Launch Copy Formulas: What the Top Launches on aat.ee Name Themselves
We looked at the highest-voted launches on aat.ee and found their names follow five patterns. A working taxonomy of launch copy — drawn from real projects, with the anti-patterns that cost clicks.

Before anyone reads your launch description, they read your name — usually in a list of thirty other names, each fighting for the same two seconds. We went through the highest-voted launches on aat.ee to see whether the top of the list shares a copy pattern. It does. Not one formula — five, plus a handful of anti-patterns that reliably appear at the bottom.
The examples below are real launches from the platform. Steal the structures, not the names.
Formula 1: What it is, stated as a noun
The most common pattern in the top ranks is aggressively literal: combine what the product touches with what it does, and stop.
- Photo to Video AI Free — the medium, the transformation, the price.
- Image Upscaler 4K — the input, the action, the spec.
- True Cost of an Employee Calculator — the question the buyer is already asking.
These names are boring by design. A buyer skimming a launch feed does not decode cleverness; a name that reads like a search query gets understood at a glance, which is exactly the unit of attention you get.
Formula 2: The compound mark
Take two short morphemes and fuse them into a brand: TextileGen, RacoonDR, WallPreview, UPCgen, vidflux. The first half names the domain, the second names the action or output.
The failure mode of this pattern is fusion that hides the domain — a name like Meridian or Karpo is memorable after you know what the product does, but in a launch list it spends its two seconds teaching pronunciation instead of meaning. If your brand is abstract, your tagline must carry the entire literal load; that is a real cost, not a stylistic choice.
Formula 3: The qualified audience
Name the buyer, not the category: IELTS Writing Checker (test-takers), Saparo AI Shopping Agent (shoppers), Ask My Wardrobe (wardrobe owners). The narrower the stated audience, the stronger the "this is for me" reflex from exactly the visitors who will convert — and the weaker reaction from everyone else is a feature, not a bug. Launch feeds reward products that are obviously for someone.
Formula 4: The stated outcome
SEOFarmer promises growth. Skippr AI promises skipping. Gas Deck promises a dashboard for gas. The name takes a side about what success looks like, which doubles as positioning: it tells the reader which competitor this product is not.
The trap is promising an outcome the product cannot prove in thirty seconds. "Optimise your marketing spend" works for a calculator because the proof is the calculator working in front of you. The same claim on a consultative product with no visible mechanism reads as noise.
Formula 5: Voice as a differentiator
A minority of top launches break every rule on purpose: just f*ing send it** reads as a personality statement, and in a list of literal names the contrast *is* the memorability. This only works when the product's own voice is irreverent everywhere else — profile, copy, and name saying three different things is the actual anti-pattern.
The anti-patterns, from the same data
- "AI-powered platform for X" — the most common construction in the bottom half. It describes a technology instead of an outcome, and every launch in the feed says AI now.
- Two abstractions in one name. A fused brand whose halves both require explanation is a riddle, and lists are not the venue for riddles.
- The name that needs a subtitle to disambiguate — the subtitle exists to add information, not to rescue an unclear name.
Stress-test your launch name in four questions
- The glance test: would someone scrolling a list know the category in two seconds?
- The audience test: does the name repel the wrong visitor at least a little?
- The outcome test: does it promise something a thirty-second demo can prove?
- The search test: would anyone ever type these words when looking for the product?
Names that pass three of four tend to hold up on a launch feed. Submit yours and find out how it performs against the current week's list — the top launches are the benchmark you are writing for.