In 2017, Ju Rhyu and two co-founders pooled $50,000 and listed one product on Amazon. It was a sticker. Five years later they sold the company for $630 million.

The sticker was Mighty Patch, a hydrocolloid acne patch. Ju found them while working for Samsung in Seoul, where every drugstore had them. In the US, nobody sold them. She had breakouts herself, tried one, and found it worked better than anything she'd used back home.

The first 10,000 units sold out in 90 days. By 2021, Hero Cosmetics had passed $100 million in sales at roughly a 40% EBITDA margin, and it never took outside money. Church & Dwight, the Arm & Hammer company, bought it in 2022 for about 14x EBITDA. At that point Hero held 63% of the US acne patch category, more than five times its nearest competitor.

A good product helped. But the moves behind it are ones any Amazon seller can copy.

1. Amazon was the test

In 2017, DTC was at its peak. Everyone was building Shopify stores and buying Facebook ads.

Hero went Amazon first, with one SKU.

Ju's thinking: "Let's see if people buy it. If people buy it, then we'll work on phase two." Shoppers were already searching for acne fixes, so they didn't need an audience. Amazon was the cheapest way to test the idea.

DTC came after. Then specialty retail. Then 1,521 Target doors in 2019. In Ju's words: "If you build the demand, the retailers will come."

2. They raised the price and it worked

Competing patches sold for about $5. Hero planned to launch at $9.99.

Then they hit a wall: they couldn't get the Prime badge. While testing ways to unlock it, they raised the price to $12.99.

That's the price that sold out in 90 days. They kept it.

Ju calls it "kind of an accident." The lesson isn't. Being the cheapest isn't a strategy. In a category full of $5 knockoffs, a higher price told shoppers this one was better. On a product that costs pennies to make, it also paid for those 40% margins.

3. They learned brand protection the hard way

At first, Hero wasn't brand gated. Other sellers jumped on the listing and undercut the Buy Box by a penny or a nickel.

Ju called it "a very dangerous game because you can just sort of discount this product to zero."

They fixed it by proving they owned the IP and locking down the listing. That control paid off later. Retailers liked working with Hero because its Amazon price matched the shelf price.

Control your listing and your price, or someone else will.\

4. They talked like their customer

This is the big one.

The skincare industry describes acne in polite language: blemishes, imperfections, "blemish-prone skin." The words sound clinical and a little embarrassed, as if the problem is too awkward to name.

Mighty Patch talked about it the way you'd tell a friend. The giant zit you wake up with on the morning of a photo shoot, a first date, or a big meeting. They named the exact moment the customer was living, in her own words.

The polite version keeps the problem at arm's length. The real version makes people feel seen. When people feel seen, they trust that you have the fix.

On Amazon, this isn't only about copy. It's about search. Ju has said shoppers find them by typing "pimple patch," "acne patch," and "pimple solution." Nobody searches "blemish-prone skin solution."

The words your customers actually use are the keywords that drive your rank. Plain language helps you twice: more people find you, and more of them buy.

5. They sold the magic moment

Hero focuses on what it calls the "magic moment." For Mighty Patch, it's the peel-off: you pull off the patch and see what it pulled out overnight.

It's gross, it's satisfying, and people share it. That clip went viral on TikTok again and again, mostly without Hero paying for it.

Find your product's magic moment. Then put it in your main video, your secondary images, and your A+ content. If your product does something people can see, show it happening.

Your 15-minute homework

Polite copy sounds like every other brand. Plain copy sounds like a friend who gets it.

How do you you scale a $10.95 jar of sauce into an 8-figure e-commerce empire?

In this episode of Marketing Misfits, Kevin King and Norm Farrar talk with serial entrepreneur Chris Lang (founder of Fresh Chile Co. and Move FWD agency) to unpack how he bootstrapped a family green chile recipe into a massive $10M+ DTC brand.

Chris shares his raw journey of managing 10 businesses simultaneously, why he shut them down to focus purely on e-commerce, and the exact 4-pillar system he uses to scale Shopify stores to 7, 8, and 9 figures.

Learn how to turn your Product Detail Pages (PDPs) into a 300-post organic content engine, navigate Meta's aggressive P&L cuts, and avoid the dangerous traps of relying on AI without real operator experience.

Finding an Amazon support email address is easy. Getting a reply is the hard part.

You've probably been through this. A case stalls. Seller Support keeps sending you copy-paste answers. So you go looking for a direct line.

You find an address ending in @amazon.com. It looks official. You spend an hour on the perfect message, with every ASIN, case ID, screenshot and timeline laid out.

You hit send.

Then nothing happens. A week later you learn the address was an old alias that nobody has checked in years.

Why @amazon.com means nothing

An @amazon.com address only tells you the domain. It doesn't tell you:

Amazon reorganizes teams, retires aliases and moves responsibilities around all the time. Seller forums and old blog posts don't keep up. An address that worked in 2021 can be a black hole now.

When you're dealing with a suspension, stranded inventory or a listing that's been hijacked, sending to the wrong inbox can cost you days. Sometimes it costs real money.

Vanessa Hung did the research

She built the Amazon Email Directory: 76 internal Amazon seller and vendor email addresses, each matched to:

Every address has a confidence rating

Each address carries one of four statuses, so you know how much to trust it before you write anything:

That status alone can save you from wasting a week on a route that was never going to work.

Real cases, not just addresses

Each row also links to a related case where she fixed one of her customers' Amazon problems, so you can see how that type of issue was investigated and escalated.

Most of those cases didn't use the email in that row. They show the path that worked instead. A few do involve the listed contact.

That matters because the right answer isn't always "find a better email." Sometimes it's a different escalation path entirely, and seeing how a similar problem was actually solved is often worth more than the address.

How to use the directory

Figuring out routing, tracking dead aliases and working out which team owns what isn't work you should have to do yourself. So we put it all in one place.

👉 GET YOUR AMAZON EMAIL DIRECTORY FREE →

Meta's new personal agent, Muse, went mainstream fast. Within a week of its US launch it passed ChatGPT to take #1 on Apple's free App Store. Meta's stock is up about 20% since launch.

Muse doesn't just answer questions. It books, fills out forms and buys things for you.

Then Amazon locked the door.

On September 20, Amazon started blocking Muse from shopping on Amazon.com. The official reason is that the agent was "unauthorized" and violated Amazon's Conditions of Use. Amazon also raised privacy concerns: it says Muse doesn't identify itself and appears to store customer login credentials.

Shopify went the other way. It partnered with Meta so Muse can complete checkouts across Shopify stores.

Amazon is playing defense and Shopify is playing offense. Sellers have a stake in both.

Why Muse scared Amazon when Perplexity didn't

Amazon has been fighting outside agents for over a year. Perplexity's Comet browser was the first real fight. It was annoying but manageable, because Comet is a niche tool used by early adopters. Most shoppers are still figuring out how to clear their cache.

Muse has no such ceiling. It comes with:

Amazon's real moat was never just price or logistics. It's discovery, the engine that decides which products a shopper sees first.

An agent running on Meta's data could rebuild that discovery layer outside Amazon's walls. If that happens, Amazon still ships the box, but someone else decides what goes in it.

Follow the $68 billion

This is where sellers should pay attention.

Amazon's ad business brought in more than $68 billion last year, and all of it depends on one thing: humans browsing pages.

Agents don't browse, and they don't click ads. An agent that goes straight from "buy me running shoes" to a finished checkout leaves nothing to sponsor. It skips the search results page, the ranked product grid and the placement you paid for.

Meta makes this worse for Amazon because Meta runs a giant performance ad business of its own. If discovery moves into a Meta assistant, those ad budgets have an obvious new home.

The "do as I say, not as I agent" problem

Amazon isn't against agentic shopping. It just wants to be the one doing it.

So Amazon's agents roam everyone else's stores while Amazon bans Meta's agent from its own. Amazon says the difference is that its agents announce themselves. That difference is real, and it's also very convenient.

Fraud verification tells you the most. Amazon wants its agent to be the one that decides what's trustworthy, while arguing that nobody else's agent can be trusted.

Can contract law hold the line?

Probably not for long.

Amazon won an early court order against Perplexity this spring. It lost on appeal this summer. The court ruled that under anti-hacking law, the user is the one accessing Amazon when an agent acts on the user's instructions. A rehearing was denied.

So Amazon switched from hacking law to contract law. It added a rule that agents must identify themselves in their browser string. But Amazon also claims Perplexity changed that string every time Amazon learned to detect it. A wall you have to rebuild every week is really just a speed bump.

You can block a company, not a category

Amazon isn't fighting one rival. It's fighting a trend.

Amazon can sue or block each of these companies one at a time. It can't block the whole shift.

What sellers should do now

Everything you've built on Amazon assumes a human sees your listing, reads your A+ content and notices your sponsored placement. That assumption is starting to crack.

Amazon is buying time for its own agents. The real question is what you're doing with that same window.

✌🏼 Have a great weekend.

See you again on Monday.

The answer to today’s STUMP BEZOS is
Bezos moved to Miami in 2023.