Everyone on YouTube right now is promising you that AI blog posts are the fastest path to passive income online. Post a hundred articles, target the right keywords, and watch Google send you thousands of visitors who click affiliate links while you sleep. That pitch sounds simple enough. But does it actually work?
Alston Godbolt decided to find out. Instead of guessing, he hired a virtual assistant, gave the VA a clear process, and ran the experiment live for one full month across 20 separate websites. The results were surprising enough to share in full, including the keyword rankings achieved, the traffic that actually showed up, and the money actually made. Spoiler: those three numbers are not all the same.
What You’ll Walk Out With
- The exact ChatGPT prompting sequence used to produce longer, better-structured blog posts
- How to find low-competition keywords using both a paid tool and a completely free alternative
- The realistic math behind scaling AI content: 4 posts per week, 20 blogs, 4 weeks, 320 posts
- What keyword difficulty score to target when you are starting a brand-new blog
- Why ranking for 500 keywords in month one did not produce any first-page traffic or revenue yet
- An honest breakdown of what “signs of success” mean at the 30-day mark versus month six
- The platform recommendation for hosting and why free site builders are risky for affiliate blogs
- Not sure which online income method fits your background? Take two minutes at finder.platformproof.com
The Claim That’s All Over YouTube
If you have spent any time watching make money online content lately, you have seen the thumbnail: a laptop, a stack of cash, and a headline like “I Made $10,000 in 30 Days With AI Blog Posts.” The creators behind those videos are not necessarily lying, but they are also not showing you the full picture. Some of them started with aged domains that already had authority. Some of them had budgets for backlinks. Some of them filmed the thumbnail before running any real experiment.
The specific question this experiment tests is narrower and more honest: can a brand-new or relatively new blog post, written entirely by AI with minimal human editing, rank on Google and eventually generate affiliate revenue? That is a testable question. One month of data is not enough to give a definitive answer, but it is enough to see early signals, which is exactly what Alston set out to capture.
The other thing worth pointing out before getting into results is that people making big income claims from AI blogging rarely show you their keyword research strategy. They show you the traffic dashboard. They show you the commission screenshot. What they do not show you is how they identified the specific low-competition keyword that made that article rank in the first place. That part of the process is what this video actually teaches.
How the Experiment Was Set Up
Alston hired a virtual assistant specifically for this task. The VA’s job was narrow and repeatable: find the keyword Alston assigned, follow the prompting sequence in ChatGPT, copy the output into WordPress, and publish. No editing. No reading through. No rewriting. Bare minimum output, on purpose.
That last part is important. The decision to skip the editing step was not laziness. It was a deliberate choice to test the floor. If the absolute minimum version of AI content can rank, then adding human editing and original research on top of that floor should only improve results. If the bare minimum fails completely, then the shortcut sellers on YouTube are selling you false hope and you need to know that before you invest time and money into a strategy that will not pay off.
The scale of the experiment: 4 blog posts per week, published across 20 different websites, over 4 weeks. That math works out to 320 blog posts written and published in roughly one month. Not all 20 blogs ran perfectly because the VA ran into some upload issues on certain sites. But the total volume was substantial enough to draw early conclusions about how Google treats AI-generated affiliate content.
The niches covered were intentionally varied. Alston tested the pet niche, specifically targeting “can dogs eat” keyword variations. He tested vacation content. He tested make money online sub-niches. He tested security cameras. He tested soundbars. Spreading across niches that different in topic, audience, and competition gave a broader read on where AI content finds early traction versus where it struggles to get any indexing at all.
Step 1: Finding Low-Competition Keywords, Paid and Free Methods
The keyword research step is the one that most beginners skip or rush through, and it is also the one that determines everything downstream. Publishing an AI article on a high-competition keyword like “best dog food 2024” is not going to rank in month one for a new site. Publishing on “can dogs eat avocado” with a keyword difficulty score of 10 or under is a completely different situation.
Alston uses a paid keyword research tool as his primary method. He pulls up a seed phrase, in this case “can dogs eat,” and then filters keyword difficulty to 10 or less. That filter alone narrows the list from potentially hundreds of keyword options down to the ones where a new blog actually has a realistic shot at showing up in search results at all. Google assigns rankings to every blog post based on quality, relevance, and domain authority, and new sites just do not have the authority to compete on high-difficulty terms in the early months.
For anyone who cannot afford a paid keyword tool right now, there is a free alternative that works. Ahrefs offers a free keyword generator. You can find it by searching Google for “Ahrefs free keyword generator” and it should be the first result. You type in your seed phrase, see keyword suggestions with difficulty scores, and identify targets without spending anything. The free and paid versions return somewhat different keyword sets, which is one of the real trade-offs of free versus paid research. But for a brand-new blogger on a tight budget, the free Ahrefs tool is enough to get started and find your first batch of publishable targets.
The target difficulty range for the best odds of early ranking is 0 to 3. Not just under 10, but ideally in the zero-to-three bucket for your first several months of content. Keywords in this range often still attract meaningful monthly search volume. “Can dogs eat avocado” was the specific example cited in the video, and it draws substantial searches despite being low competition. That combination, decent volume plus very low difficulty, is the exact sweet spot for a new affiliate blog trying to show up in results before it has built any real domain authority.
Step 2: The ChatGPT Prompting Strategy That Produces Longer Posts
Once you have a keyword in hand, the next step is generating the actual content. The mistake most beginners make is typing something like “write a 2,000-word blog post about can dogs eat avocados” into ChatGPT and expecting a complete, well-structured article. That single-prompt approach consistently produces around 1,000 words in practice, and the structure tends to be shallow and repetitive.
The prompting sequence Alston’s VA used has three distinct steps, and each step builds directly on the previous one to produce better and longer output than any single prompt can generate.
Prompt one: generate the title. Take your keyword, in this example “can dogs eat avocados,” and ask ChatGPT to “create blog post title for [keyword].” ChatGPT returns a list of title options. Pick the strongest one and move forward. The title sets the framing and focus for everything that follows, so it is worth taking a few seconds to pick the one that feels most specific and useful to a reader.
Prompt two: generate the outline. Take the title you selected and paste it into ChatGPT with the instruction “write blog post outline for [title].” ChatGPT returns a structured outline with sections and subsections mapped out. This is the skeleton of your article, and it is what allows the next step to produce content that is longer and more detailed than a single shot ever produces. The outline gives the AI a map to follow.
Prompt three: write each section. Copy the full outline and ask ChatGPT to “write blog section for [paste your full outline here].” By giving ChatGPT the complete outline, it writes each paragraph with awareness of the overall structure. The result is more developed paragraphs, a longer total word count, and better internal coherence than you get from a simple “write me a blog post” prompt. You can also go section by section through the outline if you want even more depth per heading.
This three-step approach works with both the free ChatGPT 3.5 and the paid ChatGPT 4. The VA used ChatGPT 4 in this experiment, but the structural process is accessible to anyone with a free account. The difference between the versions shows up in tone and factual accuracy, but the sequence itself is the same regardless of which version you use.
Step 3: Publishing the AI Output Without Editing
After the AI output was generated, the VA copied it directly into WordPress and clicked publish. No spell-check pass. No fact-checking. No human voice added. No reading through to catch errors. The content went live exactly as ChatGPT produced it, on purpose, to test whether the bare minimum approach could produce any results at all.
On the hosting question: Alston recommends self-hosted WordPress over free platforms like Google Sites, Wix, or Squarespace for this kind of affiliate content. The reasoning comes down to control and credibility. On a platform you own, you control all of your monetization options, you are not at risk of the free platform changing its policies or terms of service, and visitors tend to trust a real domain more than a sub-domain running on someone else’s platform. The cost difference is minimal: around $3 per month for basic hosting is enough to get started with your own site.
Whether free platforms like Wix or Google Sites would produce similar Google rankings is genuinely unknown. The video honestly acknowledges this as an open question rather than a settled fact. But for building something you actually own, can fully monetize, and can sell someday if the income is meaningful, self-hosted WordPress is the recommended foundation. The $3 per month cost is the one spending decision that Alston points to for someone starting from scratch.
Not sure which online income method is right for your starting point?
Answer a few quick questions and get a clear recommendation at finder.platformproof.com.
The Results After One Month: 500 Keywords, Zero First-Page Rankings, Zero Dollars
Here is the full, honest scorecard from the first month of the AI blog post experiment.
Keywords ranked for: close to 500. That number genuinely surprised Alston. When you are publishing bare-minimum AI content across low-competition keywords at high volume, Google does pick up on some of it and assign it a position in search results. Close to 500 keyword rankings from 320 posts in one month is a real signal that the approach is not invisible to Google’s indexing systems. Something is working at the floor level.
First-page rankings: zero. None of those 500 keywords are landing on page one of Google yet. They are showing up somewhere in the index, but not in the positions that actually attract clicks. For context, the first organic result on a Google search captures a large portion of all clicks on that query. Results on page two, three, or deeper generate almost no organic traffic in practice, regardless of how many keywords you rank for on those pages.
Actual organic traffic: one or two visitors. Not one or two hundred. Literally one or two people clicked through from Google to the blogs over the entire month across all 20 sites. That is the real-world gap between having keywords indexed somewhere in Google’s results and having keywords that actually send visitors your way at positions people click on.
Revenue: zero dollars. No affiliate commissions. No ad revenue from display ads. Nothing. With one or two total visitors across the whole experiment, that is exactly what you would expect. You cannot generate income from traffic you do not have yet. The absence of revenue at month one is not a surprise and is not a reason to abandon the approach. It is a confirmation that this is a longer game than a single month.
What These Numbers Actually Tell You
The essential context here is timing. One month is, by Alston’s own words in the video, “absolutely nothing” for a brand-new blog post. His first original blog took months before it started seeing meaningful rankings and any traffic at all. The fact that AI content is already indexing and showing up for close to 500 keywords after just 30 days is a positive early signal, not a failure.
The progression most bloggers see with new content on new sites follows a predictable shape. In months one and two, content gets indexed but sits in low positions that nobody clicks. In months three through six, positions start climbing as Google reassesses the relevance of the content and the site earns a small amount of authority from continued publishing. In months six through twelve, if the keywords were chosen well and content quality is sufficient, traffic starts to arrive at meaningful levels. That is the realistic timeline for the approach.
That timing means the real test of this experiment is not what happened in month one. It is what happens in month six when the VA has been consistently writing and publishing across all 20 blogs the entire time. By that point, hundreds more posts will have been indexed, older posts will have had time for Google to evaluate them more thoroughly, and some of the close-to-ranking keywords will likely have climbed into page one territory. That is when Alston expects to see first-page rankings and the affiliate clicks that convert keywords into real income.
The key metric Alston calls a “sign of success” at month one is ranking at all. If you put out 320 posts and Google ignores all of them, that tells you something important: your keyword selection or your site setup has a problem. Getting close to 500 keywords indexed, even in low positions, means Google sees the content as worth cataloging. That is the foundation you need for the longer-term climb.
Honest Drawbacks of the Bare-Minimum AI Approach
The experiment is designed to test the floor, but it is worth being direct about what the bare-minimum approach is missing and what those gaps might cost you over time.
No human editing means no accuracy check. AI content can state things confidently that are simply wrong. In the pet niche, incorrect information about what dogs can safely eat is not just a credibility problem for your blog. It is a potential harm to readers who follow the advice. At scale, publishing completely unedited AI content introduces real risk that you will need to manage as your blogs grow and attract real audiences who act on what they read.
No original research means no differentiation. Google’s ranking systems increasingly reward content that brings something new: original data, personal experience, real-world testing, or expert perspectives. Content that is just a reorganized version of what the AI model was trained on competes in an increasingly crowded field of similar AI-generated content. The low-competition keyword targeting partially compensates for this problem by avoiding the most contested ground, but it does not make the differentiation problem disappear permanently.
Google’s stance on AI content continues to shift. At the time of the video, Google was clearly indexing and ranking AI-written content, as the 500 keyword rankings confirm. But Google has consistently stated that what it rewards is helpful, accurate, people-first content regardless of how it was written. Sites that build a reputation for thin, unedited, mass-produced AI content may find their rankings adjusted as Google’s evaluation capabilities improve. The approach works now at a floor level; whether it continues to work at the same level in year two is not guaranteed.
Zero first-page presence means zero income-generating traffic for now. The 500 keyword rankings are an encouraging sign that Google acknowledges the content exists. They are not 500 keywords sending you actual visitors. If you need income within the next 90 days, this method in its bare-minimum form is not the right tool for that timeline. This is a six-to-twelve-month minimum play, and that is the honest framing the video itself provides.
Find Your X
AI affiliate blogging might be the right model for you, or it might not be. The right fit depends on your patience for slow compounding results, your ability to fund a virtual assistant and hosting across multiple sites, and whether you can stay consistent for six to twelve months before seeing meaningful income. Those are real requirements, not details you can skip.
Take two minutes at finder.platformproof.com to get a direct answer on which online income path actually matches your skills, time, and starting point before you commit months of effort to a method that may not fit your situation.
Frequently Asked Questions
Do AI blog posts actually rank on Google?
Yes, they can. This experiment produced close to 500 keyword rankings from 320 AI-written posts across 20 websites in one month. However, none of those rankings were on the first page of Google, and the total traffic generated was one or two visitors. Google indexes AI content, but reaching page one positions, where the vast majority of clicks actually go, requires time, accumulated domain authority, and typically more than bare-minimum AI output with no human editing.
What keyword difficulty should I target as a new blogger?
Target keywords with a difficulty score of 10 or under, and ideally aim for the 0 to 3 range for your strongest early ranking opportunities. Those very low difficulty keywords can still carry meaningful monthly search volume. “Can dogs eat avocado” was the specific example from this video, and it gets substantial searches despite sitting in the low-competition range. Higher difficulty keywords are a longer game that requires more site authority to compete and is not realistic for a new blog in its first several months.
Can I do this with the free version of ChatGPT?
Yes. The three-prompt sequence described in this video, title first, then outline, then section-by-section writing, works with ChatGPT 3.5, which is the free version available without a subscription. You will get somewhat less polished output than GPT-4 produces, but the structural approach is identical. For a beginner on a budget, start with the free version and reinvest into the paid version once your affiliate revenue starts coming in.
How long does it take for AI blog posts to rank on page one?
Based on this experiment and typical SEO timelines for new blogs, expect six to twelve months before you see consistent first-page traffic from a brand-new blog targeting low-competition keywords. One month is too early to judge the method’s ceiling. The 500 keyword rankings in month one are a positive early signal that content is being indexed, but meaningful page-one traffic and affiliate revenue come later as Google reassesses the content’s relevance and the domain accumulates authority from continued publishing.
Is the free Ahrefs keyword tool good enough to get started?
It is a legitimate starting point for a beginner. The free Ahrefs keyword generator, which you can find by searching “Ahrefs free keyword generator” on Google, gives you keyword suggestions with difficulty scores without any subscription cost. The keyword set it returns will differ somewhat from the paid version, which is one of the honest trade-offs of using a free tool. But for someone building their first content strategy without a budget for paid tools, the free version gives you enough information to identify low-competition targets and start publishing right away.
Should I publish on Wix, Google Sites, or Squarespace instead of WordPress?
Alston recommends self-hosted WordPress. The reasoning: you own the platform, you have full control over your monetization options, and visitors tend to trust a real custom domain over a free sub-domain that runs on someone else’s platform. At around $3 per month for hosting, the cost to go self-hosted is minimal. Whether Wix or Google Sites can produce comparable Google rankings is an honestly open question the video does not claim to answer. But for building an affiliate income asset you control fully and can monetize without platform restrictions, self-hosted WordPress is the cleaner foundation to build on.
How many blog posts do I need before results show up?
This experiment published 320 posts across 20 sites over one month and produced close to 500 keyword rankings but almost zero actual traffic yet. That ratio shows that volume helps with getting indexed but does not shortcut the time Google needs to evaluate and promote content to competitive positions. Rather than fixating on a specific post count, focus on consistent low-competition keyword targeting and publishing on a sustainable regular schedule over six to twelve months. The compounding happens over time, not from a single burst of volume.
What niches work best for AI affiliate blog posts?
This experiment tested the pet niche, vacation content, make money online sub-niches, security cameras, and soundbars, and all of them showed some keyword indexing within the first month. The niches with the clearest affiliate product match tend to have the most straightforward monetization once traffic eventually arrives. The pet “can dogs eat” niche connects directly to pet food and supplement affiliate programs. Tech product comparison content in categories like security cameras and soundbars connects naturally to Amazon Associates and similar programs. Pick a niche where you can clearly see which affiliate products readers would logically buy next.
Read Next
If you are building an affiliate blog and want to understand the broader online income landscape before committing fully to a single method, this post walks through ten years of attempts and honest lessons from building income online from scratch.
I Tried Making Money Online for 10 Years
Sources
- Alston Godbolt, “AI Affiliate Marketing: Do AI Blog Posts Rank On Google?” YouTube, https://youtu.be/SyXSy2JTPH0
- Ahrefs Free Keyword Generator: ahrefs.com/keyword-generator
- ChatGPT by OpenAI: openai.com/chatgpt
- Google Search Central, Creating helpful, reliable, people-first content: developers.google.com/search/docs/fundamentals/creating-helpful-content
Helping 1 million working adults make their first $3,000 online with the skills they already have. Alston Godbolt, Platform Proof.