ChatGPT for SEO FAQ: Should You Use AI for Creating Content?
What Are the Implications of Using ChatGPT for SEO?
Like insects to a bug-lamp, SEOs everywhere are swarming to this newfound option to streamline otherwise resource-consuming tasks. Some sites are taking advantage of ChatGPT to publish articles written entirely by AI—but at what cost?
As a content manager, I think AI-generated content is the bane of our existence. And while I despise those using AI to generate thousands of robot-written articles, I’ve found myself having my own love affair with the technology. I’m constantly reviewing the implications of ChatGPT and pondering the most practical use cases for SEO.
Let’s explore the opportunities, risks, and ways to keep our hats white using AI for SEO.
What is ChatGPT?
ChatGPT is like a voiceless Siri on steroids. It’s essentially a website developed by OpenAI that follows instructions typed in by users and can generate answers to questions.
The GPT stands for Generative Pre-trained Transformer. It’s a natural language processing model (NLP) trained on a wide array of published data to produce responses that look like they were written by people answering questions or giving instructions.
Greg Brockman, the president, chairman, and founder of OpenAI explained how ChatGPT was trained on over 40 terabytes of text—which includes information available on the internet from Reddit, Twitter, Wikipedia, books, and more.
The software offers a human-like conversation while following instructions, and even gives the impression of having short-term memory by recalling data from earlier in the thread.
When asked to “Sarcastically describe ChatGPT,” the program replied:
“ChatGPT, the epitome of intelligence and wit. Just your average, everyday language model that can hold a conversation and solve all the world’s problems. Nothing to see here, folks. Just a computer program that can mimic human thought and communication. No big deal.”
What is Google’s position on using ChatGPT for content creation?
Before we take advantage of the chance to crank out copious amounts of chatbot-assisted copy quickly, we must consider Google’s webmaster guidelines. So how does Google’s webspam team feel about content created using AI tools? Let’s look at how the discussion about leveraging automatically generated content is going.
Google says the appropriate use of AI-generated content is okay
In February 2023, Google published guidance about AI-generated content. It says its primary objective is rewarding high-quality content however it’s produced. The guidance says the appropriate use of AI to generate content is not against its guidelines. The primary concern is about content strategies used specifically for manipulating search results:
“If you see AI as an essential way to help you produce content that is helpful and original, it might be useful to consider. If you see AI as an inexpensive, easy way to game search engine rankings, then no.”
While the company is highlighting the importance of avoiding any content strategy scale just to rank in search, they won’t be checking or penalizing for it at this time. Google’s focus is content that satisfies the aspects of E-E-A-T.
The company is welcoming AI with open arms, expressing their understanding of the value of the new and exciting technology by explaining:
“Automation has long been used in publishing to create useful content. AI can assist with and generate useful content in exciting new ways.”
It’s encouraging to see Google’s position shift in the direction against spam instead of treating all auto-generated content as spam.
How has Google’s opinion about AI changed?
Google was advising against automatically generated content from the start.
In April 2022, Google’s Search Advocate John Mueller stated content automatically generated with AI tools is spam:
“It’s still automatically generated content, and that means for us it’s still against the Webmaster Guidelines. So we would consider that to be spam.”
In October 2022, Google added a section on auto-generated content to the spam policies for developers on their site that reads:
“Spammy automatically generated (or “auto-generated”) content is content that’s been generated programmatically without producing anything original or adding sufficient value.”
Interestingly, they added the word spammy a month later. Was this their first sign of acknowledging there is such a thing as non-spammy AI-generated content?
In November 2022, Google’s Search Liason Danny Sullivan clarified Google’s position on AI by tweeting:
“We haven’t said AI content is bad. We’ve said, pretty clearly, content written primarily for search engines rather than humans is the issue. That’s what we’re focused on. If someone fires up 100 humans to write content just to rank, or fires up a spinner, or a AI, same issue…”
So they’ve always had a clear position against creating content specifically for search engines. Sullivan expanded on his statement in a reply:
“We did talk about a focus on content *by people* for people in our post about improvements like the helpful content system. But the nuance is really that it’s unlikely some AI content is going to feel written by people without some degree of human review:”
In January 2023, the Google Search Liason Twitter account further clarified its position in a reply to a thread about Bankrate’s decision to use AI for generating content. It stated:
“For anyone who uses *any method* to generate a lot content primarily for search rankings, our core systems look at many signals to reward content clearly demonstrating E-E-A-T (experience, expertise, authoritativeness, and trustworthiness)”
At present, their primary goal is rewarding helpful content created for people—with at least some degree of human review.
Can Google detect AI content?
Google hasn’t claimed its ability to understand the difference between human and AI content.
The latest version of ChatGPT is based on version 3.5. But a great deal of the output is still detectable by GTP-2 output scanners. You’re probably not fooling Google by adjusting the copy until it clears outdated detection tools.
Despite detection-evading efforts like changing words, intentional misspellings, and other strategies, researchers have shown that AI-generated text is still identifiable. It is, however, becoming harder to identify, and some older methods of checking are less reliable than before.
OpenAI recently introduced its new AI text classifier, a tool for detecting AI written content. Keep in mind it’s less than perfect and doesn’t have complete accuracy at this time.
It’s possible detecting AI is a weak point for Google. Although, it decided to embrace the new technology in search and content creation.
What about websites already using AI to create content?
It’s no secret that some of the most authoritative news sites use robot writers for their content.
The Washington Post began publishing AI-generated articles years ago using a bot they call Heliograf. Bloomberg previously reported a third of its articles are drafted by its bot named Cyborg. Forbes has Bertie for prewriting first drafts and streamlining the creative process for reporters. And CNET is trying its luck with publishing articles written entirely by AI, however, it paused its AI efforts citing discomfort from the backlash of getting caught.
Sites like these are regarded as experts in their respective fields. They have a team of editors manually reviewing and editing articles. Since their content undergoes human review, adds value, and is regarded as helpful—it’s unlikely this will be regarded as spam.
What are the limitations of chatbots for content?
While chatbots are certainly improving, they’re not without shortcomings. The biggest criticisms of AI output are trustworthiness, quality, and originality.
Automated content must be fact-checked
One of its main criticisms is the inaccuracy of data. The bots have an impressive way of writing as if they know what they’re talking about, even if the data is rubbish. They’re unable to vet their own responses for accuracy—unless they’re instructed to with the correct information.
To put it simply, chatbots don’t always know when to say they don’t know. They fake it ‘til you make it. Robots lack judgment.
AI-generated content is low quality
While there have been significant strides in personalizing the written conversations between the software and people, the quality of the output still raises eyebrows. The bots have a tendency to be repetitive, speak in circles, and use redundant filler language. Answers to questions can look just like the original input only slightly reworded.
The generic output is unlikely to satisfy Google’s appetite for E-E-A-T content.
The chatbot output is unoriginal
The software is incapable of creating its own new discoveries. It can’t tell stories, interview experts, or complete a unique critical analysis.
Chatbots are like state-of-the-art copy machines for content on the internet designed to make collages of data that look unique. All things considered, AI-generated output is created by shuffling up previously published material in a cute and convincing fashion.
They rely on preexisting content and lack the intelligence to fact-check themselves, let alone spawn brand-new material. Chatbots are thought repeaters, not thought leaders.
The information from bots can be outdated
The freshness of the data depends on how the chatbots are trained. ChatGPT-3.5 isn’t connected to the internet for its answers. Instead, it was trained on data from 2021 and before. The output is limited to the past and results are not populated in real time.
It’s almost like using a fancy version of the Wayback Machine to search older archived files for answers to questions you have today. By contrast, search engines are pulling fresh results from the internet immediately as new information becomes available.
It’s not likely to remain this way forever as chatbots are being improved to integrate with search. Microsoft is already incorporating the ChatGPT technology into Bing and is developing OpenAI features in all of its products.
How ethical is using AI to create SEO content?
Clearly, passing the output from AI tools off as human-generated for personal or financial gain is the epiphany of unethical. Companies, organizations, and institutions everywhere are taking a position against AI content.
This is especially true in the education sector, which already has plenty of challenges fighting academic dishonesty—like plagiarism and cheating. The New York City Public Schools have already banned ChatGPT. Similarly, the International Conference on Machine Learning banned any undocumented use of ChatGPT or similar applications to create any text in academic papers.
In SEO, we must examine how we’re improving the experience and delivering more value to viewers to influence results positively. If you’re automating low-quality content creation at scale for the sake of manipulating traffic, ask yourself if it’s really something you should be doing.
It’s important to do things properly, even if we think Google isn’t checking right now. There’s a long-term SEO benefit for integrity.
Is ChatGPT a threat to Google?
Google is thinking about how people could get their answers from chatbots like ChatGPT and may not need to rely on Google Search.
The ‘Code Red’ at Google prompted by ChatGPT in December 2022 confirmed that AI chatbots could be real disruptors for the search titan. It sounds like a top priority for CEO Sundar Pichai, who has reportedly “upended the work of numerous groups inside the company to respond to the threat that ChatGPT poses.” Of course, this could result in a major blow to ad revenue. So it makes sense they’re concerned about any risk to their bottom line.
It’s hard to forecast how much Bing will close the gap with the power of ChatGPT. But with Microsoft’s $10 billion investment and 49% stake in OpenAI, you better believe they’re striving to siphon searchers away from Google.
Google still enjoys 90% of the market share of search engines, and they’ve been a leader in the AI space for quite some time. Knowing how much threat the AI bots actually pose is difficult, especially when Google—the default search engine for such an overwhelming majority—has its own competing technology.
While the search giant is not too big to fail, it does have an incredible headstart.
How is Google competing with ChatGPT?
Google has already made its own impressive strides in AI.
The company’s new chatbot “Apprentice Bard” is reportedly based on LaMDA, its Language Model for Dialogue Applications originally unveiled in May 2021. It’s expected to offer more accuracy, better logic, and fresher data compared to ChatGPT. From the looks of it, this isn’t the only chatbot contender they’re testing.
Google is also planning to unveil a demo version of its search engine supported by AI.
Furthermore, it’s been nearly a decade since a human lost the board game to AlphaGo—the AI created by Google’s subsidiary DeepMind. Now the company is preparing to launch Sparrow, a chatbot focused on safety and source citation. The new bot is supposed to provide citations alongside the responses it generates, a feature ChatGPT is currently lacking.
Google’s initial reservations about implementing the technology in search have been rooted in its concern for tarnishing a glowing reputation for accuracy and trustworthiness—something the chatbot technology isn’t exactly known for at this time.
Ironically, the Bard made mistakes during its debut demonstration. Call it Murphy’s First Law in action.
One thing is for sure, ChatGPT is accelerating Google into a new era of AI, and the whole world is along for the ride.
How will chatbots shape the future of SEO?
As impressive as ChatGPT is, it’s opened a huge can of worms for SEO as we know it, and this is only the beginning. There are a few ways SEO will be impacted by chatbots moving forward.
If people start getting their answers from a robot, we can infer we’ll see fewer clicks on links. As it stands, up to 25% of searches end without a click. Viewers won’t have to click through to get pertinent information when their questions are answered in a search chat. Likewise, we can expect to capture fewer top-of-the-funnel impressions. This will undoubtedly shift the way we track and report performance.
As more web developers jump on the AI bandwagon, we can anticipate a surge in low-quality unoriginal content created specifically for volume to manipulate results. Eventually, these SEOs will find themselves behind the game and looking for help with white hat content strategy as they scramble to do things the right way.
We’ll see a higher reward for original and creative content. New thoughts and previously undocumented concepts will stand out to search engines like diamonds in the rough.
We could see a newly created demand for chatbot optimization where the goal is to rank in bot results. Since language models are trained on published text, optimizing for NLPs might include massive PR campaigns targeting the data sources used for training the bots.
Google’s overall search market share could shrink and search optimization may be less centralized around Google than it is today. As the competition adopts the technology on a broader scale, it’s feasible that smaller engines like Bing will garner more attention. I’m just not holding my breath waiting to see if Yahoo will catch up.
In the meantime, SEOs are watching closely for trends that may be explainable by GPT.
How can you safely use ChatGPT for SEO?
Playing it safe begins with focusing on creating quality content that is valuable to readers. We want to shun any strategy that resembles publishing loads of spammy content just to manipulate results.
Comprehensive human review is a critical component of using AI that can’t be neglected. AI tools are impressive, but they’re not real replacements for writers. Nonetheless, that doesn’t mean there aren’t ways we can use them to streamline our process and reclaim time lost to mundane tasks.
It’s smart to think of ChatGPT as a digital assistant that takes care of the otherwise simple stuff so you can focus your attention on things that require more brain power. Here are several suggestions for using GPT in your day-to-day that won’t result in a penalty:
- Making a list of ideas or outlines for articles
- Generating outreach or marketing email templates
- Writing hooks for social media posts
- Creating or explaining excel formulas
- Drafting scripts for commercials
- Coming up with jokes, poems, songs, and jingles
- Writing copy in the voice of famous people
- Correcting grammatical errors
- Coping with writer’s block
To ensure quality and accuracy, avoid copying and pasting the output directly into articles or websites.
Use caution incorporating ChatGPT into workflows or becoming too reliant. The software is still in the early stages of development and isn’t foolproof. Network performance is sporadic for the free version, an issue OpenAI plans to resolve with a subscription model.
For more specific examples of using ChatGPT for SEO, review the helpful resources linked below.
How much does ChatGPT cost?
In February 2023, OpenAI announced its new subscription plan, ChatGPT Plus—touting priority access during peak times, faster responses, and new features and improvements for $20/month. It stated:
“We love our free users and will continue to offer free access to ChatGPT. By offering this subscription pricing, we will be able to help support free access availability to as many people as possible.”
The pilot of the premium version will be available to customers in the United States first. The company set up a waitlist for those patiently waiting for premium access to expand to additional countries and regions.
This is the first iteration of OpenAI’s subscription model, so we may see its offers adjust as plans are created for businesses and lower-volume users.
What should you consider about using ChatGPT for SEO?
The true impact of chatbots on search marketing has yet to be felt. We know it’ll shake up search results, strategy, and how the industry looks at reporting. But technically, we’re all beta testers wondering just how deep this rabbit hole will go.
On the one hand, there is real potential for streamlining day-to-day tasks by using it as a time-saving agent. On the other, the quality of the output is low and requires the keen oversight of an educated mind. There’s no room for skipping a human editor.
While there are multiple ways to use ChatGPT to support SEO efforts, it’s not a reliable stand-alone solution for creating quality content. For the sake of quality, it’s imperative for anything automated to undergo a thorough review. Human writers are still the most intelligent way to approach content for blog articles and website content.
Content marketing will become even more reliant on editors who understand the strengths and weaknesses of automation. Success will inevitably come to those embracing the technology in a way that prioritizes quality.
At the end of the day, the risk of penalties from Google for spammy content is no laughing matter. Some SEOs will struggle to keep their hats white using AI given the opportunity for abuse. I pity the black hats who will ultimately get zapped for automating low-quality content at scale in order to game results.
Are you interested in designing a winning content strategy? Don’t rely on ChatGPT for your SEO—we’re here to help you succeed!
Helpful resources for ChatGPT and SEO
- How to NOT Automate Your SEO & PPC with AI – Perfect Search
External links open in new windows:
- 6 Ways SEO Pros Are Using ChatGPT Right Now
- ChatGPT Examples: 5 Ways SEOs and Digital Marketers Can Use ChatGPT
- How to… use ChatGPT to boost your writing
- Spam policies for Google web search – Google
- More content by people, for people in Search – Google
- Google Search’s helpful content system and your website – Google
- Our latest update to the quality rater guidelines: E-A-T gets an extra E for Experience – Google
- English Google SEO office hours from April 1, 2022 Google (YouTube)
- ChatGPT For Content and SEO?
- Microsoft Bing With ChatGPT Reportedly Launching In March
- New York City Department of Education Bans ChatGPT
- ChatGPT Caused ‘Code Red’ at Google, Report Says
- GPT-2 Output Detector Demo
- Avoiding Zero-click Searches to Get More Traffic
- Google search responds to Bankrate, more brands using AI to write content
- CNET is quietly publishing entire articles generated by AI
- CNET and Bankrate Say They’re Pausing AI-Generated Articles Until Negative Headlines Stop
- Google plans to demo a version of its search engine with AI chatbot features
- DeepMind’s Sparrow is Google’s answer to ChatGPT
- Building safer dialogue agents (Sparrow) – DeepMind
- The inside story of ChatGPT: How OpenAI founder Sam Altman built the world’s hottest technology with billions from Microsoft
- New AI classifier for indicating AI written text – OpenAI
- Introducing ChatGPT Plus – OpenAI
- Google Is Likely Testing Its Own ChatGPT-Style AI
- Google is asking employees to test potential ChatGPT competitors, including a chatbot called ‘Apprentice Bard’
- Google Search’s guidance about AI-generated content – Google
- Reinventing search with a new AI-powered Microsoft Bing and Edge, your copilot for the web – Microsoft
As Perfect Search Media’s Content Marketing Manager, Ryan brings over 10 years of marketing and management experience to the table. He’s the company’s in-house editor and the mind behind the Marketing Leader Digest bi-weekly newsletter. When he takes off his SEO cap, he enjoys art, antiques, live music, spending time in nature, and visiting family.