The Second Sale Is a Different Business

TL;DR: Simon Owens lines up four creator product launches and finds the same curve. Prime, from Logan Paul and KSI, did $1.2 billion in its first year and then sales “dropped by nearly half in 2025.” MrBeast’s Feastables is “still technically growing” but has “largely stalled out.” Chamberlain Coffee took scathing taste reviews on a flavor expansion and lost momentum when Emma Chamberlain stopped posting. The diagnosis is that an audience gets you the first sale, and nothing about an audience gets you the second one. That distinction scales all the way down.

Every one of these launches worked. That is the part to hold onto before the schadenfreude starts, because the failure being described here is a second-year failure, and a second-year failure is only available to people who won the first year.

What is the actual mechanism?

Owens’s argument is that creators are extremely good at one specific thing and that thing is not retail. They are good at entertainment and at the parasocial bond it produces, and that bond converts to a first purchase at rates a conventional brand would need to spend enormous money to approach.

Then the product has to be a product.

The cleanest statement of the problem comes from Emma Chamberlain herself: “This brand has to exist outside of me. People aren’t going to return if they don’t like the product.”

That is the whole mechanism in two sentences, and it is worth noticing that she is describing her own business accurately and unflatteringly in public, which almost nobody does. The first sale is a transfer of trust from the person to the object. The second sale is the object standing on its own, being judged against every other object in a category, by somebody who has now tried it.

Owens’s point about category matters here. Coffee, energy drinks, and candy are, in his words, “highly saturated.” A saturated category means the competition is not other creators, it is companies who have spent decades on formulation, distribution, and shelf placement, and who will spend real money on sustained brand advertising forever. The creator brand shows up with a launch spike and no engine underneath it.

Ryan Trahan’s Joyride is the counter-example in the piece, and it is instructive for the reason Owens gives: over 6.8 million YouTube views on a flavor launch driven by professional marketing. That is a company doing marketing, not a creator doing a video. Different function.

Two failure modes, and they are not the same

Reading the four cases against each other, they separate cleanly, and this is the part that transfers.

Chamberlain Coffee’s problem is product. Scathing taste reviews on a flavor expansion. That is a quality failure, and it is fixable in the way quality failures are fixable, which is expensively and slowly. Compounding it: she stopped posting to YouTube and Spotify, which removed the top of the funnel while the product was still leaning on it.

Prime’s problem is category physics. $1.2 billion is not a fluke and half of $1.2 billion is still an enormous business. Energy drinks are a repeat-purchase category with fierce competition and shelf economics that punish anyone without a sustained ad budget. Losing half the peak may be exactly what the normal curve looks like when a launch spike settles into a real baseline.

Those require opposite responses. One says fix the thing. The other says the thing is fine, the expectations were wrong. Getting them backwards is how an operation destroys a viable product line trying to fix a problem it does not have.

The repeat rate is the only verdict on the product

Nobody reading this has thirty million subscribers. We have a list somewhere between two hundred and five thousand, and I run the log for one of them, so this is the part I checked against our own numbers before writing it down. Some of Owens’s curve survives the scale change. Some of it does not, and the split is clean.

What does not transfer: the saturated-category problem, mostly. You are not competing with Hershey. If you sell a guide about a specific thing you actually did, your category has maybe four other entrants and two of them are bad. Owens’s competitive analysis is about consumer packaged goods, and it does not describe a niche information product.

What transfers completely: the first-sale-versus-second-sale split. Your launch numbers are a measurement of how much trust you had banked, and they tell you nothing about whether the product was good. That is the uncomfortable one. A strong launch and a weak second month is the same curve Prime is on, just with three fewer zeroes, and it means the same thing.

The number to actually watch, and it is the one nobody on a small list computes:

What percentage of your buyers have bought twice?

Not list growth. Not launch revenue. Not conversion rate. The repeat rate is the only metric in the stack that is a verdict on the product rather than on the audience, because a second purchase cannot be explained by parasocial goodwill. They already spent that.

If you have never run it, run it this week. Pull your order history, count unique buyers, count how many appear more than once. If the number is low and your launches are fine, you have a Chamberlain problem, and the answer is not more marketing.

The part I would argue with

Owens frames this as creators struggling, and I would frame it as creators discovering they own two businesses and only staffed one.

The audience business and the product business have different inputs, different failure modes, and different clocks. The audience business runs on consistency of publishing. The product business runs on whether the thing is good and whether people can buy it again without friction. Chamberlain’s momentum loss when she stopped posting is exactly what it looks like when one of those is quietly subsidizing the other, and neither party noticed until the subsidy stopped.

For a one-person operation the practical version is blunt: if your product only sells when you post, you do not have a product line, you have a broadcast with a checkout attached. That is a real business and it can pay well. But it stops the day you stop, and it will not compound.

The fix is not a bigger audience. It is one product good enough that somebody buys the next one because of the last one, and the only way to find out if you have that is to count.

Frequently asked questions

Is a launch spike followed by a drop always a warning sign?

Not on its own, since a launch concentrates demand that would otherwise have spread across months, so some decline is arithmetic rather than diagnosis. What distinguishes a settling curve from a failing one is where it settles. A product that finds a lower steady baseline and holds it has found its actual market. One that keeps declining toward zero after the audience’s attention moves on was carried entirely by the announcement.

How many buyers do you need before a repeat rate means anything?

Enough that a handful of enthusiasts cannot swing it, which in practice means at least a few dozen distinct buyers and preferably a hundred. Below that the number is directional at best, and the more useful exercise is qualitative: contact the people who bought twice and ask what made them come back. On a small list that is a conversation you can actually have, which is an advantage the billion-dollar version of this problem does not get.

Does this argue against ever launching into a competitive category?

It argues for knowing which kind of competition you are in before you commit inventory to it. A saturated physical-goods category demands sustained advertising spend that most creator operations never budget for, and the launch audience cannot substitute for it. A narrow information category with few credible entrants behaves differently, because the audience relationship and the product’s credibility are closer to the same asset there.