Alston Godbolt watched a video from Marcus Campbell, the affiliate marketing dude, that claimed you could use ChatGPT 4o to hack your way onto the first page of Google. The strategy sounded simple on paper: find expired domains, pull their old URLs, feed them into ChatGPT, and publish the rewritten posts. Three posts a day, hands off, cheap to run. The question Alston had was whether it actually works in real life.
He had the perfect test case sitting right in front of him. Two years earlier, his blog at alstongodbolt.com was ranking for over 82 keywords. Then malware and spyware hit his VPS, took down every site on the server, and wiped out every ranking he had built. The site sat collecting dust for nearly two years. This method gave him a reason to try to get it back. Six days and 18 posts later, the data is in. Here is what he found.
What You’ll Walk Out With
- The exact expired-domain plus ChatGPT method Marcus Campbell teaches, broken down step by step
- Why Alston used his own dormant blog as the test bed instead of starting from scratch
- The specific ChatGPT 4o prompting process that makes this 90% hands off
- The “major in the minors” keyword strategy that lets a low-authority site actually show up on Google
- Real Google Search Console data from six days of testing, including first impressions and early keyword wins
- The honest cost breakdown, including what you cannot skip if you want this to work
- The one drawback Alston says most creators glossing over this method completely ignore
- Not sure which method fits your skills and situation? Find out in about two minutes at finder.platformproof.com
The Marcus Campbell Method That Started This Test
Marcus Campbell, known on YouTube as the affiliate marketing dude, posted a video showing a specific workflow for getting AI-written blog posts to rank on Google. The core idea is not brand new, but the twist he added makes it easier than the traditional version. Instead of brainstorming keywords from scratch, you find expired domains that already had ranked content in the past. Those domains have a history. Google once trusted them. The web pages on those domains were once showing up in search results for real queries.
The workflow Marcus lays out goes like this. You go find an expired domain using a tool like Expireddomains.net or a similar domain search site. You find a domain that has historical backlinks, a real niche, and old blog posts you can look up. Then you pull the URLs from those old posts, paste them into ChatGPT 4o, and tell the AI to rewrite each post as a fresh article based on the keyword in the URL. You copy the output, paste it into your website, and publish. That is it. No original research required, no unique insight needed, just a structured prompting workflow and a consistent publishing schedule.
The reason this catches people’s attention is the cost model. Writing a solid blog post used to mean paying a writer. Alston mentioned that he previously paid around $30 per post for well-written affiliate content. At three posts a day, that would run $90 a day or close to $2,700 a month. With ChatGPT doing the writing, that cost drops to nearly zero for the content production itself. The question is whether the output is good enough to rank.
Why Alston Had a Dead Blog Ready to Revive
Most people testing this method would start with a fresh expired domain they just bought. Alston had something better: a domain with real history that he actually built. Back in June of 2023, his site was tracking around 82 to 100 keywords in Ahrefs, a paid keyword research tool. The site had a domain rating of 15 and real backlinks. It was small but real, and it was moving in the right direction.
Then his virtual private server got hit with malware and spyware. Not just his site but every site hosted on that server went down. The spam and hack took out all his rankings. By December 2023, the keyword count had dropped to zero. Google Search Console was showing 403 error codes across multiple pages. The site was essentially a ghost, technically alive but invisible to search engines and readers alike.
When Alston saw Marcus Campbell’s video, he realized he did not need to start from scratch. He had a domain with a DR of 15 and an existing backlink profile. It had once ranked for dozens of keywords. The malware had wiped out the content and the rankings, but the domain authority was still there. Using the ChatGPT method to quickly rebuild the posts that used to rank was a direct application of the exact strategy Marcus described, just with a domain Alston already owned and understood.
How the ChatGPT 4o Process Actually Works
The process Alston followed is exactly what Marcus laid out. He took old URLs from his site, like alstongodbolt.com/harbor-freight-affiliate-program-plus-10-ways-to-make-money, and pasted them directly into ChatGPT 4o. The AI reads the URL, infers the keyword target from the slug, and writes a full blog post around that topic. You give it a brief outline or some direction if you want, but you do not have to. ChatGPT will generate a structured article with sections, tips, and affiliate program details on its own.
Once ChatGPT finishes the first section, you type “keep going” and it continues until the full post is done. Then you copy it, paste it into WordPress or your CMS, and hit publish. The whole process for one post takes maybe five to ten minutes of active time. The rest is waiting for ChatGPT to write. That is what Alston means when he says the process is about 90% hands off. You are not writing, you are cueing the AI and managing the queue.
One issue Alston ran into is post length. He wants his blog posts to be around 2,000 words, which is a standard threshold for affiliate content trying to rank. ChatGPT often produces closer to 1,000 words in a single run. You can push it to keep going and expand sections, but you will need to prompt it specifically to hit a longer output. For a first pass this is not a dealbreaker, but it is something to watch if you care about hitting a word count target for SEO reasons.
The Upload Strategy: Why Three Posts Per Day Is the Safe Number
When Alston started using this method, one of the first questions he had to answer was: how many posts can I publish per day before Google gets suspicious? There are sites out there publishing 30 AI-generated posts a day. Alston decided that was too aggressive for where his site stands. He settled on three posts per day as his limit, and he explains his reasoning clearly: three posts a day looks normal. It is the kind of publishing pace a real human blogger could maintain. It does not signal to Google that something automated is happening at scale.
Over the six days he tested this, he published 18 posts total, sticking to three per day. He keeps a queue of drafts ready so that he can publish consistently without having to generate a new post in real time every day. Whenever he has a few minutes between other tasks, he opens ChatGPT, generates a new draft, saves it, and adds it to the queue. Then each day he pulls three from the queue, reviews them briefly, and hits publish. This keeps the process sustainable without requiring a dedicated block of time every day.
Major in the Minors: The Keyword Strategy That Actually Works
The most important concept from Alston’s test is not the ChatGPT workflow itself. It is the keyword targeting strategy that determines whether any of this works. He calls it “major in the minors,” a phrase Marcus Campbell used that stuck with him. The idea is simple: stop chasing big keywords and win consistently on small ones.
Here is the problem with big keywords. “Amazon affiliate program” gets 86,000 searches per month. That sounds incredible. But the keyword difficulty is 41, which means established sites with massive domain authority are competing for it. A new or recovering blog with a DR of 15 has no realistic shot at ranking for that term in the near future. Going after it is a waste of time and content budget.
Now look at what happens when you filter by keyword difficulty of 2 or less. “Home Depot affiliate program” gets 16,000 searches per month with a KD of 2. “Sweetwater affiliate program” is searchable with almost no competition. “Farfetch affiliate program” brings in around 600 searches per month. “Shein affiliate program” gets around 400. These numbers look small in isolation, but think about what happens when you rank for 50 or 60 of these low-difficulty terms at the same time.
If 10% of the people searching “Home Depot affiliate program” click on your post, that is 1,600 visitors from one keyword. Multiply that by dozens of similar low-competition keywords and you are looking at real traffic from people who are actively interested in affiliate programs, which means they are already warm buyers. Alston used this exact strategy before the malware incident, and it was working. He was ranking for many of these small keywords simultaneously and generating real traffic because of it.
The base hit metaphor he uses is accurate. You are not trying to hit home runs. You are trying to get on base over and over. Stack enough base hits and you are scoring runs whether or not you ever hit one out of the park. Most content creators chase the big keyword and wonder why they never rank. The ones who win with this method target the keywords nobody else bothers with and publish them consistently at scale.
Six Days of Data: What the Google Search Console Actually Showed
After six days and 18 published posts, Alston pulled up his Google Search Console to show the results. He was honest that the numbers are small, but the direction matters. He got his first impression on February 23rd. That is a signal that Google has indexed at least one of the new posts and served it in a search result somewhere. Small, yes. But it was zero before he started.
More importantly, he is starting to see rankings for specific low-difficulty keywords. The Poway Barn affiliate program and the Sweetwater affiliate program both showed up in his data as early keywords where his posts are appearing in results. These are not ranking number one yet, but they are in the index and showing up. For a site that had been sitting at zero keywords for over a year, that is genuine progress in less than a week.
When Alston zooms out in Search Console to show the last 16 months, you can see the full picture. There was solid activity two-plus years ago when the original blog was thriving. Then the drop to zero when the malware hit. And now, at the right edge of the chart, a tiny uptick beginning. The site is not back to where it was yet, but the curve is pointing the right direction for the first time in nearly two years. He made no money from this in the first six days, but he was clear that six days is not enough time to expect revenue. The goal right now is indexing and early rankings.
Not sure which platform or method fits where you are right now?
Take the two-minute finder quiz at finder.platformproof.com and get a specific starting point matched to your skills, schedule, and income goal.
The Real Cost to Start This Method
One of the things Alston is direct about is cost. He knows a lot of people watching his channel are looking for completely free methods, and he is honest that this one is not free. Here is a realistic budget breakdown based on what he shared.
First, you need a domain. If you are buying an expired domain with existing backlinks and some history, expect to pay anywhere from $20 on the low end to several hundred dollars depending on the domain’s authority and niche relevance. Alston specifically says you can start with expired domains under $20, which is a real option if you filter by price on sites like Expireddomains.net. If you have your own domain with a history like Alston does, you skip this cost entirely.
Second, you need web hosting. A basic shared hosting plan can run $5 to $10 per month. A VPS, which is what Alston was using before the malware hit, costs more but gives you more control and speed. For someone just starting out, shared hosting is fine. For someone scaling to hundreds of posts, a VPS or managed WordPress host makes more sense.
Third, and this is the one that catches people off guard, you may need a paid keyword research tool. Alston used Ahrefs to see what keywords his domain used to rank for and to find which new keywords had a difficulty of 2 or less. Ahrefs starts at around $99 per month. That is not cheap. There are lower-cost alternatives like Ubersuggest or LowFruits.io that can show keyword difficulty at a fraction of the price, but you will need something. Free tools like Google Keyword Planner do not show difficulty scores in a way that makes low-competition filtering easy.
Alston estimates you need around $100 to get started realistically, not counting the keyword tool if you find a workaround. If you include a month of Ahrefs or a similar tool, your startup costs are closer to $200. That is still a low bar for a business with the potential to generate recurring affiliate income at scale.
Honest Drawbacks
Alston does not oversell this method. He lays out the real friction points clearly, and they are worth understanding before you decide whether to try it.
The first drawback is that this is not free. Almost every person asking about this method is hoping it costs nothing. It does not. You need a domain, hosting, and some kind of keyword tool. If you go in expecting to spend zero, you will hit a wall before you publish your first post.
The second drawback is that AI-written content is not always good. Google has stated repeatedly that it evaluates content on quality, not just on whether AI was used to write it. A 1,000-word AI post with generic affiliate tips and no unique angle can get indexed and never rank, or rank temporarily and then drop when Google updates its quality signals. Alston is aware of this risk, which is why he is pacing at three posts per day rather than flooding the site with 30 a day. But the quality ceiling on pure ChatGPT output without editing is a real concern for long-term ranking performance.
The third drawback is that it takes time to see results. Alston was six days in with no money earned yet. That is not a failure, that is just reality. SEO does not pay on day one or even day 30 in most cases. If you need income immediately, this method will not solve that problem. It is a slow-build strategy that starts generating returns after months of consistent publishing and indexing.
The fourth drawback is the keyword research tool cost. If you do not have access to Ahrefs or a comparable tool, finding keywords with a difficulty of 2 or less is guesswork. You might get lucky, but you might also spend weeks publishing content that goes after terms that are too competitive. The paid tool is what makes the low-difficulty filtering reliable.
Find Your X
The ChatGPT bulk content method works best when it is matched to the right situation. If you have an expired domain with a real history, some budget for tools and hosting, and the patience to publish consistently for several months, this is a legitimate way to build affiliate income at low ongoing cost. If you are starting with no domain, no budget, and a need for fast income, you may need a different starting point.
The two-minute quiz at finder.platformproof.com matches your specific skills, budget, and schedule to the online income method most likely to work for you. Instead of guessing where to start, let the quiz narrow it down so you can move faster with less wasted effort.
Frequently Asked Questions
Do you need to buy an expired domain or can you use a brand new one?
You can technically use a brand new domain, but you will wait much longer to see any results. Expired domains already have backlinks and some level of domain authority from their previous life, which gives Google a reason to trust them sooner. A brand new domain starts with zero trust and can take six months to a year to rank for anything consistently. If you can find an expired domain for under $20 in a relevant niche, it is worth it for the head start.
How does ChatGPT know what to write when you paste in a URL?
ChatGPT 4o can read the slug and structure of a URL and infer the topic from it. A URL like /harbor-freight-affiliate-program-ways-to-make-money tells the AI the target keyword and the angle. It will write a post covering the Harbor Freight affiliate program, how it works, the commission rate, and tips for promoting it. You can also give it a short outline to guide the structure. The AI does the heavy lifting; you just direct it and clean up the output.
Is Google penalizing AI-generated blog content?
Google’s official stance is that it does not penalize content for being AI-generated. It penalizes content for being low quality, thin, or spammy regardless of how it was produced. The practical risk with bulk AI content is that it can be generic and repetitive, which does not serve users well. The sites that have gotten hit by Google updates tend to be ones publishing massive volumes of low-effort AI content with no editing or added value. Publishing at a moderate pace, targeting specific low-competition keywords, and doing a basic review pass before hitting publish reduces that risk significantly.
Where do you find expired domains with good backlink history?
Marcus Campbell mentions a domain search site in his original video. Alston referenced it as “Zilla” but could not recall the exact name during filming. Expireddomains.net is a well-known free option that shows domain age, backlink counts, and category filters. GoDaddy Auctions and Namecheap also list expired domains with bidding options. For more specific backlink quality data, you would cross-reference the domain in Ahrefs or Moz to check the domain rating and the actual backlink profile before buying.
What keyword difficulty should you target when just starting out?
Alston filters for keyword difficulty of 2 or less when looking for targets on his recovering domain. For a brand new domain with no authority at all, you might want to start with KD 1 or even 0, which are the truly uncontested searches. As your site builds authority over months of publishing, you can gradually move up to KD 5 or KD 10 targets. The key is patience. Going after KD 40+ terms right out of the gate is how people spend months publishing and never see a single click from organic search.
How long does it take before you start making affiliate money from this?
Alston was honest that he made no money in the first six days. Realistically, most people see their first affiliate commissions from this method somewhere between three and six months in, assuming they are publishing consistently and targeting low-competition keywords. The timeline depends heavily on domain authority, niche competitiveness, and how well the affiliate programs convert. Some niches convert at 5% to 10%, meaning 1,000 visitors can produce 50 to 100 affiliate clicks. Others convert much lower. Research the affiliate programs in your niche before investing months of publishing into a program that pays nothing.
Can you do this method completely for free?
Not really. Alston is clear that you will need to spend some money. A domain, hosting, and a basic keyword tool add up to at least $50 to $100 to start. ChatGPT itself has a free tier, so the content generation is where you save money compared to hiring writers. But the domain and hosting are non-negotiable if you want to build something that grows and that you actually own. Platforms like Medium or Substack let you publish free, but you do not control those sites and cannot benefit from domain authority the same way.
What should you do if you do not have an old domain to revive?
Alston mentioned an idea that Marcus Campbell did not cover in the original video: you can approach businesses that have dead or inactive domains and offer to write content for them in exchange for access to the domain or a revenue share. If a local business registered a domain years ago, built a few pages, and then stopped updating it, they may welcome someone taking it over and building it back up. This is a creative way to access an aged domain without buying one outright. You bring the content production, they provide the domain history and backlinks that already exist.
Read Next
If you want to see how AI tools are being used right now to build income streams beyond affiliate blogging, the next post covers five specific methods people are using to generate $3,000 or more per month with AI assistance, many of them starting with just a laptop and a few free tools.
5 AI Side Hustles That Pay $3,000 Per Month
Sources
- Original method by Marcus Campbell, the affiliate marketing dude, YouTube
- Google Search Console data from alstongodbolt.com (February 2025, six days of new posts)
- Ahrefs keyword data shown in the video: Home Depot affiliate program (16,000 searches/month, KD 2), Amazon affiliate program (86,000 searches/month, KD 41)
- Alston Godbolt, Platform Proof, I Tried It series, YouTube
Helping 1 million working adults make their first $3,000 online with the skills they already have. Alston Godbolt, Platform Proof.