Remember the days of generating the next sections of text with Jasper? What a pain.
I’ve gone through plenty of iterations of AI usage over the past three years, and I think I finally found a system that works.
It’s not long-form articles giving AI the opportunity to hallucinate.
It’s what I would consider ‘fluff’ listicles, where all you need is a massive list of nouns to generate text.
Why listicles work
Listicles work because it’s not critical information and generative AI is low-cost so you can create en-masse.
Critical information could be classified under YMYL (yo money, yo life), but it also goes further than that. Any information that has the potential to be wrong is in my no-fly zone. If this information is incorrect, will someone or something be harmed? That’s why I’m all for data-fed generative content.
So, what is not critical? Pet names, email subject lines, copywriting headlines, examples of x, or business mottos. Short, sweet, idea generation.
Let’s expand on business mottos. What other short texts do businesses have?
- Mission statement
- Vision statement
- Values
- Tagline
- Slogan
- Brand promise
- Elevator pitch
No one will be harmed if they have the wrong mission statement for their business.
And how many business types are there? Bakeries, landscapers, architecture firms, law firms, sip-and-cycles, etc. The North American Industry Classification System shows around 700 business types exist. That’s a lot of content, and because it’s generative AI, keyword volume doesn’t matter.
Does “Bakery Slogan Examples” get hundreds of searches per month? Doesn’t matter.
And the other reason this works well is because it is inexpensive.
Guess how much content was created with a $128.88 spend in February?
All of that.
About 20 million words worth of content. About 12k posts. At that price point, I don’t give a hoot if nothing on this site gets indexed, because if it does it goes 🚀.
But, oh, wait.
If you’re wondering how 12k posts turned into 200k, I’m translating into every language with TranslatePress.
Editor’s note (also Justin LMAO): translation using TranslatePress & DeepL API also has a cost. The Bulk Publishing Spreadsheet also ran me like $200.
And the trial run I started back in September is starting to show its legs.
The ‘downside’ of this
It’s not all positive.
Monetizing the site above is going to be a pain. I’ve repeatedly been denied Adsense, so I just put up a few Adsterra ads towards the end of February. I don’t think I’ll get an RPM above $10 for this site.
But at the same time, this has the pOtEnTiAl to get an ignorant amount of traffic. I’ll take a $10 RPM if the site’s getting a million monthly visitors.
If that’s the only downside, we’ll be alright.