Building An AI Marketing Strategy: The 94-Revision System
- Tony Cooper
- Jul 11
- 5 min read
Updated: Jul 25
Have you considered using AI to find the low hanging fruit in your marketing data?
At the most basic level you can import your Google Analytics CSV to ChatGPT and ask it to find opportunities. But what if you applied a marketing strategy that built on those opportunities that AI discovered for you?
That's exactly what we did. And after 94 revisions, we developed a systematic approach that any business can follow.
Try this now: Export your last 3 months of Google Analytics data and ask ChatGPT: "What pages have high impressions but low clicks?" You'll likely find 5-10 quick wins in under 5 minutes.
What 94 Revisions Taught Us About AI Marketing
When people hear "94 revisions," they think we're obsessive. But here's what those revisions uncovered:
7,108 monthly search impressions getting only 2 clicks
37 keywords ranking on page 2 that competitors were monetising
£3,960 in monthly recurring revenue from systematic improvements we were missing
More importantly, we learned exactly when revisions add value and when they're just perfectionism.
The Truth About AI Content Creation
Here's the truth about AI marketing: if you don't set boundaries and focus on the subject, you'll end up with worthless porridge.
Problem 1:
AI Has Recency Bias LLMs train heavily on Stack Overflow and coding forums. Ask about React.js? Brilliant. Ask about converting visitors into customers? Generic waffle.
Problem 2:
AI Agrees With Everything Watch this:
You: "Should I use WordPress?"
AI: "Yes! WordPress is perfect!"
You: "Actually, maybe Shopify?"
AI: "You're right! Shopify is ideal!"
Problem 3: No Business Context AI doesn't know your ideal customer or your business goals. It can't connect content to revenue unless you teach it.
Quick Fix: Build resistance into your prompts. Ask AI to challenge your assumptions, not just agree with them.
Setting Up Your AI Marketing Assistant
You don't need complex systems to start. Here's our exact setup:
Basic Requirements:
GitHub repository for your business documentation
Clear tone of voice guidelines
Actual client examples and case studies
Revenue targets and business goals
What Makes It Work: Our AI assistant Claudia has access to our complete business documentation.
How we talk to customers
Real client success stories
Specific revenue targets
Common problems and solutions
Your Action Step: Create one document titled "How We Talk to Customers" with 5 example phrases you use and 5 you'd never say. That's your starting point.
The 4-Phase Development Process
Phase 1: Intelligence Gathering (Week 1)
What You'll Need: Google Analytics, Search Console, 10 hours
Investment: 50 revisions to perfect the brief
Start by mapping:
Your current traffic and conversion data
Competitor content gaps
Customer pain points from support tickets
Keywords you rank for but don't monetise
Common Mistake: Jumping straight to content creation without understanding your position.
Try This: List your top 10 competitors. Run their URLs through any SEO tool. Find 5 topics they all cover that you don't.
Phase 2: Strategic Planning (Week 2)
What You'll Need: Content calendar, clear business goals
Investment: 31 revisions to nail the structure
Create two content phases:
Education Phase: Build trust with helpful content
Conversion Phase: Demonstrate your expertise
The Stakeholder Challenge: Getting buy-in for quality content is tough. The accounts team says it's too expensive. Marketing worries about deadlines.
Time for some stakeholder physics: Show them that 3-revision content = £0 revenue, while 94-revision content = £990 monthly recurring. The invoice reality wins arguments.
Phase 3: Topic Selection (Week 3)
What You'll Need: Keyword research tool, revenue projections Investment: 19 revisions per topic
For each topic, map:
Search volume and commercial intent
Competition level
Revenue potential
Your unique angle
The IKEA Test: Doing 3-revision content is like assembling IKEA furniture with half the screws missing. Sure, it might stand up initially, but wait until someone actually sits on it. Your content needs all its screws tightened properly if you want it to support actual business weight.
Phase 4: Content Creation (Weeks 4-8)
What You'll Need: 4-6 hours per piece, quality standards Investment: Up to 94 revisions for cornerstone content.
Revision Breakdown:
Revisions 1-10: Structure and basic content
Revisions 11-25: Voice and personality
Revisions 26-40: Strategic alignment
Revisions 41-60: Evidence and examples
Revisions 61-75: Conversion optimisation
Revisions 76-94: Final polish
When to Stop: When adding more revisions doesn't improve business impact. We learned this at revision 71 when we unlocked the £3,960 MRR opportunity.
Your Implementation Roadmap
Do This Today (30 minutes)
Export your Google Analytics data
Upload to ChatGPT with this prompt: "Find pages with high impressions but low CTR"
Fix the top 3 page titles you find
Do This This Week (4 hours)
Create your "How We Talk to Customers" document
Map your top 10 competitor content gaps
Choose 5 topics that match your expertise
Do This This Month (20 hours)
Write your first strategic content piece
Track your revision count and what each adds
Measure impact after 30 days
The Revenue Reality
Time for some revenue physics - quality content compounds. That article you spend 6 hours perfecting generates leads for years. The quick post you knocked out in 30 minutes? Forgotten tomorrow.
Our Actual Results:
Month 1: 12 revisions average = 0 clients
Month 3: 53 revisions average = 1 client
Month 6: 94 revisions average = 8 clients
Every additional revision adds approximately £13.50 in monthly value. Not because revisions are magic, but because they force you to improve what matters.
Common Questions
"Do I really need 94 revisions?" No. You need as many revisions as it takes to create content that generates revenue. For some pieces, that's 20. For cornerstone content, it might be 100.
"What if I don't have time?" Start with your highest-value pages. One great piece beats ten mediocre ones.
"How do I track ROI?" Simple: Track enquiries before and after. If content doesn't drive enquiries, revise or remove it.
Your Next Steps
You have three options:
Option 1: DIY Implementation Follow the roadmap above. Expect results in 3-6 months.
Option 2: Accelerated Learning Get our templates and frameworks. Cut your learning curve in half.
Option 3: Done-With-You Implementation We guide you through the process. You learn while building.
Want Our Help?
We can handle 10 audit requests per month while maintaining quality standards.
Get Started:
Phone: 01952 407599
Ask for the Marketing Data Audit. We'll show you exactly what opportunities are hiding in your analytics and how to systematically capture them.
P.S. This article? 51 revisions. Not 94 like our earlier content. Why? Because our AI agents are getting better.
Claudia now starts closer to the target, understanding our voice and standards from the beginning. This is what systematic improvement looks like - when your 41st revision achieves what used to take 94.
