← Documents
Startup pressure test

Sponic Gardens website pressure test

A startup-pressure-test pass for sponicgardens.com and the go-to-market page. This uses the local Codex startup-pressure-test skill: real users, painful problems, current behavior, manual traction, and the smallest test that proves or kills the idea.

Prepared 2026-05-05 · Targets: /, /gtm, and canonical source /gtm.html · Method: installed skill workflow, live-page fetch, and repo-source review.
Contents
  1. How to use the pressure-test skill
  2. Verdict and scorecard
  3. Core assumption
  4. Fatal flaws
  5. First customers and MVP
  6. Website fixes
  7. Rerun checklist

How to use the pressure-test skill

The skill is now installed locally at /Users/rahulio/.codex/skills/startup-pressure-test. It came from Greg Brockman’s repost of the open-source Codex startup pressure-test skill. Install on another machine with:

npx --yes codex-startup-pressure-test-skill@latest

After installing, restart Codex so the skill is discoverable, then run:

Use the startup pressure test skill.

Pressure-test this startup idea and use these pages as the current articulation:
- https://sponicgardens.com/
- https://sponicgardens.com/gtm
- https://sponicgardens.com/gtm.html as a fallback/canonical check

Idea:
Sponic Gardens is an AI-native social space and urban growing venue in Warsaw:
plants, wellness, food, music, maker infrastructure, and community commerce
run by an AI operating system that optimizes member satisfaction over time.

Target customers:
- founding members who want healthier social spaces
- Warsaw venue/land partners
- investors backing the first pilot
- early builders/operators

Return:
- verdict
- scorecard
- core assumption
- fatal flaws
- problem reality
- competition
- first 10 customers
- 2-week MVP
- exact website fixes
Workflow note. The skill says to test real behavior, not compliments. Use this page as the working snapshot, then replace the result sections after each rerun.

Verdict and scorecard

Verdict: Pivot required, but not away from Sponic. Pivot the wedge. The full AI-native urban growing social space is vivid, differentiated, and fundable as a world, but it is too broad to validate as one startup hypothesis. The first company should be a recurring paid Warsaw Sunday social ritual that proves people want healthier community spaces built around plants, wellness, food, and guided connection.
3/5 Pain intensity

Social isolation and wellness fatigue are real, but the page needs evidence that this exact format is a must-have.

2/5 Buyer clarity

Founding members, venue partners, investors, and builders are all being sold at once. The first buyer is blurry.

3/5 Urgency

Warsaw 2026 gives a clock, but the why-now argument should be made explicit.

4/5 Differentiation

AI-native physical social space plus growing infrastructure is memorable and hard to confuse with another SaaS pitch.

4/5 Speed to validate

A Sunday prototype can test demand before building the full venue.

3/5 Founder advantage

The ambition feels founder-led, but the site needs more proof of operational ability, Warsaw access, and unfair network.

Core assumption

People in Warsaw will repeatedly pay, attend, and bring friends to a structured plant-wellness-social Sunday experience before the permanent AI venue exists.

Why this matters: If this assumption is true, the warehouse-greenhouse vision has a wedge. If it is false, better sensors, more amenities, and more AI will not save the model.

Fatal flaws

RiskSeverityWhy it mattersFast test
The product is a beautiful world, not a painful wedge High People may love the idea and still choose brunch, sauna, gym, bar, coworking event, or staying home. Sell 30 paid seats to a no-build Sunday pilot. No free interest list as proof.
AI is over-positioned relative to what members buy High Early members probably buy belonging, novelty, health, and taste. "AI OS optimizing satisfaction" may sound abstract or creepy before trust exists. Run two event pages: AI-forward versus experience-forward. Compare paid reservations.
Venue complexity arrives too early High Real estate, permits, food, wellness safety, staffing, sensors, GDPR, and community ops are each hard. Bundled together, they become lethal. Use borrowed or partner space, manual ops, portable growing stations, and one weekly ritual.

First customers and MVP

Problem reality

Competition

First 10 customers

  1. Recruit from specific Warsaw circles: expat/founder groups, yoga/sauna communities, urban gardening people, makers, creative operators, and remote workers. Ask for a paid Sunday seat, not feedback.
  2. Start with 30 direct invites to get 10 paid attendees. Price it enough to create truth, for example EUR 20-40 equivalent, even if subsidized.
  3. After the first event, only count success if at least five attendees ask for the next date or bring a friend. Compliments do not count.

2-week MVP

Website fixes

Homepage: https://sponicgardens.com/

GTM: https://sponicgardens.com/gtm and /gtm.html

Priority fixes

  1. Fix the homepage tagline typo: change get high on your own supplai to the intended phrase or replace it with a clearer public-facing tagline.
  2. Replace the first hero explanation: use Sponic Gardens is a Warsaw social space prototype where members meet through plants, wellness, food, music, and guided community rituals.
  3. Move the AI operating system below the human promise: lead with the member outcome; make AI the mechanism.
  4. Add a First Sunday Pilot block: date, city, capacity, price/deposit, what happens, and what the team is testing.
  5. Add hard validation targets: paid reservations, show rate, repeat intent, referrals, venue partner interest, and event margin.
  6. Split CTAs by audience: Join founding Sundays, Offer a venue, Request investor deck, and Help build.
  7. Normalize GTM URLs: choose /gtm or /gtm.html as canonical. Redirect the other path and align sitemap, Open Graph URL, nav, docs, and internal links.
  8. Add "What we need now" to the GTM page: venue partner, 30 founding guests, five volunteers/operators, and pilot capital.
  9. Add "What we are not building yet": full spa, full AI OS, marketplace, app, and permanent site. This makes the plan more credible.
  10. Add operator proof: named team roles, Warsaw partner pipeline, rough unit economics, safety/compliance checklist, and exact next event milestone.

Rerun checklist

After edits, rerun the pressure test and record the date, tested commit, and public URLs.

# Fast route/header check
curl -I https://sponicgardens.com/
curl -I https://sponicgardens.com/gtm
curl -I https://sponicgardens.com/gtm.html

# Visual smoke check, if using a browser test runner
npx playwright screenshot https://sponicgardens.com/ /tmp/sponic-home.png
npx playwright screenshot https://sponicgardens.com/gtm /tmp/sponic-gtm.png
Open item: after changing the public pages, rerun the skill from a fresh Codex session so it loads $startup-pressure-test directly.