AI Niche Sites: Real Numbers Behind the Passive Income Pitch

Two Operators, One Crash, Different Fates

In May 2023, Jon Dykstra — a former lawyer who had spent a decade building content websites — netted $89,396 in a single month from display advertising on his portfolio of niche sites. By October of the same year, he was reporting "the lowest revenue earned in many years." Same sites. Same strategy. Five months apart.

Bryan Collins ran a similar operation: seven content sites, a full editorial team, "multiple five-figure monthly incomes for several years." When Google's Helpful Content Update hit, his traffic dropped 90%. He laid off his entire writing and editing staff and publicly announced he was leaving the business. "The passive income gold rush is over," he wrote. "AI slammed a nail in the coffin."

Both men ran legitimate businesses. Both got hit by the same algorithm update in the same month. One rebuilt. One quit.

If you're evaluating niche site publishing as a side hustle or career pivot — especially now that AI tools make the entry point look cheaper than ever — the question worth asking isn't "can this work?" It's what actually determined which of these two men survived. That's a more useful question, and the answer turns out to be surprisingly specific.

But before getting to what separated them, the aggregate data deserves a hard look. Individual stories are exceptional by definition. The average outcome is something very different.

What the Numbers Actually Say

The income distribution for bloggers and niche site owners is a power law, not a bell curve. The median outcome is near zero. The timeline to anything meaningful is measured in years. And the top earners pull the average far above what most people will ever see.

Start with the failure rate. Eighty-one percent of bloggers never make even $100 total from their blogs — not $100 per month, $100 ever. This comes from a survey of 1,000 U.S. bloggers, corroborated by multiple independent studies. The median blogger in year one earns $4.44 per month. That's not a typo.

The timeline to meaningful income is equally sobering. The median time to a full-time blogging income — defined as enough to replace a salary — is 48 months. Four years. And that figure has been getting worse, not better, declining from 28% of bloggers reaching full-time income within two years to 24% in the most recent survey period.

Google doesn't owe niche website owners ANY traffic. If you play the SEO game, expect to get burnt.
— Bryan Collins, Niche Website Publisher

For those who do persist, the 2026 Blogging Income Survey found that bloggers with 500 or more published posts average $9,460 per month. Bloggers with fewer than 100 posts average $38 per month. The inflection point is real — but it requires roughly 300 or more posts before meaningful income begins.

This is a tournament, not a career ladder. Most entrants exit before the payout window opens, and the payout window is measured in years, not months. For anyone considering this as a side hustle while employed, the relevant question is whether you can sustain two to four years of near-zero financial return on this specific project. If you need income from it within 18 months to justify continuing, the data says this model is a poor fit — not because the opportunity is fake, but because the timeline doesn't match the need. That holds whether you're a marketer, a nurse, or an HR director. The 48-month median isn't niche-specific. It's structural.

What Separated the Man Who Rebuilt from the Man Who Quit

Here's the answer to the opening puzzle.

Before Google's Helpful Content Update hit in late 2023, Dykstra had spent months building a newsletter audience — an owned channel Google cannot touch. When the HCU devastated his search traffic, he had somewhere to pivot. Collins' sites were Google-dependent. When traffic dropped 90%, there was no fallback. Dykstra's 2026 update: he's now generating "more traffic than Google alone" through newsletter and multi-source strategy. Collins' conclusion: "Google doesn't owe niche website owners ANY traffic. If you play the SEO game, expect to get burnt."

The difference wasn't talent, niche selection, or effort. It was one decision made before the crash — whether to treat Google as the business, or as one distribution channel inside a business that didn't need Google's permission to survive.

This variable matters even more now than it did in 2023. Google AI Overviews reduce clicks to top-ranking content by 34.5% to 58%, depending on query type. Seer Interactive found that organic click-through rates on queries with AI Overviews dropped from 1.76% to 0.61% between June 2024 and September 2025 — a 61% decline in roughly 15 months. Sixty percent of Google searches now end without any click to a website at all. The addressable traffic pool for informational niche sites — the exact content type that dominates the model — is structurally shrinking.

Even operators who successfully reach the 500-post threshold and $9,000-per-month in revenue are now sitting on an asset with a permanent structural headwind baked in. The 2026 version of this model requires building two things simultaneously: a content library that ranks and an audience that doesn't need Google to find you. That's a significantly more demanding project than the "write content, get traffic, earn money" pitch implies.

This isn't only a problem for bloggers. Anyone building a content-dependent online business — newsletters, YouTube, social media — faces the same single-platform dependency risk. The lesson from Dykstra and Collins applies wherever one algorithm controls your livelihood.

Where AI Actually Changes the Math

There is one place where AI tools shift the economics in an operator's favor. It's just not where most people think.

One anonymous Reddit operator documented a fully automated AI content site built for $1,571 total — OpenAI API for content generation, bulk WordPress publishing, programmatic SEO across 1,023 articles in the home improvement niche. Revenue went from zero to $3,916 per month in 14 months. The site sold for $108,000. The same operator manages a 41-site portfolio and stated plainly: "AI and Google updates have essentially killed the traditional content website business."

AI and Google updates have essentially killed the traditional content website business.
— Anonymous Reddit Operator, 41-Site Portfolio Manager

They're profiting from a model they believe is dying — and exiting before it does.

The exit math is real. Content sites currently sell for 28 to 38 times monthly profit on established marketplaces. A site earning $3,000 per month could sell for $84,000 to $114,000. That asset value is genuine. But it assumes a buyer believes the revenue will hold, which is an increasingly uncertain bet as AI Overviews expand and zero-click search grows.

AI automation lowers the cost of entry and compresses the time to monetization. It does not fix the Google dependency problem. It does not reverse the zero-click trend. The operators succeeding with AI automation in 2025 and 2026 are treating it as a sprint to an exit, not a marathon to passive income. If building and selling a content asset is the goal, AI is a legitimate tool. If indefinite passive income is the goal, the automation doesn't solve the structural problem.

The Filter That Separates Survivors from Casualties

The niche site model doesn't fail universally or succeed universally. It filters. The filter criteria are identifiable, and applying them honestly before investing time and capital is the highest-leverage decision you can make.

Drawn directly from what separated Dykstra's rebuild from Collins' exit, this model probably works for you if you have genuine expertise — original research, professional credentials, or lived experience — that AI and commodity writers cannot easily replicate. It works if you can sustain two to four years of near-zero financial return without needing the income to cover essential costs. It works if you're willing to build a non-Google traffic channel from day one, not as an afterthought. And it works if you're treating it as a business asset to build and potentially sell, not a passive income stream to live off indefinitely.

It probably doesn't work if you're entering a niche where you have no real expertise and plan to outsource all content to low-cost writers. It doesn't work if you need meaningful income within 18 months to justify continuing. It doesn't work if your entire traffic plan is "write SEO-optimized content and rank on Google." And it doesn't work if your primary source of information about the opportunity is a course seller with earnings screenshots.

These criteria don't change based on professional background. The relevant variable is whether you have expertise that produces content a search engine's AI can't summarize away — not what industry you come from.

What to Do Before You Register a Domain

The difference between Dykstra's rebuild and Collins' exit came down to one decision made before the crash: whether Google was the business, or just one distribution channel inside a business that didn't need Google's permission to survive.

The "passive income from niche sites" framing was always aspirational. Income from content sites is volatile, algorithm-dependent, and concentrated among a small fraction of persistent operators with genuine expertise. AI tools have made the entry cost lower and the exit value potentially higher, but they haven't changed the fundamental filter.

The people most incentivized to sell you this dream often derive their income from selling the dream, not living it. Course sales, coaching programs, and affiliate referrals to hosting companies are frequently larger revenue sources for "successful niche site operators" than the niche sites themselves. In April 2026, the FTC ordered Publishing.com to pay $1.5 million to settle charges that it misled consumers about income potential from its "foolproof, passive income system" — and that most buyers never came close to the promised results. Treat any income screenshot from a course seller as marketing copy, not data.

Before registering a domain, run a two-question audit. First, write down — in two paragraphs — what you know about your proposed niche that a generalist AI assistant would get wrong or miss entirely. If you can't fill two paragraphs, you don't have differentiated expertise yet. Second, map out your finances for 24 months assuming zero income from this project. If that math doesn't work, the timeline math doesn't work either.

This isn't a door closing. It's a filter clarifying who the door was ever really open for.

Recommended Tools & Resources