About Prompt & Profit
You already use ChatGPT or Claude for some of your marketing. The question is whether you can rely on it — week after week, across briefs, drafts, distribution, lifecycle, and reporting — without the quality drifting and without your team turning into prompt cleanup. The Playbook is the operator’s manual that closes that gap.
The Playbook in one paragraph
This is a practitioner’s Playbook for applying large language models to the core jobs of a modern marketing team. It moves from prompt (the craft of asking) to practice (the workflows that make output reliable across people and weeks) to profit (the measurement and governance that keeps the work defensible to your finance team and your legal team).
Every chapter ends with prompt templates you can run, declared variables and stated failure modes included. Every claim is sourced. Every case names the company.
What is in it
Fifteen chapters across five parts.
- Part 1 — Foundations. The shift from “tool you sometimes use” to “augmentation you ship through.” Voice and evidence standards.
- Part 2 — Content & Creative. Long-form writing, social copy, brand voice consistency at team scale, visual brief generation.
- Part 3 — Strategy & Research. Audience research, competitor teardowns, positioning sprints, persona maintenance.
- Part 4 — Performance & Lifecycle. SEO content briefs, paid-search ad variants, lifecycle email programmes, win-back sequences, A/B test design.
- Part 5 — Operations & Governance. Measurement, review workflows, audit trails, the compliance posture you can defend in a vendor review.
Every chapter follows the same eight-section rhythm — opening hook, job-to-be-done, pain points, productivity gains, named case study, prompt templates, pro tips, common mistakes. You learn the rhythm once and work every chapter faster.
What you walk away with
- Sixty-plus prompt templates in YAML, each with a worked example, declared variables, and stated failure modes. Downloadable as runnable YAML inside the Premium bundle.
- A verified case-study register. Every case names the company, the date window, and the source. Anonymous “a major retailer” cases are not used. The public standards state the bar; the register itself reads inside the subscriber library.
- A measurement framework for AI-augmented marketing that reports productivity gain in numbers your finance team will accept — drafts per writer per week, time-to-publish, cost per qualified piece.
- A governance posture you can defend — review checkpoints, source-tier discipline, audit trails for AI-generated copy. Tested against the kinds of questions a procurement or legal team actually asks.
What it does not cover
This is not a primer on how large language models work internally. It is not a vendor tutorial — there is no “how to set up Jasper” chapter, no annotated screenshots that age in a quarter. It is not a generalist AI Playbook — every example is a marketing example. And it is not a hype Playbook — if a claim does not carry a Tier 1, 2, or 3 source by the evidence standard, it is not in the Playbook.
The Playbook also does not cover non-marketing applications of LLMs (engineering, customer support, finance), agentic systems beyond what a marketing team can ship today, or model fine-tuning. Those are different Playbooks.
The Marketing Playbook, and what comes after
The Marketing Playbook is the first in a series. Every subsequent Playbook follows the same operator’s-manual format — public standards, named-source evidence, prompt templates with declared variables and stated failure modes.
The standards tell you what you would be holding any of them to.
How to use it
Most chapters are independently usable — you can jump straight to Chapter 5 (Email Marketing) if that is the work you ship next week. Two chapters are prerequisites for everything else: Chapter 1 sets the worldview and the voice the Playbook runs on; Chapter 2 establishes the measurement framework that every productivity claim refers back to. Read those first; then work in any order.
Free samples of Chapter 1 and Chapter 3 are on the site, no email gate.
Shape and scope
| Property | Value |
|---|---|
| Format | Web-first Playbook with downloadable working files |
| Scope | About 95,000 words across fifteen chapters |
| Working time | A chapter takes 35–45 minutes to read; the prompt templates are the artefact you keep using afterwards |
| Prompt templates | 60+, with 200+ worked-example variations |
| Case studies | 40+ named, each clearing the case study standards; the verified register reads inside the subscriber library |
| Edition | First — the v1.x update track ships substantive corrections within sixty days of report |