After the second AdSense rejection, I did something I hadn't done before. I actually read Google's content policies. Not skimmed them. Read them. Word for word. And somewhere in the middle of the "low value content" section, I found a line about AI-generated content.
Checking AdSense status on my phone
Google's position, as far as I could understand it, was this: content generated primarily by AI, without human oversight or added value, may be considered low-quality and could affect AdSense eligibility. Fair enough. I get it. Nobody wants a search result that reads like a machine wrote it. Google's job is to show people useful, human content. If my blog posts read like a robot spat them out, Google has every right to reject them.
But here's where it gets interesting.
The Company That Sells AI Rejecting Sites For Using AI
The same week I got rejected for AI content, Google ran an ad in my Gmail sidebar promoting Gemini , their AI writing assistant. "Write faster with Google AI," the ad said. A few days later, Google announced a new feature in Google Docs that uses AI to generate entire documents from a prompt. They're investing billions in AI. They want everyone to use it. They'll even sell it to you.
So let me get this straight. Google will happily sell me a tool to write content. They'll promote it in their own products. They'll push it in ads. But if I use that tool . or any AI tool ( to write content on my website, and then apply for their advertising program, they might reject me for it.
I'm not complaining. I'm just observing the irony.
The Honest Part
Here's where I have to be honest with myself. Some of my early blog posts were rough. Not all of them were AI-generated, but some were AI-assisted. I'd write an outline, feed it into a tool, and get back a draft. Then I'd edit it : sometimes lightly, sometimes heavily. The ones that got rejected were probably the ones where I edited lightly. Where the structure was too clean, the sentences too perfect, the tone too generic. Google could tell. Not because they have some magical AI detector, but because the content didn't sound like a real person wrote it.
And honestly? They were right to reject it.
If I'm being really blunt, some of those early posts were garbage. They existed to fill pages. They existed to host affiliate links. They said what a hundred other sites said, in slightly different words, and added nothing that a reader couldn't find elsewhere. Whether a human wrote them or a machine wrote them didn't matter — they weren't good enough either way.
The Double Standard Is Real, Though
But I'd be lying if I said the double standard didn't bother me. Google's own search results are increasingly filled with AI-generated content. Their AI Overviews — the summaries that appear at the top of search results — are literally AI reading web pages and writing a response. They're doing exactly what they tell website owners not to do: using AI to generate content for users.
The difference, presumably, is that Google's AI is good and mine wasn't. Or that Google's AI serves their users and mine served my affiliate links. Or that Google is Google and I'm not.
Take your pick.
What I Actually Did About It
I could have written an angry blog post about the hypocrisy. I could have ranted on Twitter. I could have complained in forums. Instead, I did the boring thing: I changed how I write.
I stopped using AI to generate blog posts. Not because Google told me to, but because the rejection made me realize something I'd been avoiding: AI-assisted content wasn't helping me. It was making me lazy. When I used AI to draft, I wrote faster but I wrote worse. The posts didn't have my voice. They didn't have my actual experiences. They didn't have the specific details that only I would know — the name of the Hostinger rep I talked to, the exact dollar amount of my first commission, the way it felt to stare at a rejection email at 11 PM on a Tuesday.
AI can write a thousand words about affiliate marketing. It can't write what it felt like to get my first $42.50 commission. It doesn't know. It wasn't there.
So I went back to writing like I write in my journal. Messy. Honest. Sometimes too long, sometimes too short. With details that no AI would think to include because they're too small, too human, too specific. The kind of writing that doesn't sound polished because it isn't. The kind that sounds like a real person because it is one.
The thing I carried forward
The irony of Google selling AI while rejecting sites for using it isn't the point. The point is that I was using AI as a shortcut, and shortcuts don't build businesses. They build pages. And pages without substance get rejected — by Google, by readers, by the market.
Google's policies might be hypocritical. Their timing might be convenient. But their rejection was accurate. My content wasn't good enough. Not because a machine helped write it, but because I didn't put enough of myself into it.
The fix wasn't to stop using AI. The fix was to start being honest. To write about what actually happened to me, in my own words, with the details only I would know. That's not something you can prompt a machine to do. That's something you have to live first and write second.
So yeah. Google rejects AI content while selling AI tools. That's funny. But it's also fine. Because the content that got me rejected wasn't rejected for being AI. It was rejected for being empty. And that was on me, not on the machine.
— Justice
The Red Banner on the AdSense Dashboard
The rejection arrived on a rainy Sunday evening while I was riding a commuter minibus taxi from Park Station back toward Rosebank. I pulled out my phone to check my Google AdSense application status on my Chrome browser. Sitting right at the top of the account home page was an unmistakable, bright red notification banner: "Site Status: Needs Attention - Your site is not ready to show ads. Low Value Content / Policy Violation."
The rejection felt like a heavy blow. I had spent R450 on domain registration, server setup, and three weeks of intensive labor drafting twenty-five comprehensive articles covering tech tools, web software, and productivity guides. I opened the policy explanation link hoping for constructive, actionable feedback. Instead, I was met with vague corporate boilerplate language stating that my site lacked sufficient original value, provided poor user experience, or contained thin, automatically generated content. Receiving an automated refusal after weeks of dedicated work was deeply frustrating.
The Irony of the Google Search Results Page
In a mood of cynical curiosity, I typed the exact rejection string into Google search: "how to fix low value content AdSense rejection". What appeared at the top of the search engine results page (SERP) was mind-boggling in its corporate irony. The very top of the screen was dominated by a large Google AI Overview box, aggregating AI-generated text summaries directly into the search results page! Right beneath the AI Overview were three paid Google Ads promoting "Automated AI Content Writers" and "One-Click AI Blog Generators"!
The hypocrisy was breathtaking. Google was aggressively deploying Generative AI models across its core search product to summarize publisher content without sending traffic to original sites, while simultaneously running paid ad networks monetizing AI writing tools. Yet, when an independent publisher utilized AI assistance during draft research, Google's AdSense compliance bots slapped the site with a "Low Value AI Content" penalty. The system felt bizarrely hypocritical, setting strict rules for independent creators while breaking those exact rules on its own platform.
Auditing the Fluff of Generic AI Drafts
Once the initial annoyance passed, I decided to conduct a cold, objective audit of my published articles to understand why Google's automated compliance filters had flagged my content. I printed out five of my published review articles and read through them with a red pen in hand. Reading my own published work on physical paper was a humbling experience. I quickly began noticing distinct, unnatural language patterns that modern Large Language Models (LLMs) output when generating unedited text drafts.
My articles were filled with empty, repetitive introductory fluff phrases like: "In today's fast-paced digital landscape, finding the right tool is paramount...", "It is important to remember that...", and "Furthermore, delving into this aspect reveals...". The content lacked genuine personal voice, original testing photos, local South African context, exact Rand pricing, terminal execution logs, and real human emotion. The text was technically correct but completely devoid of EEAT (Experience, Expertise, Authoritativeness, and Trustworthiness). Google's automated quality filters had correctly identified that my articles read like generic, synthesized web summaries rather than authentic, hands-on reviews written by a real person.
Injecting Real South African Blood, Sweat, and Code
Accepting this reality, I embarked on a complete editorial overhaul of all twenty-five articles on my blog. I systematically stripped out every single paragraph of generic AI intro fluff and replaced them with raw, authentic first-person experiences. I injected rich local South African context into every piece: describing my real-world struggles with Stage 4 loadshedding, paying R149 for prepaid Vodacom mobile data bundles, navigating Capitec banking payment gateways, and running Python automation scripts on an old desktop tower in Johannesburg.
I added original high-resolution screenshots taken directly from my own phone and laptop browser, complete with annotated error logs and terminal command line outputs. I replaced generic product feature summaries with transparent pros-and-cons breakdowns, exact pricing tables converted into ZAR (Rand), and honest personal opinions based on real hands-on usage. By injecting my own blood, sweat, code, and local living experience into every article, I transformed my blog from a cold AI text directory into an authentic personal publishing log.
Beyond AdSense: Focusing on High-Ticket Affiliate Revenue
The ultimate realization from this AdSense rejection was a fundamental shift in business monetization strategy. AdSense display advertising is a high-volume, low-margin business model. In South Africa, AdSense CPM rates (earnings per 1,000 ad impressions) are notoriously low, often paying between R15 and R40 ZAR per thousand pageviews. To earn a modest R10,000 per month from display ads, a blog needs over 250,000 monthly pageviews—a massive traffic threshold for a new independent site.
Direct affiliate marketing, on the other hand, operates on high-ticket performance commissions. A single successful referral lead or software signup on an affiliate network like MaxBounty or ClickBank can pay between $25 USD and $100 USD (R460 to R1,850 ZAR) per conversion. Suddenly, achieving your monthly revenue targets requires only fifteen to twenty qualified, high-intent readers taking action rather than hundreds of thousands of passive ad impressions. Converting that initial AdSense rejection into a drive for direct affiliate publishing transformed my site into a far more lucrative, independent media business.
Reclaiming control over my site's monetization strategy also meant I didn't have to clutter my clean layout with intrusive banner ads, pop-ups, or auto-playing video units. My readers enjoy a fast, distraction-free reading experience that loads instantly over 3G mobile data. By focusing exclusively on high-value affiliate recommendations that genuinely solve problems, I built a lean, highly profitable publishing model that doesn't rely on Google's permission or display ad networks to thrive.
Comments
Post a Comment