The question was never whether to build them by hand
I wanted landing pages for fifteen different industries. A page for real estate brokerages, one for dental groups, one for franchises, one for freight and logistics, and so on. Each one speaking that industry's language, showing that industry's content, answering that industry's version of the same problem.
With AI, the question was never "should I build fifteen pages by hand." Of course not. That is the easy thing to delegate. The real question, the one that actually decides whether this works, is different and a little uncomfortable: which one page is good enough that copying it fifteen times is a good idea?
Because that is what a template really is. Not a shortcut for making junk faster. A decision to make one thing excellent and let the machine carry that quality everywhere. If the one page is mediocre, you now have fifteen mediocre pages and a faster way to make a sixteenth. The leverage is real, but it points at whatever you point it at.
So I spent the time on one page, then turned its shape into a fill-in-the-blank
Here is where the work actually went. We took one industry and made the page as good as we could make it. We mocked up the real product inside the page, the actual interface, coded and on brand, not a stock screenshot. We invented a fake company in that industry and gave it believable assets to fill the library, real-sounding files with real-sounding details, so the page felt like a true example of that business using Masset, not a generic demo.
Once that one page was right, we did the part that makes it scale. We turned the shape of the page into a fill-in-the-blank. Every headline, every asset, every comparison, every example became a labeled slot. The design and the structure are fixed and live in one place. The words and the examples are just data. To add an industry, you fill in the blanks for that industry. You do not touch the design.
That is the move. The thing that took real taste and time, the design, you do once. The thing that is mostly research and writing, the industry-specific words, the AI is genuinely good at. So the human does the part that needs judgment, and the machine does the part that needs repetition, fifteen times over.
Two details made this safe instead of reckless. The blanks are strict, so a page cannot ship missing a piece or shaped wrong. And it is additive: the new pages stand beside the old ones without disturbing them, so nothing that was already working, search rankings included, got knocked over to do this.
Why these fifteen industries, and not the usual tech crowd
The idea did not come from a spreadsheet. It came from a customer. We work with a freight company, and in passing they told us about another group they know, a business with genuinely bad asset management, and how much better what we have is. That offhand comment did something in my head.
Most software, including a lot of ours historically, sells to the tech-savvy crowd that is already comfortable with AI. But the freight comment pointed somewhere else. There is a whole world of less-technical industries that are still figuring AI out, that do not have a person whose job is to wrangle their content, and who would feel the value of a simple, AI-ready home for their stuff more than anyone. Real estate. Dental groups. Senior living. Home services. Manufacturers. Franchises. The kind of business that will never build this themselves.
So the fifteen pages are a bet on those buyers. Each one targets a phrase like "digital asset management for [their industry]," on the theory that when someone in that world finally goes looking for a better home for their content, the page that speaks their language should be there waiting.
The honest result, five days in
Now the only question that matters. Is any of this working yet? I wired all fifteen pages and their target keywords into our own analytics tool so I could watch instead of guess. Here is the dated snapshot, pulled five days after the pages went live.
It is nothing. And that does not surprise me at all, because the pages just went live this week.
I want to be clear about why this is not a failure. Search engines and AI assistants have to find a page, crawl it, decide it is trustworthy, and only then start ranking it. That takes weeks to months, not days. A brand-new page showing zero is the normal starting line, not a verdict. The few visits the pages do have are direct, meaning people I or the team sent there, not anyone who found them through search.
AI sped up the building. It did nothing to the clock
This is the part I want every marketer to take away, because it is easy to get drunk on the speed and forget it. AI collapsed the cost of building fifteen good pages from weeks of work down to about a day. It changed nothing about how long Google and the AI engines take to notice them.
What that means in practice is that the scarce resource moved. When making the thing is nearly free, the bottleneck is no longer effort. It is two things instead. First, the quality of the one thing you choose to template, because the machine will multiply whatever you give it. Second, patience and honest measurement, because the result lives on a clock you do not control. The teams that win with AI are not the ones who ship fastest. They are the ones who ship something good and then actually watch what it does.
“AI collapsed the cost of building fifteen good pages to about a day. It changed nothing about how long Google takes to notice them.”
What I am actually betting on
So what do I expect from these fifteen pages? Eventually, hopefully, we rank for them on Google and inside the AI engines, so that when someone in a specific industry goes looking for a good, AI-ready, simple home for their content, they find a page built for them. If even a couple of these pages land, they will have paid for the whole batch many times over, because the cost of making them was so low. That is the shape of a smart AI bet: cheap to place, asymmetric to win.
If you want to try the same move, it is not really about landing pages. It is a pattern. Find the thing you make over and over in slightly different flavors. Spend your real effort making one version excellent. Turn the shape of it into fill-in-the-blanks. Let the AI fill the blanks. Then, and this is the part people skip, wire it to a number so you can tell the truth about whether it worked, instead of admiring how fast you made it.
I will post the follow-up to this in thirty and sixty days with the real numbers, good or bad. That is the whole point of building in public. You do not get to only show the wins.
Key Takeaways
- With AI, building many pages by hand is never the question. The real work is getting one page good enough to be worth copying.
- Make one page excellent (we coded a real product mockup and built believable industry-specific example assets), then turn its shape into fill-in-the-blanks so every new industry is just data, not design.
- The idea came from a freight customer pointing at another business with worse asset management, which surfaced an underserved buyer: less-technical industries still figuring AI out.
- Honest result, five days in: all 15 pages rank nowhere and have zero organic or AI traffic, just about 20 direct visits. That is the normal starting line, because indexing and ranking take weeks to months.
- AI compressed the cost of building, not the clock for ranking. The scarce resource moved to the quality of the one template and the patience to measure it.


