In my very first ChatGPT prompt, I asked it to write a funny Christmas limerick about my wife. It was December 2022, roughly three weeks after OpenAI unveiled the technology that threatened to erase my livelihood. It did indeed return a limerick about both Christmas and my wife. And it was… moderately funny.
I shuddered. The response might as well have ended with a postscript: “Too bad you didn’t study something useful like engineering.” On a spectrum between publishable poetry and crumpled-draft-in-garbage, the robot returned a composition I’d consider memorable–but embarrassing if read by others.
As a professional writer, I didn’t fear the quality of this particular response. But I did fear the acceleration. That’s the promise of Artificial Intelligence, after all: exponential learning and improvement. If it was this good three weeks after launch, where would we be in six months?
That’s probably why my night sweats started.
We’re now almost two years into the generative AI revolution. As marketers, the tech has undoubtedly changed the way we work forever. We use it every day. For certain tasks, it’s absolutely incredible tech. Yet, for all that’s changed, we’re finally acknowledging one thing: AI is terrible at writing. It’s not getting better. And the volume of that writing is making internet search far less useful.
That’s why Google–as the (nearly) omnipotent arbiter of our collective search experience–has taken drastic action in its March algorithm update to clean up the internet’s AI litter. As a result, the update gives marketers the clearest picture yet on the acceptable and unacceptable uses of AI for SEO.
Not to mention, the update also helped me sleep great last night.
Google's New Approach to AI
3 Takeaways from Google’s March Algorithm Update
On March 5, Google rolled out a Core Update that’s had massive ramifications on what sites it indexes, and where they rank. Here’s what the company said in its release, emphasis mine:
“We’re enhancing Search so you see more useful information, and fewer results that feel made for search engines.”
Translation: AI is not an SEO cheat code.
Here are three key takeaways from the update:
Useful Content is Better than a Lot of Content
There’s no doubt volume is a key component of SEO success. Sites that publish fresh information, regularly, will rank higher than those that remain static. When GenAI arrived, it provided too tempting a loophole. Why bother spending hours on a single blog when ChatGPT would write 500 or 5,000 in the same amount of time? Flood keywords with content, and the traffic will follow.
Instead, we drowned. SEO experts agree that Google’s search results are getting worse. This latest update, then, is meant to be a course correction. Here’s what Google has to say:
"This update involves refining some of our core ranking systems to help us better understand if webpages are unhelpful, have a poor user experience or feel like they were created for search engines instead of people. This could include sites created primarily to match very specific search queries.
We believe these updates will reduce the amount of low-quality content on Search and send more traffic to helpful and high-quality sites. Based on our evaluations, we expect that the combination of this update and our previous efforts will collectively reduce low-quality, unoriginal content in search results by 40%."
In just three weeks, the company has dropped the hammer.
Whole Websites are Disappearing
Google isn’t playing around. According to Search Engine Land, they have completely de-indexed (removed from their search results) at least 10 websites (albeit malicious and spammy ones) that each used to garner millions of clicks every month.
Beyond that, it’s penalizing large swaths of reputable sites that tried to game the system–removing content or lowering results that it no longer deems original or “useful.”
AI is “Technically” Still Fine to Use (but Proceed With Caution)
Oddly, Google’s own blog announcing the changes doesn’t mention the word “Artificial Intelligence” a single time. It’s not changing its policy–from 2023–that allows for AI-produced content, as long as that content is useful.
But the subtext of this update is clear. High-volume, poorly-written content will be penalized.
Using AI to write content really only gives you one advantage: It allows you to produce high-volume, poorly-written content.
Go ahead and connect the dots.
It’s still unknown how good the Google algorithm even is at identifying–at the granular level–which content was authored by AI and which was authored by a human. It likely can’t tell every single word composed by a GPT.
What it can do? Flag drastic shifts in volume. If you’re a reputable site that suddenly publishes significantly more articles, you will likely get penalized.
There’s only one fundamental (algorithm-proof) law in SEO: High-quality, useful content always wins.
Who is the AI Writing Even For?
But let’s pretend spewing AI-produced articles into the ether actually worked. I’d be out of a job and my robot friend would have produced this blog. Here, with no editing, is how ChatGPT would start.
“In the ever-evolving landscape of technology and innovation, one constant remains: the profound impact of artificial intelligence (AI). Recently, Google, one of the foremost pioneers in AI research and development, made a significant proclamation that reverberated throughout the tech community and beyond. With a decisive statement, Google declared that AI is no longer merely a cheat code – it's a fundamental tool reshaping the way we approach problems, unlock insights, and drive progress.”
That’s just… terrible writing.
Now, imagine you’re a potential client vetting a marketing agency that proclaims on its website to produce world-class content marketing (we do). You find this article at the top of your Google search (because the AI worked. Awesome!), and you start to read. You’d immediately lose trust in our content credentials. You’d think we hired an eighth grader who just discovered a thesaurus. You’d never hire us.
Too many marketers think SEO is the finish line. Any tactic that brings their site to the top of the results page is a win. They believe their job is done when you’ve clicked. But it’s not. The job is done when the sale closes. That doesn’t happen without… You guessed it. High-quality, useful content.
Imagine if Venice used its best real estate to open a frozen yogurt shop. The product looks kind of like gelato and feels the same temperature as gelato, so a lot of people queue up because they think they’re getting gelato.
If you exclusively use AI to produce content so that you will rank higher in Google, you might get a lot of people to come to your website. But when they arrive, they won’t linger. You sold them (metaphorical) frozen yogurt. They wanted gelato.
AI writing is not for humans. AI writing is for an algorithm. But the algorithm doesn’t buy from you. Humans do. Logically then, this means your writing should be for humans, not the algorithm.
And humans will be able to distinguish frozen yogurt from gelato–every time.
How We Actually Use AI in Our Writing
I’d be naive (and arrogant) to say good writers should never use AI. That would be like trying to sell you an artisanal, organic, grass-fed mode of transportation after our competitors started selling the internal combustion engine.
There really are great, efficient ways in which AI can aid marketers who write. Soup-to-nuts SEO content execution isn’t one. But here are the most useful AI tasks that our content team uses:
- Brainstorming: I recently needed a real-world example of a wonky scientific phenomenon. The simple ChatGPT prompt, “Give me a practical example of the Linear Viscoelastic Region,” returned a fantastic, intelligible result. The writing still sucked, but even with a full edit, it saved me plenty of time searching for ideas on my own.
- Copyediting: AI has become great at catching grammatical mistakes. Not just spelling, but passive voice, dangling participles, and other granular errors that normally take significant chunks of editing time.
- Summarizing: We convert all our blogs into emails to share with potential clients. When we paste a full blog into a GPT, AI spits out a decent summary. We still need to edit the finished product, but it saves us from starting with a blank page.
- Organizing: Last week, I needed to re-order more than 100 terms alphabetically. By hand, it would have taken me 2 hours. ChatGPT did it in 30 seconds.
Conclusion: Maybe The Cheat Code Was Good Content All Along
Google has many reasons (48 billion of them, to be exact), to provide you with the best results possible. An influx of terrible, AI-produced content has turned search results pages into a pretty miserable user experience.
Google’s March update is a business decision. They want to make Google better, so you continue to use it, so they continue to print money. In doing so, they also gave marketers the clearest proclamation yet that artificial intelligence cannot produce high-quality, useful content by itself.
Sure, that’s encouraging news for copywriters who still have some job security. But I’d argue that it’s better news for everyone.
It means the internet just got much better. Coincidentally, so did my sleep.
Ready to elevate your B2B marketing?
We help leading business-to-business brands hit their marketing goals. Get in touch to learn how Altitude Marketing can help you reach your peak performance.
