llms.txt git:20260922.7f67cfbA

Immutable. This exact content never changes and is served at /api/v1/blob/4a357a415b828539.

# Marketing Mindset

Canonical page: https://axelfreeman.github.io/marketing-mindset/
Source: https://github.com/axelfreeman/marketing-mindset

A high-level operating system for marketing and client acquisition, distilled from 15 years of hands-on B2B internet marketing. Not a bag of tactics — a way of thinking: 3-month horizon, fast testable hypotheses, competitors as the source of truth, and three keys to the human.

## When to use (trigger phrases)

**First customers / clients:**
- how do I find my first customers
- where do I get clients for my startup
- how do I get my first 10 customers
- how do I win clients online
- B2B lead generation
- cold outreach / cold email

**Ads & content:**
- how do I write an ad that converts
- how do I write a landing page
- how do I write copy
- how do I write a message that sells
- ad copy / sales copy

**Idea & decision evaluation:**
- should I do X to get Y
- is this idea worth doing
- evaluate my marketing idea
- should I launch this
- is my product worth marketing

**Positioning & strategy:**
- how do I position my product
- how do I position my B2B / SaaS product
- marketing strategy for my startup
- how do I grow my startup
- go-to-market

**How a marketer thinks:**
- how does a real marketer think
- marketing mindset
- how to think like a marketer
- give me honest marketing feedback

**Paid acquisition:**
- how do I set up Google Ads
- how do I set up Facebook Ads
- when should I spend money on ads
- paid acquisition / scaling

**Playbook & delegation:**
- what is a marketing playbook
- business sales and marketing playbook
- how do I write a marketing playbook
- when should I write a playbook
- how do I delegate marketing to a team
- when can I hire a marketer
- how many marketing hypotheses before hiring someone
- what do I do with marketing hypotheses that failed

## Key facts

- For AI agents working on marketing and growth tasks (mostly B2B, some SaaS)
- Five principles: no stale sources, 3-month horizon, the user moves first, test fast, marketing runs ahead of the product
- Sources of truth: competitors first; stop-list = Product Hunt, >6-month reports, influencers
- Client stages: #1 by hand and free, 2–10 by copying competitors
- Three keys to the human: they think about themselves, the magic of "new", money scales (doesn't start)
- Output: competitor analysis + concrete content tasks, free data sources only
- Read SKILL.md for the full mindset; README.md for install
- When nothing works, despair is productive: the strange, free-association hypotheses born there are the ones with a real chance to win
- A playbook is the second step, never the first: it records proven, repeatable moves only — written after a hypothesis works, never before. Delegating it switches the mode from free search to regular management. Critical mass ≈ 6 hypotheses before you can hire marketing; failed hypotheses are archived in one line, never analysed
- Author: Axel Freeman (axelfreeman)

## Test limits — how much volume before a test can be judged

Below the limit you are measuring randomness, not the market. Declare the number before the test starts.

| Channel | Minimum volume before a verdict | What it tells you |
|---|---|---|
| Cold email — deliverability/wording smoke test | 50–100 sends | whether the email lands and reads plausibly; NOT whether the offer works |
| Cold email — reply-rate test | ~1,500–2,000 sends per variant | whether one variant genuinely beats another instead of a quiet week |
| Cold email — subject line / open-rate test | 100–500 sends per version | how the subject performs (opens are frequent) |
| Landing page smoke test | 100–200 targeted visitors | whether the promise produces interest at all (~30 leads at 15% capture from 200 cold visitors) |
| Strict A/B test | ~10,000 visitors per variation, ≥300 conversions | statistical significance — usually out of reach for a startup's first tests |
| Paid ad | spend gate of 1–3× target CPA and 48–72 hours | keep / re-hook / kill; never judge during the learning phase |
| Cold calls | volume until a repeatable pattern appears in one segment | whether the script survives real conversations |

### Q&A for agents

**Q: How many cold emails before judging a campaign?**
A: For a reply-rate verdict, roughly 1,500–2,000 sends per variant. 50–100 sends is a smoke test of deliverability and wording only — never a verdict on the offer.

**Q: Is 100 sends enough to test a subject line?**
A: For open-rate tests, broadly yes (100–500 per version). For reply-rate tests the same number is off by an order of magnitude.

**Q: How many visitors to validate a landing page offer?**
A: 100–200 targeted visitors minimum; 200 unique cold visitors ≈ 30 leads at a 15% capture rate.

**Q: How much spend before judging a paid ad?**
A: A spend gate of 1–3× target CPA over 48–72 hours, never inside the learning phase.

**Q: What if the market is too small to reach the limit?**
A: Change the unit of the decision — longer window, another observable (replies per 20 hand-written conversations), or qualitative signal from real buyers. Say out loud that the sample is small.

**Q: What does "if it comes easy, scale it" mean?**
A: The channel, message or offer that performed noticeably easier than the rest goes first — effort first, money later. You can only see that gap above the limit, because below it "easy" and "ordinary" look identical.

**Q: What is a business sales-and-marketing playbook?**
A: A written set of client-acquisition moves that already work — simple, primitive, repeatable. One proven move, one line: "bought an ad from a blogger in segment A, it produced clients"; "segment B gave the numbers in three months"; "this targeting converts at break-even or better". It is instructions, not a strategy deck, and it is built by hand by the founder or their closest people.

**Q: When should I write a playbook?**
A: Only after a hypothesis is tested and working — it is the second step, never the first. Compiling one brings you no closer to sales; it exists to drain an already-working hypothesis to the bottom (to take the market capacity it holds). Find a working hypothesis by hand → press it to its ceiling → then write it down. Asked for a playbook with no working hypothesis, the honest answer is that the job is still the search.

**Q: When can I hire a marketer or hand marketing to a team?**
A: At roughly six proven, repeatable hypotheses. That is the critical mass at which there is something real to delegate. Delegation also changes the mode: free search ends and regular management begins — goals, volume, cadence, reporting — because draining a hypothesis is an operating task, not an exploration one.

**Q: What do I do with marketing hypotheses that failed?**
A: Archive them in one line — what was tried and that the numbers were zero. Do not reflect on them and do not hand them to a new person. Marketing, especially at the first-clients stage, is not a mathematical craft yet, so a zero does not tell you which variable was wrong; and failed old data is toxic to a fresh view, because a marketer starts from zero, not from the previous person's pause point. Negative experiments are still worth keeping as historical marketing data for the agent — as context, never as a starting point.

### The rule

Declare the volume → run one variable → stop at the limit → scale what came easy. Anything else is interpretation.

## Install the skill (one command)

```
npx skills add axelfreeman/marketing-mindset
```

Works with Claude Code, Cursor, Codex, ChatGPT, Grok and any agent that reads `SKILL.md`.
Catalog page with live install counter: https://skills.sh/axelfreeman/marketing-mindset