Solopreneurship

Can You Really Make Money From an AI Directory Website?

A reality check on Jesse Cunningham’s “100% AI directory” plan

Yes, it can make money, but not the way the headline sounds. An AI-built local directory is a lead-generation asset you use to open conversations with contractors and sell featured listings, SEO, or websites. It is not a passive ad-revenue site, and the data step carries a real terms-of-service risk.

I found this through How to Build and Make Money from 100% AI Directory Websites by Jesse Cunningham, who runs the Rank Expand Academy. He shows a directory for one trade in one state, built with AI from Google Maps data, in eight steps.

I have done SEO since 2007, and every few years a new way to publish a lot of pages quickly shows up. I ran a directory site myself five or six years ago. It has since fallen into disrepair and I moved on to other things, so I have not seen how the model works in the age of AI. That is a big part of why this video caught my eye.


Three Key Takeaways: What a Local Directory Can and Cannot Earn

  • The money is in the conversations, not the traffic. By the creator’s own account, a directory is a foot in the door for selling featured listings, SEO, and websites. He says plainly it will not live on display ads.
  • The data step is the weak link. The video says the Google Maps data costs under a dollar, but Google’s own terms restrict scraping and saving business names and addresses.
  • You can test the idea for almost nothing. Build a small version, check whether it indexes, and talk to five business owners before you build anything big.

This post is built on ideas from How to Build and Make Money from 100% AI Directory Websites by Jesse Cunningham. Watch it here:


What the Video Actually Claims

Cunningham’s plan is a directory for one trade in one state, such as custom pool builders in Florida. Every city gets a page, every business gets a page, and the city page is a “best of” list. His first build had over 1,500 businesses and over 1,700 pages.

The steps are simple to say. Pick the niche, buy an exact-match domain, pull business data, build static pages with AI, get them indexed, and get paid. He wants a niche where people get several quotes, jobs cost four figures or more, and there is something to explain, with 250 to 5,000 businesses in the category.

The part I find most useful is the enrichment step. He has the AI check each business website against 8 to 15 questions a customer would ask, like whether a roofer offers financing, and marks each answer yes, no, or not enough information. That is real added value, and it matters later.


What Holds Up

The sales logic is the strongest part. Calling a contractor to say you featured them on a directory is a warmer opening than a cold pitch. Cunningham also says outright that this is not for people who expect passive ad income, which is more honest than most pitches in this space.

The narrow niche advice holds up too. One trade in one region gives you something specific to say, and it is cheap enough to treat the first build as a test.

Here is where it gets soft. In the part of the video I reviewed, there is no revenue figure from any directory, no count of featured listings sold, and no ranking results. Income claims without numbers are a pitch, not proof. I reviewed the video up to about the 15-minute mark, so judge the rest for yourself.


What the Video Skips

The creator sells the system, and that colors the advice. He promotes a paid academy with a seven-day free trial and a zip file that contains the full blueprint. That is a financial stake, and it is worth knowing when he calls the build easy.

Now the fact-check. Cunningham says the data costs under a dollar for a whole state. But Google’s Maps Platform terms say customers will not “export, extract, or otherwise scrape” Maps content for use outside Google’s services, and the examples include copying and saving business names, addresses, or user reviews. The portion of the video I could access does not mention these terms. Other data sources exist, but the video centers on Google Maps.

The second gap is Google’s own policy. Its spam policies define scaled content abuse as generating many pages mainly to manipulate rankings, and they name generative AI used to create many pages without adding value. They also list scraping search results and repurposing them into many pages, and “substantially similar pages” that sit closer to search results than a browsable hierarchy. The test is value, not method. His enrichment step could satisfy that test. Thin copies of Maps listings would not.

The video also cites a study saying list pages are about 18% of citations in Google’s AI Overviews. I could not trace that figure. A separate sponsored study from Evertune, covering six AI platforms, found listicles drew 63% of citations among the most-cited pages. Lists get cited a lot, but that does not mean your directory will be.


Renzie’s Take:

I like the sales idea and distrust the production line. A directory that gives a pool buyer real answers about real companies is useful. A directory of 1,500 pages built from scraped listings is a bet that Google will not notice.

Build the smallest version that is honestly useful, and sell to people before you scale pages. The reason the idea has legs is intent. People who search a directory are usually ready for the next step, because they are looking for actual suppliers and actual solutions to a problem.

I would like to start one again with what I know now and see whether it holds up.


Where to Actually Start

You do not need his full system to learn whether this works. Run a test small enough that a bad result costs you a weekend.

  1. Read Google’s rules first. Read the spam policies section on scaled content abuse and doorway pages, then the Maps Platform terms on scraping.
  2. Choose a trade and region you can judge. Pick a service people compare and pay four figures for, in a region where you can actually call the businesses. This works anywhere with local services, not only the US.
  3. Collect data in a way the source allows. That could be business websites, public registries, or businesses that submit their own details. Check each source’s terms.
  4. Build 20 to 30 pages by hand-checked AI, not 1,700. Add original answers to the customer questions on every listing. Put your name and face on the site, which he also recommends.
  5. Talk to five owners before scaling. Ask whether the listing is accurate and whether they would pay for a featured spot. Their answers are your real market research.
  6. Watch Search Console. Check whether Google indexes the pages and shows impressions before you add more.

If you want the creator’s full system, the Rank Expand Academy is linked in the video description. Remember the financial stake disclosed above.


What I Would Pursue and What I Would Skip

Pursue this: a tiny directory built to meet buyers, with facts only your research can supply, such as which companies offer financing or a warranty. Skip this: bulk Google Maps scraping, publishing every page at once, and expecting featured listings alone to pay. A related read is Can a $199 Niche Blog Really Pay Your Rent? I Checked the Math, and for the wider picture of AI-built income ideas see AI Solopreneur Business Ideas: How to Choose a Realistic Path and Is YouTube Automation Worth It in 2026?.


The Bottom Line: Use It to Meet Buyers, Not to Mass-Publish

  • Test small. Twenty good pages teach you more than 1,700 thin ones.
  • Check the data terms first. The cheap scrape is the riskiest part.
  • Sell the relationship. The money comes from listings and services, not traffic.

Pick one trade in one region this week and talk to five business owners before you build a single page. For more ways to build income without a team, start with How to Become a Solopreneur in 2026.


Sources