitor Yonghui, one of China's biggest chains, was closing more than 200 stores and losing money for the fifth consecutive year running.

Here's what else you need to know. There was no advertising, no loyalty app or personalisation engine. The fame came from people filming the place and posting it, and there's now an intercity train from Zhengzhou airport that locals have nicknamed the Pang Dong Lai line.

So let's be razor sharp and clear here about what this is, because I can already feel this industry about to file it under "an interesting China story" and then move on to the next AI vendor pitch.

This is exactly the proof of the position IGU has taken for years and maybe I’ve been told is too soft, too emotional or not commercial enough. Trust is not a value. Trust is the competitive advantage and the only one left that can't be copied by the brand next door with a bigger tech budget or advertising budget and it's built through one thing and one thing only.

It’s not what you say. It’s what you do as a brand and business.

What you say versus what you do are two different things and that's the whole argument here, I'm also not going to soften it to make you feel better about what you're currently doing. Every brand has trust on their values slide or plastered to their walls and logos and in almost every one of them it sits there as a claim the brand makes and the store is supposed to somehow deliver. Pang Dong Lai doesn't claim it anywhere, instead every mechanism in the business is built so the customer can watch it being done, and trust that's watched compounds in a way trust that's promised never ever can.

The Pang Dong Lai retail model prints what your brand book can't.

A shirt at 118 yuan, and the ticket beside it shows the buy price, 84.6, and the margin, 28.3 percent. Gross margin on everyday goods is capped at 30 percent, no supplier pays a slotting fee, no promo kickbacks, and the customer reads all of it while standing in the aisle making decisions.

Now let's think about what that does to any human being, because this is the bit our industry refuses to sit with.

The question running under every retail transaction now is "am I being had", and it happens whether the customer says it out loud or not. That's not a rational calculation, it's a feeling, a low hum of vigilance the brain carries into every single store, and it costs you conversion every single day without ever showing up in your reports. The price tag here switches it off before it's even asked. Why, because this retailer has voluntarily handed back the information advantage the whole industry is built on, and when the seller gives up its edge the brain reads that as safety, and safety is the precondition for spending.

That's emotional intelligence as a commercial discipline. It’s not empathy training, not a tone of voice document but knowing exactly what the customer is feeling in the moment of decision and engineering the environment to change it.

What this shows us is that transparency isn't some comms tactic, it's a pricing architecture, and if yours lives in a deck rather than on the shelf you don't have one.

Freshness is a clock the customer can see running.

Produce is pesticide-tested daily and the results are displayed in the store. Meat doesn't stay overnight. Cut fruit reportedly runs on a visible discount ladder by the hour until it's pulled and when a supplier's noodles were found to have breached food safety procedures in 2024, Pang Dong Lai went and paid 8.8 million yuan in compensation to its customers, said so publicly, and then kept going.

None of that is a promise. It's a smart rule with a clock on it and the customer watches the rule operate every single time they visit.

This is the mechanism of compounding, and it's why we say trust is infrastructure and not marketing. A promise has to be re-evaluated every time, the customer checks it against experience and the checking costs them something. Then a rule they've watched run a hundred times doesn't need checking any more. Trust stops being a judgement and moves to a habit, and habit is where your margin lives, and the retailer who gets there first owns that customer's default for many many years.

Your competitor can copy your range in a season but they cannot copy a hundred visits that's for sure.

The people running it own it.

So the average pay is 9,000 yuan a month, which is more than double the provincial average. Seven-hour days, no overtime, forty days of leave, and then ten more days of what Yu Donglai calls unhappy leave, if you're not happy don't come in. Since 2000 he's distributed shares to staff and kept ten percent himself, and in March this year he moved four billion yuan of company assets over and into the hands of employees and management, split down the middle.

He said on national television that if you treat employees well they'll treat the business well, and it sounds like a poster until you follow it to the floor. The person restocking the cut fruit is a shareholder, the person putting the pesticide sheet on the wall is a shareholder. The behaviour doesn’t need to be policed, it's owned, and that's the only version of behaviour that survives a bad Tuesday.

What this means is that trust with the customer is downstream of trust with your people, and I hate to say it but most of the businesses currently running culture workshops are just trying to get the first without actually paying for the second.

Here's why your dashboard says this can't work.

Run this model through any conventional P&L and it fails on pretty much every line. Margin capped. Labour costs roughly triple the category. Product pulled that could still be sold. Compensation paid that wasn't legally owed.

Now look at what the customer did. Ten point seven billion yuan in 2023. Nearly seventeen in 2024. Twenty three and a half in 2025, which is close to five billion Australian dollars. Profit of more than 800 million yuan in 2024, around 165 million Australian dollars, from just thirteen stores. Yonghui lost nearly twice that across more than 800.

And when Yonghui handed stores over to be rebuilt on the Pang Dong Lai retail model, its Xi'an Zhongmao store went from 200,000 yuan a day to 1.6 million, around 330,000 Australian dollars, in the first two days after reopening, and footfall from 3,000 to over 14,000.

Miniso's founder, who took over Yonghui's reform, has said publicly the Pang Dong Lai model is the only way out for Chinese supermarkets.

So the metric that says trust is a cost is measuring the wrong side of the transaction unfortunately. The cost lands immediately and the return lands in behaviour, and behaviour is the only thing your customer ever actually gives you.

And if your finance team can't see it, I need to say it's not a finance problem, it's a measurement problem, and it's the one we spend most of our time fixing.

The tech isn't the moat. The human is. Unapologetically.

Here's the part our industry doesn't want to hear, and I'm going to say it anyway.

Pang Dong Lai has no personalisation engine to speak of. No AI merchandising, no chatbot, no app nudging you back. It has a person at the seafood counter who'll clean and pack the fish, a printed card on the shelf telling you how to eat the fruit in order of sweetness, and a founder who closes profitable stores because the old ones no longer meet what customers expect.

We've spent five years, and I include myself, chasing the illusion that the next system would do the trust for us. Personalise harder, automate the checkout, put an agent on the enquiry line. And every one of those moves takes a human out of the moment where trust is actually decided, and then hands the customer another reason to wonder who's accountable when it goes wrong.

Let me be precise here, because I might get quoted on this lol. AI has its place, it belongs in your supply chain, your forecasting, your back office, and it'll do those things better than any human. It does not however belong in the moment the customer decides whether to believe you. That moment is human, it's emotional, and every dollar you spend automating it is a dollar spent making yourself indistinguishable from the competitor doing the same thing.

What this tells us is that the more the industry automates, the scarcer an accountable human becomes, and scarcity is the definition of luxury. A human made retail experience, where you can see the rule, see the person, and see who owns the outcome, is the premium tier now. Not nostalgia but rarity. And it’s only getting rarer every quarter the rest of the industry chases the demo.

This is the position. IGU Global exists because we believe brand performance is an emotional engineering problem, not a marketing one and not a technology one, and Pang Dong Lai is what it looks like when someone actually builds a business on that instead of putting it in a strategy day.

Here's three tests you could run from Monday.

Print the price tag. Any one category, one store and one month. Put the cost price and margin on the ticket and watch what happens to conversion and to the conversations your floor staff have. I’d say if you can't do it, that tells you something about your margin, not your customer.

Put a clock on something perishable and let the discount run by the hour where the customer can see the rule working. Then measure whether they trust the full-price product more.

Ask your floor staff what they'd change if they owned the place and then ask why they don't already behave that way, and be honest about whether the answer is culture or whether it's pay.

The villain isn't China and it isn't the founder.

The easy out is that this only works because it's China, or because Yu Donglai is a one-off, or because it's fourteen stores and you've got four hundred. The scalability question is fair. Yonghui's remodelled stores are profitable and the group as a whole isn't yet, and a business where the founder is the story carries its own risk.

But that's the excuse, not the lesson.

The mechanism isn't Chinese and it isn't charisma. Show the customer the rule, let them watch it run, and give the people running it a reason to run it when nobody's looking.

That's it. That's the whole competitive advantage, and it's available to every single brand and business reading this.

There's a lot that won't take it and will buy the software instead.

We've all had the values slide. The difference here is that Pang Dong Lai printed it on the price tag.

The customer read it, and then came back.

Nick Gray

About the author: Nick Gray is the founder of IGU Global, a Sydney brand strategy and retail consultancy, and was named a Top Retail Expert 2026 by Rethink Retail. With 25+ years across Nike, Adidas, Diesel, Sneakerboy and Westfield, he works with startups, mid-size brands and enterprise clients on emotionally intelligent brand strategy, consumer psychology and AI's role in retail.

Work with Nick at  iguglobal.com.

Common questions

What is the Pang Dong Lai retail model?

Pang Dong Lai is a Chinese supermarket group running 14 stores across two cities in Henan province. Its model makes trust visible rather than claimed: cost price and margin are printed on shelf tickets, gross margin on everyday goods is capped at about 30 percent, suppliers pay no slotting fees, and freshness rules run on a clock customers can watch. Revenue reached 23.5 billion yuan in 2025, up from 10.7 billion yuan in 2023, with no advertising.

How does Pang Dong Lai grow without advertising?

Growth comes from customers filming and sharing the stores rather than from paid media, a loyalty app or a personalisation engine. More than a million people have travelled to the fourth-tier Chinese cities where it operates, and an intercity train from Zhengzhou airport is nicknamed the Pang Dong Lai line. Trust customers can watch being delivered compounds into habit, which removes the need to buy attention.

Does Pang Dong Lai print its profit margin on price tags?

Yes. A shirt priced at 118 yuan carries a ticket showing the buy price of 84.6 yuan and a margin of 28.3 percent. By handing back the information advantage retail is usually built on, the retailer switches off the customer question of whether they are being taken advantage of, and that sense of safety is the precondition for spending.

How much does Pang Dong Lai pay its employees?

Average pay is about 9,000 yuan a month, more than double the provincial average. Staff work seven-hour days with no overtime, take 40 days of leave plus 10 days of what founder Yu Donglai calls unhappy leave, and hold shares in the business. In March 2026 he moved four billion yuan of company assets to employees and management. Customer trust is downstream of employee ownership.

Can the Pang Dong Lai model work outside China?

The mechanism is neither cultural nor dependent on its founder. It is three transferable moves: show the customer the rule, let them watch it run, and give the people running it a reason to run it when nobody is looking. Yonghui stores rebuilt on the model produced the same effect: the Xi'an Zhongmao store went from 200,000 yuan a day to 1.6 million, with footfall rising from 3,000 to over 14,000.

Book a 30-min Clarity Call →

No pitch. No pressure. A first conversation to see if there's a fit.