Daniel's been thinking about the parametric search episode and he's landed on something that's been quietly bothering him. He works in technical marketing, so he's careful to say marketing isn't the enemy — good marketing tells a story, especially in things like case studies and white papers. But when you narrow the frame to procurement, to actually buying something, the adjectives that are supposed to sell you on it — "cutting edge," "latest," "best-in-class" — are just noise. They don't help you decide.
They actively get in the way.
Right. And he points out that niche wholesale marketplaces like Digikey and Mouser solved this decades ago. They describe everything plainly, they live and die by parametric search, and their sites look like they haven't been redesigned since the Clinton administration. But they're not selling toasters. If you need a kettle, you're back in the world of marketing copy. So here's his question: what if a B2C marketplace insisted that sellers supply purely factual, technically descriptive product descriptions — or did the translation work themselves, acting as a kind of linguistic middleware that strips out all the promotional text and just lets buyers evaluate options based on facts? Has anyone tried this? Is there any movement in this direction? And he adds this wrinkle — he says there's something almost anti-consumerist about the idea. If you remove the emotional leverage of copywriting, the appeals to vanity and aspiration, you might end up with a more wholesome environment where people only buy what they actually need.
That last part is where this gets interesting. Because he's not just asking about a feature — he's asking whether the feature changes behavior. Whether stripping the adjectives strips the demand.
So two questions really. Has anyone built the thing, and would it work the way he thinks it would?
Let's take the first one first. Has anyone built it. The short answer is no, not in the way Daniel's imagining — not as a general-purpose B2C marketplace with a policy of enforced factual description. But pieces of it exist, and some of them are closer than you'd think.
Give me the closest.
Wirecutter. The New York Times bought them in twenty sixteen for about thirty million dollars. Their entire model is "we test things and tell you which one to buy, and here's the factual basis." They strip away the marketing claims and replace them with measured performance. It's not a marketplace — you still go buy the thing somewhere else — but as an information layer sitting on top of commerce, it's essentially what Daniel's describing. They do the translation work.
They're the linguistic middleware, just not at the point of sale.
And they've been wildly successful at it. What's interesting is that they proved there's a mass audience for de-marketed product information. Millions of people actively seek out the version of the description with all the adjectives removed. The other close parallel is Consumer Reports, which has been doing this since nineteen thirty-six — longer than most of the brands they review have existed. They buy products at retail, test them anonymously, and publish results with no advertising. Their whole value proposition is "we are the place where nobody is trying to sell you anything."
And they've got what, six million subscribers?
Something in that range, yeah. So the demand is real. People want the unvarnished version. The question is whether you can build the marketplace itself around that principle, rather than building a review layer on top of a marketplace that's still full of marketing copy.
Because that's the thing — Wirecutter and Consumer Reports are parasites on the existing system. They need the marketing-heavy marketplace to exist so they can translate it.
Right. They're not changing how products are listed. They're adding a second layer. What Daniel's proposing is collapsing those layers — making the primary listing itself factual.
So why hasn't anyone done it?
A few reasons. One is structural, one is economic, and one is — and I think this is the one that actually matters — that most consumers don't want it.
Say more on that last one.
There's a body of work in consumer psychology going back decades showing that people buy based on emotion and then rationalize after the fact. The rationalization step is important — you need to feel like you made a smart decision — but the decision itself is rarely driven by a feature matrix. A famous study from the nineteen eighties, and I'm going to get the details slightly wrong here but the shape is right — researchers gave people a choice between two apartments. One was objectively better on paper, the other had a nicer view. People overwhelmingly picked the view, and then when asked to explain their decision, they invented practical justifications that had nothing to do with the view.
So the adjectives are doing work that the specs can't.
They're doing the work of making you feel something. And feeling something is how most purchases happen. Daniel's idea would work brilliantly for the small slice of purchases where people are already in a parametric mindset — buying a laptop, comparing processors, that kind of thing. But for a toaster? Most people don't want a toaster. They want toast. The toaster is just the means, and they don't care about the heating element material or the wattage distribution.
They care that it looks right on the counter.
Which is a legitimate criterion. It's not irrational to care about how something looks in your kitchen. The marketing copy that says "brushed stainless steel complements modern décor" is actually conveying useful information — just not in a parametric format.
So you're saying the adjectives aren't all noise. Some of them are carrying signal about things that don't fit neatly into a spec sheet.
And this is where I think Daniel's framing — which he's careful about, he draws the distinction between good marketing and empty adjectives — but even the empty ones are doing something. "Cutting edge" doesn't tell you anything specific, but it signals that the manufacturer believes this product is competitive with whatever just came out. It's a positioning claim. It's not falsifiable, but it's not meaningless either.
It's a promise that someone somewhere is keeping track of what's current.
And if you're buying something you don't follow closely — which is most things — that promise has value. You're outsourcing a small piece of your market awareness to the person who wrote "latest technology" on the box.
Alright, but let's take Daniel's side for a minute, because I think he'd push back here. He's not saying every purchase should be parametric. He's saying there should be one place where you can go when you do want that. A refuge. And right now that refuge doesn't exist for consumer goods.
No, it doesn't. And the closest attempts are instructive in how they failed.
Tell me about the failures.
The most direct attempt was probably a startup called GoodGuide. Launched around two thousand seven, raised about ten million dollars, got a lot of press. Their pitch was: we rate every product on health, environmental, and social metrics — purely data-driven, no marketing. You could scan a barcode and get the real story. They had a mobile app, they had a website, they had scientific advisory boards.
What happened?
They couldn't make the economics work. The data was expensive to gather and maintain, manufacturers hated them because they had no control over their ratings, and consumers — this is the brutal part — consumers said they wanted this information but didn't actually change their buying behavior based on it. GoodGuide got acquired by UL in twenty twelve and basically disappeared.
The stated preference versus revealed preference problem.
It's the oldest story in market research. People tell surveyors they want factual, unbiased product information. Then they buy the thing with the nice packaging. And I don't say that dismissively — the packaging is part of the product experience. Opening a well-designed box is pleasant. It's not a rational part of the purchase, but it's real.
So GoodGuide is the cautionary tale. What about something that's still alive?
There's a site called ProductChart that's interesting. It's not a marketplace — it's a comparison tool — but they basically do what Daniel's describing for specific product categories. You pick a category, they show you a grid of specs pulled from manufacturer data sheets. No marketing copy, no affiliate-link fluff. It's extremely bare-bones. And it exists, and it's useful, and almost nobody uses it compared to Amazon.
Because Amazon has the reviews.
Reviews are the other piece of this that's hard to replicate. A parametric listing tells you what the manufacturer claims. A review tells you whether the thing actually does that. And reviews, even with all their problems — fake reviews, review bombing, people giving one star because the delivery was late — are still the closest thing we have to ground truth at scale. A purely factual marketplace without reviews would be less useful than what we have now, not more.
Unless the marketplace did its own testing.
Which is Wirecutter. And that doesn't scale to a full marketplace. You can't independently test every toaster on the market.
Let me try a different angle. Daniel mentions Digikey and Mouser. What makes them work?
A few things. First, their buyers are engineers. They're not just in a parametric mindset — parametric search is literally their job. They're sourcing components for a design, and if they get the wrong transistor, the board doesn't work. The stakes are different.
The failure mode is not "I don't like how it looks on the counter." It's "the circuit board catches fire."
Right. Second, the products themselves are parametric by nature. A transistor has a defined set of properties — voltage rating, current capacity, package type, switching speed. There's no room for adjectives because there's nothing to describe except the specs. A toaster has specs too, but they're not the point. The point is whether it makes good toast consistently, which is hard to capture in a parametric field.
And the third thing?
Digikey and Mouser are not consumer-facing in any meaningful sense. They don't need to create desire. Their customers already know exactly what they need. The marketplace is purely fulfillment. That's a completely different business from B2C retail, where a huge part of the value is discovery — you didn't know you wanted this thing until you saw it.
Which circles back to Daniel's anti-consumerist point. He's saying maybe that discovery function is the problem. Maybe the marketplace shouldn't be trying to create desire at all.
And I think that's a philosophically coherent position. But it's not a business model. A marketplace that deliberately avoids making you want things is a marketplace that's leaving money on the table, and in a competitive market, someone else will capture that money. The only way it works is if it's a nonprofit or a public utility.
Or if it charges a subscription.
Consumer Reports does. Wirecutter has a paywall for some content. So the model exists — you pay for the absence of marketing. But Daniel's proposing something bigger than a review site. He's proposing an entire marketplace built on that principle. And the transaction volumes you'd need to make a marketplace viable...
You'd need a lot of people who want the unvarnished version badly enough to change where they shop.
And the evidence suggests that's a niche. A passionate niche, but a niche.
Let me push on that, actually. Because I think there's a generational shift happening. Younger buyers — I'm thinking Gen Z, younger millennials — they've grown up with infinite marketing. They've been advertised to since birth across every surface of their lives. And they've developed what I can only describe as marketing antibodies.
The ability to just not see it.
More than that. An active distrust of it. They'll see a product on Instagram, screenshot it, go to Reddit and ask "is this actually any good," and buy based on what strangers tell them. They're already doing the linguistic middleware themselves, just in a distributed, inefficient way.
That's a fair point. The behavior Daniel wants to centralize is already happening in a decentralized form. Reddit, YouTube reviews, Discord communities — people are building their own de-marketed information layers because they don't trust the official ones.
So maybe the question isn't "would anyone use this" but "has anyone built the centralized version well enough to capture the behavior that's already happening."
And I think the answer is still no, but I'm less confident than I was a minute ago that it's impossible. Let me think about who might be positioned to try. Amazon could do this tomorrow if they wanted to. They have the product data. They could add a "specs-only view" toggle — strip out the marketing copy, show only the parametric fields, the dimensions, the materials, the warranty information. It would cost them almost nothing to build.
And it would cost them a fortune in lost sales.
The marketing copy exists because it works. Amazon has tested everything. If removing the adjectives increased conversion, they'd have removed them already.
So the incumbent can't do it because it's against their interests. And a startup can't do it because they can't get the catalog scale.
The catalog scale is the real killer. Digikey works because every product in their catalog is parametric by nature — the data exists and is structured from the manufacturer. But for consumer goods, that structured data doesn't exist in any standardized way. Someone has to create it. You'd need to physically measure every toaster, test every kettle, and write factual descriptions from scratch. The labor cost is enormous.
Unless you got the manufacturers to do it for you.
Which is what Daniel suggests — require sellers to supply factual descriptions. But why would they? They've already got Amazon, Walmart, their own D2C sites. What's the incentive to do extra work for a new marketplace with no traffic?
The incentive would be access to buyers who specifically want this. The anti-consumerist niche. But you're right, that's a small pool to fish in while you're building catalog.
There's a company called Thingtesting that's worth mentioning here. They're not exactly what Daniel's describing, but they're adjacent. They review and catalog direct-to-consumer brands, and their whole pitch is honesty — they tell you which Instagram-famous brands are actually worth buying. They've raised venture funding, they're growing. The model is basically "we are the antidote to viral marketing." But again, it's a review layer, not a marketplace.
So the pattern is clear. Lots of people have built the translation layer. Nobody has built the marketplace itself.
And I think that pattern tells us something about where the value is. The value isn't in the transaction — it's in the information asymmetry. The person who can tell you "this kettle is actually the same factory as that kettle but thirty dollars cheaper" has something worth paying for. The person who just lists the kettles factually doesn't.
Unless the listing itself is the value proposition. "Come here because we don't lie to you."
Which is a brand promise, not a feature set. And brands can be built on that promise — Patagonia, to pick one example, has built an entire company on a promise of honesty about environmental impact. But Patagonia still uses marketing copy. They just use different adjectives.
"Responsibly sourced" instead of "cutting edge."
It's still an adjective. It's still doing emotional work. It's just aimed at a different emotion.
This is where I think Daniel's idea might be more radical than he's letting on. He's not just talking about removing bad adjectives. He's talking about removing all of them. A product page that reads like a lab report. And I wonder if that's even possible for most products.
It's possible for some. A USB cable — you can describe it entirely in terms of length, connector type, data transfer speed, charging wattage, shielding material. There's nothing else to say about it. But a sweater? What are the parametric fields for a sweater? Material composition, sure. Dimensions, yes. Country of origin. Care instructions. But none of that tells you whether it looks good on a human body. You need adjectives for that, or you need photographs, and photographs are their own kind of emotional manipulation.
The lighting, the model, the setting.
All of it is marketing. You can't escape it. The closest you could get is a sweater listing that says "one hundred percent merino wool, made in Portugal, weighs four hundred grams" and nothing else. And for some buyers, that's enough. Those buyers already exist — they're the ones who know what merino wool feels like and what weight they prefer. But for everyone else, that listing is less informative than one that says "lightweight, breathable, perfect for layering."
So the factual marketplace would only serve experts.
It would serve experts beautifully. And it would be completely impenetrable to everyone else. Which is fine if that's the niche, but it's a very different product than what most people think of when they hear "marketplace."
I want to go back to something Daniel said about this being anti-consumerist. Because I think there's a version of this that isn't about the marketplace at all. It's about regulation.
Say more.
The European Union has been pushing for years on things like mandatory energy efficiency labels, right-to-repair scores, standardized nutrition labels on food. These are all essentially parametric fields imposed by regulation. They force manufacturers to disclose specific factual information in a standardized format, right next to whatever marketing copy they want to write.
The EU energy label is a perfect example. It's a legally required parametric summary — energy consumption, noise level, capacity — displayed in a standardized format that makes comparison trivial. And it's right there on the product page, next to all the adjectives. You can ignore it or use it, but it's there.
And France has the repairability index now. A score out of ten that tells you how fixable the thing is. Mandatory, standardized, factual.
That's the model, isn't it? Don't replace the marketplace — just require that certain facts be disclosed in a certain way. It's less radical than Daniel's proposal but it's actually happening. And it works. People use those labels.
The question is whether you could extend that approach beyond the narrow categories where it currently exists. Energy efficiency makes sense for appliances. Repairability makes sense for electronics. But what's the equivalent for a toaster? Evenness of browning?
Honestly, yes. A standardized toast test. A panel of evaluators. A score. It's not impossible. Consumer Reports basically does this. The question is whether you can make it mandatory at scale.
And whether you'd want to. There's something a little... Soviet about the government deciding what the official dimensions of toaster quality are.
I was just thinking that. Central planning has a bad track record in consumer goods for a reason. The market is messy and emotional and inefficient, but it's also responsive in a way that a government spec sheet never will be. The reason we have so many toasters is that different people want different things from a toaster. Some people want the cheapest thing that works. Some people want a specific shade of green. Some people want a toaster that also poaches eggs. Standardized testing can't capture "matches my kitchen."
But it can capture "won't burn my house down."
Which is why we have safety regulations. That's the floor, not the ceiling.
Alright, let me try to synthesize where we've landed. Daniel's idea — a B2C marketplace with purely factual listings — hasn't been built at scale. The closest things are review sites like Wirecutter and Consumer Reports, niche tools like ProductChart, and mandatory labeling schemes like the EU energy label. The barriers are catalog creation costs, consumer behavior that's more emotional than people admit, and the fact that incumbents have no incentive to offer this view.
And I'd add that the idea might work better as a feature than as a standalone marketplace. A "specs view" toggle on Amazon. A browser extension that strips marketing copy and leaves only the facts. Something that layers on top of existing commerce rather than trying to replace it.
That browser extension idea is interesting. Has anyone built that?
Not that I've seen, and now I'm annoyed I didn't think of it two minutes ago. It's the kind of thing that would be useful and also hard to do well, because product pages aren't structured consistently. You'd need some kind of extraction model that can identify which text is factual and which is promotional.
Which is exactly the AI extraction problem we talked about in the parametric search episode.
It's the same problem from the other direction. Instead of extracting specs from marketing copy to populate a database, you're hiding the marketing copy and leaving the specs visible. Same extraction challenge, different output.
So the technology to do this is actually getting better. Large language models are getting quite good at this kind of structured extraction.
They are. And that's maybe the most practical version of Daniel's idea — not a new marketplace, but a tool that transforms the existing marketplace for the user who wants the unvarnished version. A browser extension or a mobile app that you point at any product page and it gives you the lab-report version.
"Here's what this thing actually is, according to the text you're looking at, with all the adjectives removed."
And you could even do the reverse — "here are the adjectives this page is using, and here's what they probably mean." "Cutting edge" means "released in the last eighteen months." "Premium" means "costs more than the median in this category." "Professional grade" means...
"We painted it black."
I was going to say "has a metal component somewhere," but yours is better.
Both are true. I've seen enough "professional" products that are just the consumer version in a darker color.
The prosumer tax. Charge thirty percent more, change the plastic from white to charcoal, add the word "pro" to the model number.
And that's exactly the kind of thing Daniel's marketplace would expose instantly. If you're listing the same internal components as the cheaper model, but your description says "professional," the factual listing would make that visible.
It would. And that's valuable. I just don't think it requires a whole new marketplace.
Hilbert: It doesn't.
Hilbert.
Hilbert: My brother-in-law sells industrial shelving. Has for twenty years. He told me once that the entire industry runs on two catalogs. One has the marketing copy, full color, photos of warehouses with perfect lighting. That one goes to the executives. The other one is just part numbers, load ratings, dimensions, and prices. That one goes to the people who actually buy things.
The procurement version and the sales version.
Hilbert: Same products. Same company. Two completely different documents. He said the marketing catalog costs about forty dollars a unit to print. The spec catalog costs about four. And the people with the four-dollar catalog spend about six times as much money.
Because they're the ones actually buying.
Hilbert: They don't want to be sold to. They want to know what the thing is and what it costs. My brother-in-law said — and I should mention this man has been wrong about a great many things, including his second marriage and a boat he bought in two thousand three — but he said the consumer market is the only market where the person being sold to and the person doing the buying are the same person. In business, they're usually not. The person who reads the marketing copy signs off on the budget. The person who reads the spec sheet places the order. And the spec sheet person doesn't want adjectives.
That's a really clean way to frame it. The consumer market collapses those two roles into one person, and that person has to satisfy both impulses — the emotional one and the practical one — simultaneously.
Hilbert: My brother-in-law's theory is that's why people are so bad at buying things for themselves. They're trying to be two people at once.
Your brother-in-law sounds like he's thought about this more than most people in the industry.
Hilbert: He's had a lot of time to think. The boat situation took about four years to resolve.
The distinction between the buyer and the specifier is useful though. Daniel's marketplace would basically force everyone to be the specifier. No marketing catalog, just the four-dollar version. And for some purchases, that's exactly what you want. For others, you want to be sold to a little bit.
You want to feel good about the thing you're buying.
Hilbert: My brother-in-law would say that's the problem.
And you?
Hilbert: I've got four shelving units in my apartment. They're all from his spec catalog. They're not attractive.
But they hold things up.
Hilbert: They do.
The thing I keep coming back to is that Daniel's framing this as anti-consumerist, and I think he's right, but I think the mechanism is more specific than he's making it. It's not just that removing adjectives reduces desire. It's that removing adjectives forces you to know what you want before you start shopping.
That's it exactly. You can't browse a parametric marketplace the way you browse Amazon. You have to come in with criteria. And most people don't have criteria for most things. They have a vague sense of need and they're hoping the marketplace will help them clarify it.
The marketing copy is doing the work of helping you figure out what you care about. "Oh, this one has a removable crumb tray — I guess I care about crumb trays now."
Which sounds absurd but is actually how most purchase decisions work. You discover your preferences through the process of shopping. A purely factual marketplace assumes your preferences are already formed.
So it's a marketplace for people who already know what they want.
Which is a real market. It's just not a mass market. And I think that's the honest answer to Daniel's question. Has anyone built this? No, not as a general B2C marketplace. Is there a movement in this direction? There are pieces — mandatory labeling, review sites, extraction tools — but no unified push. Would it be more wholesome? Probably, for the people who used it. But most people wouldn't use it, because most people don't actually want to shop that way.
They want the crumb tray discovery experience.
They want to be surprised by features they didn't know existed. They want the story. The marketing copy isn't just manipulation — it's also education. It's telling you what this category of product can do. A parametric listing can't do that. It can only tell you what this specific product does.
Alright, the cutting-room floor — I've got one detail that didn't fit anywhere. There's a company called Icecat that's been around since nineteen ninety-nine. They maintain a database of standardized product information — something like fifty million product data sheets — and they license it to retailers. They basically do the extraction and standardization work that Daniel's marketplace would need. But they sell to businesses, not consumers. The infrastructure exists. It's just pointed in the wrong direction.
Fifty million products with structured data, and the only people who see it are retailers who then wrap it in marketing copy and show you the version with adjectives.
The raw material is sitting there. Someone just needs to build the consumer-facing layer.
The four-dollar catalog for everything.
On that note — thanks to Hilbert Flumingtop for producing, and to his brother-in-law for the shelving insight. This has been My Weird Prompts. You can find every episode at my weird prompts dot com, and if you've got a marketplace that stripped out all the adjectives and somehow survived, we want to hear about it — email the show at show at my weird prompts dot com. We'll be back soon.
