We all use large language models (LLMs) like ChatGPT.
As a content marketer, I use it every day for research, blog briefs, ideation, and writing support.
But even though I use it often, I never publish a ChatGPT blog without editing it first. Here's why.
ChatGPT-Written Blogs (Without Edits) May Harm Your Rankings
To be clear: I don’t have an issue with using ChatGPT for writing. But it seriously bothers me when I see a company using articles straight from LLMs without edits.
It makes me lose trust in the brand. They aren’t creating content for their audience; they’re pumping out as much as possible to rank in search. The irony? Doing so can actually worsen your SEO performance.
Sure, Google won’t penalize AI-written blog content. But it does downrank content that doesn’t show E-E-A-T or if it’s written for search engines—not humans. According to Search Engine Journal, “using automation or AI strictly to manipulate rankings in search results is considered a violation of Google's spam policies.”
Google also prefers content with inbound and outbound links to make it trustworthy. (Trustworthiness is the most important E-E-A-T factor, after all.) As of right now, the free version of ChatGPT can’t quite do that.
As someone who reads and engages with a ton of blog content, I know a ChatGPT blog when I see it.
Here's how you can tell if a blog was written by ChatGPT—without edits from a human.
How To Tell If a Blog Was Written By ChatGPT
#1. A distinct introductory sentence.
A ChatGPT blog almost exclusively opens with a variation of “In today’s world…” Especially if the content is about a fast-paced industry like tech.
Other examples include:
“In today’s digital landscape”
“In the bustling world of [insert subject]”
“In the dynamic landscape of [insert subject]”
“In today’s fast-paced digital world”
“In the realm of [insert subject]”
As soon as I see this opener, my brain checks out. I stop reading, and I’m sure you do too. News flash: low time on page doesn’t bode well for search rankings.
To get your audience to read the rest of the article, you need to entice them. Start your introduction with a statistic. Or a fact. Or skip the introduction altogether.
But whenever possible, avoid an opening sentence like the examples above. Cut them out of your ChatGPT blog—they won’t get you anywhere.
#2. An obvious introduction.
Another telltale sign in the introduction? When the last introductory sentence begins with: “In this article, we’ll talk about x, y, and z.”
See the screenshot below for an example.
This writing style is common in academia but serves little purpose for informal blogs. Humans have short attention spans. Meet them where they’re at. Skip the “in this article” line and dive right in.
#3. Specific lists.
A ChatGPT blog also has one too many bullet points/numbered lists. And they’re all formatted the same way: with bolded, title-case introductory text.
See an example below.
Bullet Point Sentence 1: The first sentence about bullet points.
Bullet Point Sentence 2: The second sentence about bullet points.
Bullet Point Sentence 3: The third sentence about bullet points.
Case in point: the ChatGPT conversation below. (Note the classic “realm of digital marketing” line, too.)
The semi-colon is fine, but the capitalized letters aren’t necessary. Usually, informal writing styles might format bullet points like this:
Bullet point sentence 1: The first sentence about bullet points.
Bullet point sentence 2: The second sentence about bullet points.
Bullet Point sentence 3: The third sentence about bullet points.
Notice the difference? It reads better. To summarize: add bullet points, but cater them to your (most likely informal) audience.
#4. Redundant descriptive words.
ChatGPT also uses too many descriptive words that all mean the same thing.
Take a look at this sentence from the blog outline above:
“High-quality blog content has the potential to attract backlinks from other websites, particularly if it provides valuable information, insights, or solutions to common problems.”
Saying “information, insights, or solutions” isn’t inherently bad. But it’s wordy. Most writers would use one of them for brevity, but for some reason, ChatGPT uses all three.
At the end of the day, these terms are redundant. Edit your ChatGPT blog down to one—both to avoid looking like an LLM and to improve your writing.
#5. No links.
ChatGPT also doesn’t provide direct links for sources (at least in its free version). You can ask for general sources at the end, but it won’t show you what each one supports in the article.
See an example below:
Also, each source is also from before 2021—not ideal from a content marketing perspective.
For now, ChatGPT notoriously pulls content from before 2022. And if you're in a fast-paced industry like tech, that's a major problem if you’re trying to write with ChatGPT alone.
Your information won't be up-to-date. And to demonstrate E-E-A-T, it needs to be.
I’m sure OpenAI will improve this in the future. But for now, a lack of up-to-date information is a glaring problem for content managers using ChatGPT without editing its output.
#6. A conclusion statement.
Last but not least, almost every ChatGPT-written article ends with: “In conclusion…”
Like the introductory cues above, readers don’t need a sentence telling them that an article is about to end. Put simply, it’s boring and unnecessary.
Content managers, take the extra minute and remove it. You don’t need it, or even a conclusion at all.
How to Write Better Blog Content Without ChatGPT
I previously wrote about how marketers can use ChatGPT and human effort to create search-optimized content. Here’s a step-by-step guide.
Another tip? Hire a content and SEO expert who actually knows what they’re doing. If your content manager is only pumping out ChatGPT articles, they aren’t helping your brand or your SEO. In fact, they might be making both worse.
Looking for content and SEO support? Get in touch.
Comments