On this page
In short
- Market asks one question: can we reach them repeatably, and will they pay?
- A product that works but looks unfinished loses to one that looks finished. Interface quality is part of distribution.
- Every launch is packaged the same way, so each release adds to the last instead of starting from zero.
- AI search and assistants now describe products to buyers. If they can't read you, they describe you wrong.
Most teams treat distribution as the thing that happens after the product is done. By the time they get to it, the product has been described three different ways, the website still says what it said at the start, and every launch is a one-off. The Market stage is where a product people keep using becomes one the right people can find, again and again.
01What is the Market stage for?
Market is the fifth stage of Stuck-to-Scale. Its question: can we reach them repeatably, and will they pay?
"Repeatably" is the word that matters. A launch that lands once is luck. A way of reaching buyers that you can use again, and that gets cheaper as you use it, is a market. "Will they pay" is the other half: the price, what's included, and the moment in the sale when the buyer first meets the number. Pricing logic is strategy work, but buyers meet it through the page, the demo and the proposal, so it's tested here.
Market comes after PMF on purpose. Reach multiplies whatever you have, so it should multiply something people keep.
02What does being stuck at Market look like?
Two patterns come up again and again in the products we've looked at: distribution treated as a problem for after launch, and interface quality holding back a product whose technology is strong. Both show up as the signals below.
| Signal | What it looks like | What it usually means |
|---|---|---|
| Every launch starts from zero | A new page, message and look each time | Nothing is packaged, so nothing adds up |
| The product looks unfinished | Strong technology, weak interface | The interface is costing sales the technology earned |
| Every call gets a new pitch | Sales describes it differently each time | There's no single description to repeat |
| Campaigns that work once | Spend rises; results don't repeat | Reach is being bought, not built |
| AI search gets you wrong | Assistants describe you vaguely, or confuse you with someone else | Your facts aren't written anywhere a machine can read them |
03Who is it for, in one sentence?
The first thing we fix is the description: what the product is, who it's for and why it's different, in the buyer's own vocabulary. Then it's worded the same everywhere: the site, the profiles, the deck, the press kit and the product's own copy.
The same words, not similar ones. People remember what they hear repeated, and machines depend on it. An assistant that finds three different descriptions of you has to guess which one is true.
Decision · one description of Graylemon, everywhere
- What we chose: one sentence for the studio, "an AI-native product and venture studio based in Ahmedabad, India", used word for word on the About page, in the footer and in the press kit.
- What we gave up: variety. The same sentence gets tiresome to write.
- Why: our old website described us several ways, and some of them were wrong. Buyers and AI systems both took each version literally. We tore it down in Teardown: our old website.
- What it fixed: anyone describing us now has one sentence to quote.
04How do we make every launch count?
By packaging it the same way every time. In continuous launch support, each feature release gets a demo video, short cuts for paid social, launch visuals for each channel, and updates to the landing page sections it affects. A release that ships without its package reaches the people already using the product. A packaged one reaches the people who aren't yet.
The product itself is part of the package. One of our convictions is that brand and product are one foundation: your interface is your marketing, your docs are your SEO, and your demo is your deck. A strong product with an unfinished interface loses, in the first minute, to a weaker one that looks finished.
Our own ventures are our first creative clients. otlo's launch demo and apprn's product demo are being made with the same service we sell, and they'll be labelled as our own work when they're published.
Insight
A release without a launch package only reaches the people who already use the product.
05Can AI search describe you correctly?
AI search and assistants now sit between products and buyers. They answer "what's a good tool for this?" from whatever they can read about you, so how you're described and documented is read, and repeated, by machines. We design for them as carefully as for people.
Machine Legibility sets out four layers, each built on the one below:
- One description, worded the same everywhere.
- Crawlable content: pages that can be read without running scripts.
- Machine-readable documentation: structured data and a sitemap that describe only what's really on the page.
- Genuine mentions: other people describing you, in their own words.
We rebuilt our own site on the first three layers; the fourth can only be earned. Nobody can promise that an assistant will cite you. What you can do is remove every reason for it to leave you out, or to get you wrong.
If a machine can't read what you are, it guesses. It usually guesses generic.
06Will they pay, and how do you find out?
By watching where buyers stop. When interest drops at the price, the number is only one suspect. The other is the sequence: when the buyer meets the price, what they've seen before it, and who else has to approve it. Sequence problems often look like conversion problems, and they're far cheaper to fix than a price.
So at Market we test the whole path a buyer takes, not just the headline: the page that explains the product, the demo that proves it, the proposal that prices it, and the person on the buyer's side who has to say yes.
When the right buyers can be reached again and again, and they pay, one question is left: does it compound without us pushing it? That's the last stage, Grow.
07Questions founders ask about Market
When should we start marketing?
In earnest, once the right people keep using the product. Before that, spend on learning who stays and why. The groundwork, one description and a site that tells the truth, can start as early as Define.
Do we need a brand before the Market stage?
You need a position before Market, and a brand system during it. A logo isn't a brand. A brand is the set of rules that keeps everything you publish consistent, with someone who owns it.
How do we show up in AI search?
Make your facts easy to read and consistent: one description, pages that don't need scripts to read, documentation machines can parse, and genuine mentions in other people's words. Nobody can promise a citation; you can remove the reasons you'd be left out.
What does Graylemon do at the Market stage?
Positioning and verbal identity, product and interface design, a launch package for every release, and the work that makes the product legible to AI search, with a named owner on your side for all of it.
Where this fits