Whenever I see discussions like “Casting a Wide Net with AI Job Applications Could Backfire,” another scene immediately pops into my head: the marketing department schedules five posts a day, one AI generates them, another AI summarizes them, and the humans handle the end-of-month question: “So where are the customers?”
Congratulations, the content production line is fully automated. You’ve even eliminated the readers. 😂
But ask me whether content marketing still works, and I lean toward an answer that might earn me a punch: Yes. And some content may be closer to doing its job the less it needs to be clicked.
Don’t throw the traffic report at me just yet.
Just Because Customers Aren’t Reading Doesn’t Mean the Content Isn’t Working
Suppose someone asks AI: “Is there a tradesperson in Taichung who’s willing to repair old wooden windows, can give an estimate from photos first, and is available on Saturdays?”
One business’s website says: “Professional team, guaranteed quality, dedicated service.” Another says: “Localized rot in wooden frames can be assessed for repair; a warped frame may need replacing. Send photos of all four corners of the window frame and the tracks first. Saturday site visits require an appointment.”
The first sounds like it’s running for neighborhood chief. The second at least lets people know whether it’ll take the job.
I can’t guarantee AI will recommend the second business. Different systems use different data sources, crawling methods, and rankings, and they may miss or misread things. But the second business has left information that can be checked against someone’s needs; the first has left an inspirational slogan shared by businesses across Taiwan.
The assumption being challenged here is that content must earn clicks before it can earn business.
In AI search scenarios, there’s another route: content is first used to understand whom you’re a good fit for, and customers then approach you with more specific questions. The article may not rack up views, but someone calling can ask, “Can you repair just the window frame?”
When Everyone Can Post Daily, Posting Daily Isn’t So Special Anymore
AI makes writing cheaper, and businesses’ instinct is usually to increase volume. Two posts a week becomes twenty a day, as if the internet were a hot pot and the more ingredients you dumped in, the better the deal.
The problem is, anyone can generate “Five Reasons Your AC Isn’t Cooling.” After reading it, you still don’t know whether this technician will dismantle interior fixtures, how they charge for remote locations, or which faults they simply won’t take on.
By contrast, a slightly sobering case note—“We didn’t recommend replacing the unit this time because the problem was with drainage; however, parts for this older model are inconsistently available, so the next breakdown may not be worth repairing”—contains more information than ten repetitions of “Your comfort, our commitment.” That one belongs embroidered on a throw pillow.
This kind of content may help AI match a business to someone’s needs, while also showing real people how decisions are made. It also exposes limitations: you can’t do everything, and your prices aren’t guaranteed to be the lowest. The marketing department may frown, but a customer looking for someone to put out a fire may breathe a sigh of relief.
Of course, people who make a living from content will push back: if the answers all get summarized and the website gets no visitors, who pays for production? That’s a very real financial question. For ad-supported media, being seen without anyone visiting the site can hurt; for a studio that can only take on a few projects a month, a few hundred fewer passersby and one more suitable inquiry may add up differently.
So I’d want to put another sheet of paper beside the report and record the first thing customers ask. If digital marketing looks only at traffic, it’s easy to count “I finally found someone willing to take this job” as a failure nobody read.
On Friday night, the wooden-window repairer gets a message: “After reading what you wrote, this window might be beyond patching. I took photos of all four corners. Could you have a look first?”
He zooms in on his phone and studies the rotten section. The article’s view count in the dashboard still isn’t much to throw a celebration over.