gate 02 · the lab · graveyard
Killed Ideas
Ideas that were cut, paused permanently or deliberately not pursued. Not because thinking was wasted — because the signal was weak, the scope was wrong, or the market did not justify the effort.
system: weak signal / idea terminated
Killing an idea is not failure.
It is a filter.
- Some ideas are killed because the market is too crowded.
- Some are killed because the scope is insane.
- Some are killed because distribution is too weak.
- Some are killed because they are interesting, but not worth building now.
The graveyard
killed idea = a real archive entry · discarded variant = a direction cut from a living idea- #016killed ideakilled
Social Echo
A fully decentralised social media scheduler that cross-posts using personal servers.
- why killed
- Too complex, too much technical maintenance, too dependent on platform rules and too hard to explain to normal users.
- lesson learned
- Decentralisation is not a selling point unless the buyer already feels the pain strongly.
Scope too complexbetter direction → A simple browser extension or local-first content helper for one specific platform.
full record - V-01discarded variantkilled
100 Chrome Extensions Sprint
A plan to launch dozens or hundreds of tiny Chrome extensions and monetise them through support, ads or premium upgrades.
- why killed
- It creates maintenance chaos before any distribution signal exists. Quantity without validation becomes clutter.
- lesson learned
- Build three focused extensions, measure installs and retention, then decide. Do not build one hundred random tools.
Strategy too unfocusedbetter direction → Chrome Extension Micro-Lab with 3 controlled experiments.
- V-02discarded variantkilled
Generic PDF Tools Site
A broad PDF tools website competing directly with every generic PDF utility site.
- why killed
- Too generic, too crowded and too hard to differentiate without a wedge.
- lesson learned
- Privacy-first, local-first or workflow-specific positioning is necessary. “PDF tools” alone is not enough.
Too genericbetter direction → LeafDeck, with a sharper privacy and document-workflow angle.
- V-03discarded variantkilled
Generic AI Logo Generator
A basic AI logo generator that creates static logos from prompts.
- why killed
- The market is crowded and the output risks feeling cheap, generic and legally unclear.
- lesson learned
- Static logo generation is not enough. A sharper wedge is needed.
Crowded marketbetter direction → LogoMotion AI, focused on simple logo + motion asset packages.
- V-04discarded variantkilled
Hardware-first StoryGlow
A physical night light that reveals a family memory every evening.
- why killed
- Starting with hardware would create cost, production, support and logistics problems before proving the ritual matters.
- lesson learned
- Validate the emotional ritual first with a web-based or printable experience.
Wrong MVPbetter direction → StoryGlow as a web-first nightly family memory ritual.
- V-05discarded variantkilled
All-in-One EU Tender Platform
A platform matching every company with every possible EU, national and local tender.
- why killed
- Too broad, too complex and too hard to make accurate for every sector.
- lesson learned
- Tender matching only works if verticalised by industry, geography or buyer type.
Scope too broadbetter direction → EU Tender Matcher for one specific niche first.
- V-06discarded variantkilled
Generic Recipe Social App
A recipe app with social features, discovery, public sharing and a large recipe community.
- why killed
- Social and discovery features create complexity before solving the core personal recipe problem.
- lesson learned
- Start with private recipe storage, ingredient scaling and local-first search.
Feature creepbetter direction → RecipeVault Local.
- V-07discarded variantkilled
Generic Tarot App
A standard tarot app with daily cards and generic readings.
- why killed
- The market is crowded, retention is hard and generic readings are not enough to stand out.
- lesson learned
- Atmosphere, ritual and distribution matter more than the card-drawing feature itself.
Crowded marketbetter direction → A tiny web MVP with one distinctive paid reading flow before building a full app.
- V-08discarded variantkilled
Fully Custom Local Business Agency
A broad custom agency doing websites, photography, Google Maps, menus, branding, AI content and marketing for any local business.
- why killed
- Too broad and too exposed to custom client work, unclear packages and scope creep.
- lesson learned
- Local services need productised offers with fixed deliverables, fixed prices and clear boundaries.
Scope creepbetter direction → Tagra Studios Local Business Toolkit with strict packages.
- V-09discarded variantkilled
Full AR Balcony Gardening App
A full AR app where users place plants and watch them grow virtually on their balcony over time.
- why killed
- True AR adds complexity before proving users care about previewing plants.
- lesson learned
- Start with photo-based mockups and plant-care reminders before building AR.
Wrong technical starting pointbetter direction → GrowGhost as a photo and mockup-first plant preview.
Nothing buried matches these filters.
Kill criteria
any one of these is enough to stop- 01No clear buyer
- 02Weak distribution path
- 03Too expensive to maintain
- 04Too complex for a fast MVP
- 05Too dependent on platforms I do not control
- 06Too generic compared to existing tools
- 07Interesting, but not urgent enough
- 08Better as a feature than a business
Think one of these should not be dead?
If you know the market, have distribution, or can explain why the idea deserves another test, send a clear argument.