Every few weeks somebody sends me a link to a tool that writes the copy, builds the site, cuts the video and makes the logo. Usually with a note that says some version of: doesn’t this worry you?
It doesn’t, but not because the tools are bad. They are genuinely good, they are getting better, and we use them here. It doesn’t worry me because of what happens when everybody has them.
Efficiency stops being a differentiator the moment everyone has it
For about twenty years, being able to produce marketing material faster or cheaper than the shop down the road was a real advantage. Desktop publishing did it. Stock photography did it. Template builders did it. Each time, the advantage lasted right up until the tool became common, and then it wasn’t an advantage any more — it was the floor.
That is where we are with generative tools now. If a competitor can produce forty social posts in an afternoon, so can you. If you can generate a hundred headline variations, so can they. Nobody wins that race, because the race ends in a tie the moment the starting gun goes off.
What doesn’t tie is the decision about which of those forty posts is worth publishing. That decision is the work.
The bottleneck moved
It used to be that making things was slow and deciding was fast. You had an idea in a meeting, then spent three weeks producing it. Now producing takes an afternoon, and the slow part is knowing what deserves to exist.
That flips where the value sits:
- Knowing what the business actually needs. A tool will happily build the thing you asked for. It has no opinion about whether that thing will move the number you care about.
- Editing. Generating ten options is free. Knowing which one is right, and why the other nine are not, is the whole job.
- Consistency across time. A model has no memory of the decisions your brand made two years ago, or why. Somebody has to hold that.
- Knowing when the answer isn’t a deliverable. Sometimes the right recommendation is a phone script, a price change or a better front-desk process. No generator is going to tell you that.
The sameness problem
Here is the practical risk, and it has nothing to do with quality.
These tools are trained on what already exists, which means they are very good at producing the median. The average layout. The average headline. The tone every other company in your category is already using. Each individual piece looks fine — clean, competent, nothing wrong with it.
The problem shows up in aggregate. Scroll a category where everyone is using the same tools with the same prompts and it all blurs together. Same gradient, same stock-photo warmth, same three-word headline followed by a supportive subhead. Competent and invisible.
That is not a tooling failure. It’s what happens when the input is “make me something good” instead of “here is what we actually believe, made specific.”
What actually separates work now
Four things, none of which a tool produces on its own:
A point of view. Saying something specific means some people will disagree. That is the cost, and it is also the only way to be memorable. The safest possible message is also the most forgettable one.
Real material. Your people, your building, your product, your customers. A generated image of a generic technician is fine. A photograph of your technician, in your shop, is proof. Nobody can generate your proof.
Restraint. When production is cheap, the temptation is to make more of everything. The brands that read as expensive are usually doing fewer things at a higher standard.
Judgment about what not to ship. The most valuable thing an experienced person does is kill work. That instinct comes from having watched things succeed and fail, which is a thing you earn rather than install.
How we actually use it
Nothing precious here. We use these tools the way we use any other: where they save time on work that doesn’t need a human hand.
Drafting and sorting research, cleaning up transcripts, writing the first pass of something that will be rewritten anyway, generating options to argue with, handling code that is more tedious than hard. All fine. All faster.
What we don’t do is let it make the decisions that require knowing your business — the strategy, the positioning, what the photograph should be of, what the site needs to accomplish before anyone chooses a typeface. And we don’t publish anything nobody read carefully first.
The efficiency goes into the work, not instead of it.
The bet
Our bet is that as the cost of producing content falls to roughly zero, the value of taste and judgment goes up, not down. When everything is well-produced, being well-produced stops being impressive. What impresses is a brand that clearly knows who it is, says something true, and shows you real evidence.
That was always the actual job. The tools just removed the excuse that producing the work was hard.
Wondering where these tools fit in your own marketing? We will give you a straight answer about what they help with and what still needs a person.
