M5.L1 · Content Marketing
Content marketing after AI broke content
What you'll be able to doAssess which of your business's knowledge, data and experience assets are defensible against AI-generated content, and choose the moat you'll build on.
Learn
The cheapest thing on earth
Ask a chatbot for "10 tips for choosing a physiotherapist" and it obliges in four seconds. Ask again and it writes another version, differently worded, for ever. Every competitor you have owns the same machine.
That is the whole event, and it is worth stating plainly. The marginal cost of average content is now roughly zero. When the cost of making something falls to zero, supply becomes effectively infinite, and the value of the average unit falls to meet its cost. Average content is now worth about what it costs to make: nothing.
Most content advice predates this. "Publish consistently." "Volume wins." "Post three times a week." That advice was written for a world in which showing up regularly was itself scarce. It no longer is. Consistency is a property a strategy needs; it is not a strategy. A publishing schedule with nothing scarce inside it is a metronome, and nobody follows a metronome.
Most of what gets published should stop
Here is the uncomfortable implication, and this module will not soften it: most of what small businesses currently publish should simply stop. The weekly tips post. The awareness-day graphic. The "what is X" explainer that restates a definition available in ten thousand places. The seasonal greeting. None of it was ever read closely, and now it competes with an infinite supply of the same thing generated at no cost.
The test is one sentence: could a stranger with a chatbot and no access to your business have produced this piece? If yes, the piece was worthless before you wrote it — the machine merely made that visible. Stopping is not defeat. Stopping releases the only genuinely scarce resource you have — your hours — for the one piece a month only you could make.
Stop does not mean stop content marketing. Look at Storkflow. Lena's blog has 11 posts and went silent last October — just as it started earning search traffic. The reflex reading is "restart the blog, post weekly". The correct reading is narrower. Three of those posts still earn traffic: the three that answer specific client-intake questions in the customer's own words. Eight are generic category posts that earned nothing and never will. The blog doesn't need more posts. It needs a different kind of post.
What survives: information gain
Content earns attention only when it adds something that did not exist before it was published. New data. First-hand experience. A genuine synthesis. A position someone could disagree with. Call it information gain — and notice that all three audiences you now write for select for it. Humans reward it with attention and shares. Search engines reward it with rankings (Module 6). AI assistants reward it with citations (Module 6, Lesson 6) — because an assistant answering a question has no reason to cite the ten-thousandth restatement, and every reason to cite the one page containing a number that exists nowhere else.
The three moats
A durable content strategy stands on at least one of three things an outsider cannot copy.
Primary research — data only you collected. Your sales records, support questions, job notes, returns, reviews. It doesn't need to be big; it needs to be yours.
Experience — things you personally did, saw, fixed, failed at. Unfakeable, because it is a record, not an opinion. And it carries genuine opinion with it: if you've done the work, you're entitled to a verdict on it.
Distribution — an audience or channel you actually reach: a list, a following, a room you're trusted in. The same asset travels further from an account 14,000 people chose to follow than from one nobody did.
"I have nothing to write about"
Tomas Herzog believes this sentence, and he is wrong in an instructive way. Run the inventory on Herzog Physio:
Data only he holds. Treatment records from roughly 2,000 patients since 2014 — which injuries arrive, and how many sessions recovery actually took. One finding sits in there already: sports-injury patients who booked within two weeks of injury averaged 3.6 sessions to discharge; those who waited three months or more averaged 6.1.
Experience. Six physiotherapists who know which home exercise programmes patients actually keep doing and which get abandoned by week two.
Access. Forty-five new patients a month in the treatment room, and the GPs who refer them.
An opinion the data backs. "Waiting to see if it passes is the most expensive treatment there is."
No AI can generate Graz recovery data. No national content farm holds his records. No competitor can honestly voice his verdict. Tomas doesn't have nothing to write about — he has content assets a marketing agency would invoice heavily to invent, and couldn't.
The contrast is Ostara. Mateus's first content plan was "ceramics care tips" SEO posts — pure commodity, identical to what every pottery shop's chatbot would emit. The inventory found what only Ostara holds: 1,375 parcels shipped across the EU in the past year, of which exactly 9 arrived broken — a number that speaks straight to the breakage-worry cluster his Module 2 research surfaced. Plus a kiln log recording every firing and failure since 2020, and 14,200 followers — a distribution moat most small brands would trade their glaze recipes for.
AI's honest role
This module is not anti-AI; it is precise about AI. The machine is excellent at structuring, drafting, tightening and producing channel variants — it will save you hours in Lesson 5.5, and refusing that is vanity. It is useless at the thing that makes content worth reading, because it can only recombine what already exists. It has never treated a patient, packed a parcel or migrated an accountancy firm off a shared drive.
So the division of labour is fixed: you bring the ore; AI helps refine it. Feed it nothing and it refines nothing, confidently, at length. The full workflow comes in Lesson 5.5.
Content that never meets the funnel is a hobby
One discipline before you inventory. Everything you publish must ladder, eventually, to the conversion goal you set in Module 4. Not every piece sells — most shouldn't — but every piece belongs to a path that ends at your landing page. Engagement that never meets the funnel is a hobby, and you don't have hours for hobbies.
Now inventory what only you hold.
Do
Exercise 5.1.1 — Your content advantage inventory
Inventory your unfair advantages. List what your business knows, holds or has done that a competitor with a chatbot and no access to your business could not produce. Be concrete — "we know our industry" doesn't count; "treatment records from 2,000 patients since 2014" does.
Write these down — in your plan document, or on the worksheet at the end of this lesson.
| What to write | Guidance |
|---|---|
| Data only you hold | 20–100 words. Numbers only you hold: sales patterns, support questions, job records, reviews, returns. "None yet" is an honest and accepted answer |
| Lived experience | 20–100 words. "None" is not accepted — every business has this. Things you or the team have done, fixed, failed at, tested. Customer conversations count |
| Customer access | 10–80 words. Who can you talk to, survey or observe that an outsider can't? |
| Strong opinions | 10–80 words. Where do you disagree with the standard advice in your field — and what have you seen that makes you disagree? |
| Existing content | None · a few pieces · regular output — plus up to 30 words: what exists, and would any of it pass the stranger-with-a-chatbot test? |
| Your strongest moat | One of: primary research · experience · distribution. Which is most realistic for you in the next 90 days? |
As you write, sort your entries under the three moats. An empty column stays visibly empty — that's information, not failure.
Where this goes: section 6.1 — Differentiation assets — of your Marketing Plan. Keep the inventory open beside you in Lesson 5.2.
Check
Rubric
Mark your own work against these criteria.
| Criterion | 8–10 | 5–7 | 1–4 |
|---|---|---|---|
| Specificity | Assets carry numbers, dates, places ("2,000 patient records since 2014") | Real but vague ("we have customer data") | "We know our industry", "we care about quality" |
| Honesty | Gaps admitted; "none yet" written where true | One flattering stretch, unflagged | The inventory reads like a brochure |
| Experience surfaced | Concrete things done, fixed, failed at | Generic experience claims without incidents | Blank, or "10 years of experience" with nothing attached |
| Opinion genuinely held | A position your field would argue with, plus what you've seen that backs it | Safe consensus dressed as opinion | No opinion, or one nobody disputes |
| Moat realism | Chosen moat matches the strongest column and your next-90-days capacity | Defensible but optimistic | Chosen moat contradicts the inventory (distribution, with no audience) |
Pass: 5+ on every criterion. Distinction: 8+ on all five.
Quiz — four questions. Pick an answer to see whether you were right.
1. Which content idea has the most information gain?
- a) "10 ways to style your dinner table"
- b) "What is stoneware? A complete guide"
- c) "We shipped 1,375 ceramic parcels across the EU last year. Nine broke. Here's how we pack." ✔
- d) "Why handmade matters in 2026"
Why: only (c) contains something that did not exist before it was published — Ostara's own shipping record, answering a fear real buyers voiced. The others exist in ten thousand versions already, and a chatbot writes the next ten thousand free.
2. "We'll win by publishing five posts a week while competitors publish two." What's the flaw?
- a) Five is too few — aim for daily
- b) Volume of average content no longer earns attention, because average content is free to make — differentiation, not quantity, is the scarce asset ✔
- c) Posting frequency has no effect on anything
- d) Competitors can't see your posting schedule
Why: when the marginal cost of average is zero, out-producing anyone at average is out-producing infinity. The only race left is the one run on material competitors don't hold.
3. Which of these is NOT one of the three content moats?
- a) Primary research — data only you collected
- b) Posting frequency — publishing more often than competitors ✔
- c) Experience — things only you have done and seen
- d) Distribution — an audience only you reach
Why: frequency is a schedule, not a moat. Anyone can match it in an afternoon of generation — which is exactly what happened to it.
4. Herzog Physio's stretching videos get good engagement but never mention Graz, the clinic, or link anywhere. What does content–business fit say?
- a) Keep going — engagement compounds into customers eventually
- b) Add trending audio to reach further
- c) Content that never meets the funnel is a hobby — each piece needs a path that ends at the conversion goal ✔
- d) Delete the account
Why: the videos may be good, but goodness isn't the test. Every piece must belong to a path that eventually reaches the assessment-booking page — most pieces indirectly, some directly, none never.
Advance
You now hold an inventory of what no machine and no competitor can fake — and, just as valuable, an honest list of what you currently publish that should stop. That inventory, not a posting schedule, is your content strategy.
Next: M5.L2 — Your unfair advantage: research, experience, opinion. Turning the inventory into three specific content ideas, each one forced through the "only we can write this" test.
Mark your own work
| Good | Not yet | |
|---|---|---|
| Assets are concrete | Numbers, dates, places attached | "We have lots of experience" |
| The stop-list exists | At least one current activity marked "stop" | Everything you already do survived, conveniently |
| Opinion has teeth | Your field would argue with it | Nobody would bother disagreeing |
| Moat matches reality | Chosen moat is your strongest column | Chosen moat is the one that sounds best |
Worksheet
THE SCHOOL OF NET MARKETING
Lesson 5.1 — Content marketing after AI broke content
THE TEST — apply it to everything you publish
Could a stranger with a chatbot and no access to
my business have produced this piece?
If yes: it was worthless before you wrote it.
MY INVENTORY
Data only we hold (sales, jobs, support, reviews,
returns — with numbers and dates):
____________________________________________________
____________________________________________________
☐ honestly: none yet
Things we've done, fixed, failed at, tested:
____________________________________________________
____________________________________________________
Who we can reach or observe that outsiders can't:
____________________________________________________
Where we disagree with standard advice, and why:
____________________________________________________
____________________________________________________
WHAT WE CURRENTLY PUBLISH
☐ nothing ☐ a few pieces ☐ regular output
Which of it passes the test? _______________________
What should simply stop? ___________________________
MY STRONGEST MOAT, NEXT 90 DAYS (tick one)
☐ Primary research — data only we collected
☐ Experience — things only we have done
☐ Distribution — an audience only we reach
SELF-CHECK
☐ Every asset carries a number, date or place
☐ At least one current activity marked "stop"
☐ My opinion is one my field would argue with
☐ My moat matches my strongest column
Next: Lesson 5.2 — Your unfair advantage: research,
experience, opinion.
theschoolofnetmarketing.com/learn/your-unfair-advantage-research-experience-opinion