Other Income Ideas Half-true — works only if you do the unspoken work

By the Income Reality Check Editorial Team · Updated 2026-07-31 · 7 min read

AI built a directory site in minutes — the traffic part it skipped

Verdict: Half-true — works only if you do the unspoken work. The build is genuinely fast and real; the income is not.

In “I Asked AI for a Money-Making Website (It Built One in Minutes),” Chris Koerner of The Koerner Office Podcast builds a live business directory on screen using Abacus.AI’s ChatLLM agent — searchable homepage, category pages, individual listings, and a working Stripe checkout that charges businesses $30/month for a featured spot. It’s a clean demo. The pitch is that once built, a directory “basically runs itself.” Is that real? The building is. The money is not, and the gap between the two is the whole story.

What the video actually claims

Koerner’s premise is that directory sites are boring, durable businesses — the kind that “were here before AI and will be here long after.” He uses ChatLLM (chatllm.abacus.ai) to do three things: brainstorm an underserved niche, compile a starter list of real businesses, and then hand that list to an AI agent that builds and deploys the site.

He pits GPT-5.6, Opus 4.8, Gemini 3.5 and others against each other, lands on an “estate sale cleanout” directory, and has the agent scaffold “Estate Cleanout Pros” with a Stripe-powered $30/month featured upgrade. He deploys it to an Abacus subdomain on camera.

On monetization, he names three routes: featured placements (businesses pay to rank higher), banner ads, and lead generation (collect consumer inquiries, then sell each lead to the listed businesses “to the highest bidder”). He’s candid that businesses “only need to pay if they want a higher placement, and it only really makes sense for them to pay once they’re actually getting some leads and some traffic from our website.” That single sentence is the pin in the balloon — and he moves right past it.

What the method actually requires

Here’s the thing the demo quietly assumes: someone visiting the site. A directory with no traffic has no leads, and businesses don’t pay $30/month for a featured spot on a page nobody sees. The build is the easy 5%. Traffic is the other 95%, and it’s slow.

For a brand-new domain with no backlinks, meaningful organic traffic typically takes six to twelve months of consistent SEO work, according to Shopify’s SEO timeline breakdown. Directories are actually harder than average, because thin, auto-generated listing pages are exactly what Google’s helpful-content systems demote. Twenty scraped listings — the number the agent used — is not a business; it’s a placeholder.

The costs are modest, which is the honest part of the pitch. ChatLLM is real and cheap: Abacus.AI’s own FAQ lists it at $10/month per user, including agent access and image generation. Stripe takes 2.9% + 30¢ per charge. A domain runs about $12/year. So your out-of-pocket to launch is genuinely under $25. But cheap-to-build and profitable are different claims.

Now run the banner-ad route with a calculator. AdSense pays by RPM — revenue per thousand page views. Google’s own help page defines it and gives a worked example of $4.00 RPM. General-interest niches run roughly $2–$8 per thousand views. So a directory pulling 5,000 visitors a month — already a real achievement for a new niche site — earns something like $10–$40/month from ads. That’s the passive part. It buys back the tool subscription and not much else.

Every column has the same prerequisite. Traffic. And traffic is the one thing the AI agent can’t generate for you.

Who actually wins this game?

Directory operators who succeed almost always bring something to the table before they open the tool. Some have an existing audience — Koerner himself has a podcast and a following, which is its own distribution channel that a first-timer copying the video does not have. Others have SEO chops and will spend months building listing depth, earning local citations, and getting links. A few are early movers in a niche so obscure that even weak pages rank.

There’s also a quieter problem the video glides over: the site lists businesses “without their permission,” using AI-compiled data that Koerner himself says you “want to fact check” before publishing. He caught his own bad prompt on camera (he asked for “real estate cleanout” businesses and got junk-removal companies). Publishing wrong phone numbers or addresses for real businesses isn’t just sloppy — it can invite complaints and takedown requests. Accuracy is manual labor, and it doesn’t stop after launch.

What you’d realistically earn

Koerner never puts a dollar figure on his own site, which is to his credit — he’s showing a method, not claiming a paycheck. But the video’s framing (“make me a lot of money,” “quite passive,” “basically run themselves”) sets an expectation worth grounding.

A realistic path for a beginner who does the unspoken work looks like this: roughly $0 for the first three to six months while the site earns its search rankings; maybe $10–$50/month from ads once traffic reaches a few thousand visitors; and featured-listing revenue only after you can prove to a business that your page sends them calls. Ten paying businesses at $30/month is $300 — a real number, but one that arrives in year one only for people who treat it as a job, not a set-and-forget asset. Passive it is not, at least not until it’s been actively grown first.

Who this is (and isn’t) for

This makes sense for someone who already understands local SEO, can commit five to ten hours a week for six-plus months, and treats the AI build as a head start rather than a finish line. If you enjoy the grind of link-building, listing curation, and outreach to local businesses, the low upfront cost is genuinely attractive.

It’s a poor fit if you’re expecting the “in minutes” from the title to describe the income as well as the build. If you have no audience, no SEO experience, and no patience for a half-year with zero revenue, you’ll likely abandon the site long before it ranks — which is exactly what happens to most of the directories that stall at a few hundred monthly visitors.

One more note for U.S. readers: if you ever turn this into a paid course or coaching pitch that promises earnings, the FTC’s proposed Earnings Claim Rule would require you to keep written substantiation for any income claim you make. Building the site is legal and fine. Selling exaggerated dreams about it is where regulators start paying attention.

What to remember

The demo is honest about the build and vague about the business. ChatLLM really does scaffold a working, payment-connected directory in minutes for about $10/month — that part isn’t hype. What it can’t do is send you visitors, and without visitors there are no leads, no featured upgrades, and only pocket change from ads. Treat the AI as the fastest part of a slow project, budget six to twelve months of real SEO effort, and the “passive income” label starts to look like a description of month eighteen, not month one. For more on where these “AI builds you a business” demos oversell, see our looks at Fable 5’s one-prompt business and the $10k website built in minutes.