This one is a bit speculative, so treat it as a thinking exercise rather than a to-do list. But it is less speculative than it sounds, and the exercises at the end are worth doing regardless of whether I am right about the timing.
Set aside an hour. You will need something to write on.
The premise
For roughly seventy years, marketing has been about packaging.
Feelings. Hooks. Triggers. The clever headline, the hero image, the thing that makes somebody stop scrolling. An entire industry built on understanding how humans decide, and then arranging things so they decide in your favour.
It worked because the person on the other end was a person.
That is changing. Slowly for now, then probably quite quickly. People are increasingly asking an AI to find them a supplier, compare two options, or book the thing, rather than doing the browsing themselves. Every time that happens, the decision is being mediated by something that does not have feelings to appeal to.
The internet is turning from an interface into an API. Something goes and fetches, compares and decides on your behalf, and reports back.
And the thing doing the fetching does not care about your hero image.
What it does care about
Specifics.
What the thing actually is. What it does. What it costs. Who it is for and who it is not for. How it works. What it is made of. What happens if it goes wrong. What somebody else said about it who was not paid to say it.
Facts, in other words. Checkable, comparable, unambiguous facts.
The bit that made me laugh
Here is what struck me when I properly thought this through.
This is not a new kind of advertising. It is a very old one.
Before the industry discovered emotional persuasion, advertising was largely about reasons. Claude Hopkins, writing in the 1920s, argued that good advertising was salesmanship in print: specifics, evidence, a stated reason why this product rather than that one. Adverts described how the thing was made, what went into it, and what it would do. Dull by modern standards, and extremely informative.
Then the industry worked out that feelings sold more than facts, and spent seventy years getting very good at feelings.
And now the customer is a machine, and we are going back to reasons. Ingredients, specifications, proof, who it is for. The whole thing has come round in a circle, and the businesses that never learned the packaging tricks in the first place are oddly well placed.
Before you start
One honest caveat, because I do not want you burning a Saturday on my say-so.
Humans have not gone anywhere. Most of your customers this year will be people, and people still want to like you, still respond to a decent photograph, and still ring the person who sounds like they know what they are doing. Do not strip the humanity out of your business because of a blog post.
The point of this exercise is not to replace what you have. It is to find out whether there is anything underneath it, because right now most businesses have packaging and nothing else. If the wrapping came off, there would be no product description at all.
Right. Here are the exercises.
Exercise one: the stripping test
Time: 15 minutes.
Open your homepage. Now imagine everything visual is gone. No photographs, no logo, no colours, no layout. Just the words, as plain text.
Read what is left, out loud.
Now answer honestly:
- Could a stranger tell exactly what you sell?
- Could they tell where you operate?
- Could they tell roughly what it costs?
- Could they tell who it is for, and who it is not for?
- Is there a single specific, checkable fact anywhere in it?
What most people find: the page says almost nothing. It communicated entirely through pictures, layout and tone, and once those are removed there is very little content underneath.
That is exactly what an agent sees.
Exercise two: write your spec sheet
Time: 30 minutes.
Forget marketing entirely. Write your business the way somebody would write a technical datasheet for a component.
No adjectives. No “passionate,” no “leading,” no “trusted.” Just facts, in a list.
Cover:
- What you do, precisely. Not “property maintenance,” but the actual jobs.
- Where. Named towns, not “the North West.”
- What it costs. Real numbers or real ranges, and what moves them.
- How long things take.
- What you will not do, and who you are not for.
- What is included and what is extra.
- Qualifications, certifications, insurance, guarantees.
- What happens when something goes wrong.
The test: could somebody compare you against a competitor using only this document? If not, keep going.
Most businesses find this uncomfortable, because it forces decisions they have been avoiding. That discomfort is the useful part.
Exercise three: the tacit knowledge dump
Time: 20 minutes, and you will not finish it.
Write down five things you know about your trade that are not written anywhere in your business.
The problem that always comes up on a particular type of property round here. The material you stopped using and why. The question customers always ask that nobody answers properly. The thing that goes wrong when people try it themselves. The bit of the job that takes twice as long as anyone expects.
This is the most valuable content you own and it is currently stored in one person’s head.
It is also the only content a machine cannot generate, because it did not happen to anybody else. If you want a reason to bother with any of this, that is the reason.
Exercise four: interrogate yourself
Time: 15 minutes.
Now go and find out what a machine currently thinks. Open ChatGPT or Google’s AI mode and run these.
“What does [your business name] in [your town] do, what do they charge, and who are they best suited for? Tell me what you cannot find out.”
That last clause is the important bit. It will tell you what is missing.
“I need [your service] in [your town]. Compare the options and recommend one, explaining why.”
See whether you appear. See what the recommendation is based on. See who beats you and go and look at what they have published that you have not.
“What are the risks or downsides of hiring [your business name]? What would you want to know before committing?”
Uncomfortable, and the most useful of the three. It surfaces the questions a cautious buyer has that your website does not answer.
Exercise five: the gap list
Time: 10 minutes.
Put the four exercises side by side.
Everything an agent could not find out, everything your spec sheet forced you to decide, and everything sitting in your head. That is your content plan for the next six months, and it will be more useful than anything a marketing agency would sell you.
Work down it. One item at a time, written down plainly, published where it can be found.
What you have actually done
Not marketing, in the sense the last seventy years would recognise it.
You have documented your business. Which is the thing everybody skips, because it is not creative and nobody claps.
But it is the thing that makes you legible to a machine that is deciding who to recommend, and it happens to be the thing that makes you clearer to humans as well. Nobody ever lost a customer by being too easy to understand.
The packaging still matters. Make it beautiful, by all means. Just make sure there is something in the box.
Want a hand working out what is missing? Get your free audit, or email hello@inkon.co.uk. I read every message myself.