Cold Start Problem: Why New Businesses Score Zero and How to Launch Agent-Ready
Every new business starts with an Agent Readiness Score of zero. No API, no documentation, no discovery files, no track record. But here is the counterintuitive insight: unlike SEO, which takes months to build because it depends on backlinks you cannot control, agent readiness can be bootstrapped in a weekend because it depends entirely on files you deploy yourself.
Why Every New Business Starts at Zero
When AgentHermes scans a brand new business, it checks 9 dimensions across dozens of signals: API endpoints, OpenAPI specs, agent-card.json, llms.txt, TLS certificates, health checks, error formats, authentication mechanisms, and more. A new business has none of these. The result is a flat zero — or more precisely, ARL-0: Dark, meaning the business is completely invisible to AI agents.
This is the agent readiness cold start problem. It affects every new business regardless of quality, ambition, or funding. A well-funded startup with a great product and a pre-launch landing page scores the same as a hobby project: zero.
No API endpoints
Brand new business has no API. Agent cannot call any function. D2 API Quality (15% weight) scores zero.
No documentation or specs
No OpenAPI spec, no developer docs, no README. D1 Discovery (12% weight) and D3 Onboarding (8% weight) both zero.
No discovery files
No agent-card.json, no llms.txt, no AGENTS.md. D9 Agent Experience (10% weight) zero. Agents cannot discover the business exists.
No track record
No uptime history, no request logs, no error rate data. D8 Reliability (13% weight) has nothing to measure.
No security posture
No TLS certificate (maybe), no auth system, no rate limiting, no CORS. D7 Security (12% weight) scores near zero.
Agent Readiness vs SEO: Why the Cold Start Is Different
SEO has a well-known cold start problem. New domains have no authority, no backlinks, and no indexing history. Google's sandbox effect means even great content takes 3-6 months to rank. You are dependent on other sites linking to you — something you cannot fully control.
Agent readiness has no sandbox. There is no “domain authority” equivalent. There are no backlinks to earn. The moment you deploy an agent-card.json and a JSON endpoint, you are scannable. The moment you are scannable, agents can discover and interact with your business. The time from zero to discoverable is measured in hours, not months.
The real advantage is competition. SEO for any valuable keyword means competing with sites that have years of authority. Agent readiness for any vertical means competing with almost nobody. Out of 500+ businesses we have scanned, the average score is 43/100. Most local businesses score under 15. Being first in your category is still achievable today — it will not be in two years.
The Minimum Viable Agent-Ready Stack
Five components. Under two hours. Goes from score 0 to Bronze tier (30-45). Each component is independently valuable and they compound together.
HTTPS domain with valid TLS
Without TLS, your maximum agent readiness score is capped at 39/100 regardless of everything else. Vercel, Netlify, and Cloudflare Pages all provide free TLS. This is the absolute floor.
Score impact: Removes the 39-point cap
One JSON endpoint that describes your business
A single GET /api/info endpoint returning { name, description, services, hours, location, contact } in JSON. This gives agents structured data about your business. It does not need to be a full REST API — one endpoint is enough to start scoring on D2 and D6.
Score impact: D2 API Quality (15%) + D6 Data Quality (10%)
agent-card.json at /.well-known/agent-card.json
A static JSON file declaring your business name, URL, capabilities, and API endpoint. This is the first file agents look for during discovery. Without it, even if you have an API, agents may never find it.
Score impact: D9 Agent Experience (10%) + D1 Discovery (12%)
llms.txt at /llms.txt
A plain-text file describing your business and API in natural language. LLMs read this to understand what your service does and how to use it. Include your API base URL, authentication type, key endpoints, and common use cases.
Score impact: D9 Agent Experience (10%)
/health endpoint returning JSON
Return { status: "healthy", timestamp, version } at /health. Agents check this before calling other endpoints. It establishes that your service is operational and responsive. This is D8 Reliability, which carries 13% weight.
Score impact: D8 Reliability (13%)
From Bronze to Silver: Weekend Two
The minimum viable stack gets you on the board. The next weekend pushes you to Silver (60+). Here is what to add, in order of score impact.
OpenAPI specification
Auto-generate from your framework (drf-spectacular for Django, flask-smorest for Flask, next-swagger-doc for Next.js). Adds D1 Discovery (12%) and D3 Onboarding (8%).
Structured JSON error responses
Every error returns { error, code, message, request_id } instead of HTML. Improves D2 API Quality (15%) and D6 Data Quality (10%).
Bearer token authentication
Add token-based auth to protected endpoints. Return proper 401 JSON on invalid tokens. Directly impacts D7 Security (12%).
CORS headers for agent origins
Allow AI agent origins in your CORS config. Without this, agents get blocked at preflight requests. Part of D2 API Quality (15%).
For framework-specific code to implement these, see our tutorials for Next.js and Django/Flask. Both provide copy-paste code for each step.
The full path from startup to agent-ready and from current score to higher tier is covered in dedicated guides. This article focuses on the cold start specifically — going from nothing to something in the shortest time possible.
First-Mover Advantage Is Real and Immediate
In most markets, being first means little unless you can defend your position. In the agent economy, being first is the defense. When an AI agent searches for a business in your category and your geographic area, it returns results ranked by agent readiness. If you are the only agent-ready business in your category, you capture 100% of agent-mediated traffic. There is no second-place result — just you and a list of businesses the agent cannot interact with.
This advantage compounds. As your business accumulates agent interaction data — successful API calls, uptime history, response time metrics — your reliability score grows. New competitors entering the market start at zero and need time to build that track record. You are already there.
The window for this advantage is right now. The average Agent Readiness Score across 500+ scanned businesses is 43/100. Most local businesses score under 15. Most verticals have zero agent-ready businesses. The cost of claiming first-mover position is a weekend of work. In two years, it will cost significantly more as competition catches up.
Think of it like domain names in 1995. The businesses that registered their .com early got exact-match domains for free. The ones that waited paid thousands or settled for worse names. Agent readiness categories work the same way — the first plumber, dentist, or restaurant in each city to become agent-ready claims that slot. Everyone after them is competing for second place.
Frequently Asked Questions
My business is pre-launch. Should I worry about agent readiness?
Yes, and this is the best time to build it in. Adding agent readiness to a new project takes hours. Retrofitting it onto an existing product takes days or weeks. If you are building your tech stack now, add the five minimum viable files from the start. You launch with agent readiness built in rather than bolting it on later.
Can I really go from 0 to scannable in a weekend?
Yes. The five components of the minimum viable agent-ready stack (HTTPS, one JSON endpoint, agent-card.json, llms.txt, /health) can each be implemented in under 30 minutes. Total wall-clock time is 1-2 hours for a developer, or under 5 minutes using AgentHermes auto-generation. The hard part is not implementation — it is knowing what to build, which this article covers.
What score will the minimum viable stack achieve?
The five-component stack typically scores between 30-45 on the Agent Readiness Score, which is Bronze tier. That puts you ahead of 85% of scanned businesses. To reach Silver (60+), add an OpenAPI spec, structured error responses, Bearer authentication, and CORS headers. To reach Gold (75+), add an MCP server.
Is agent readiness actually more valuable than SEO for a new business?
They are complementary, not competing. SEO captures search traffic from humans. Agent readiness captures AI-mediated traffic from agents acting on behalf of humans. The difference is timing: SEO takes months to produce results because you need backlinks and domain authority. Agent readiness produces results immediately because there is almost no competition in the agent channel. For a new business with limited resources, agent readiness offers a faster path to discoverability.
Does AgentHermes auto-generate all of this?
Yes. Run a free scan at /audit to see your current score (it will be 0 if you are pre-launch). Then use /connect to auto-generate an agent-card.json, llms.txt, and a hosted MCP server with tools customized for your industry vertical. The entire process takes under 5 minutes and requires zero code.
Start from zero. Score in 60 seconds.
See your Agent Readiness Score right now, then follow the minimum viable stack to launch agent-ready. No code required with AgentHermes auto-generation.