It’s perhaps no exaggeration to say that the biggest hype storm in months has swept the world with the launch of ChatGPT. The artificial intelligence-based language model makes English super easy to communicate with and answers almost any question.
But the question is, how can all this be used in SEO? Before we get our diving suits on and dive into the deepest depths of AI-based search engine optimization, let’s take a look at what ChatGPT is.
What is ChatGPT?
Perhaps the easiest way if it gives the answer itself.
“ChatGPT is a large language model developed by OpenAI. It is designed to answer questions and engage in conversation with humans in natural language. The model is trained on a massive dataset of text from the internet, which allows it to generate coherent and contextually appropriate responses to a wide variety of questions and prompts. ChatGPT is part of a family of models known as GPT (Generative Pre-trained Transformer), which are widely used in natural language processing applications.”
It’s true that it’s a bit sloppy in Hungarian (at least for now), but it still gets the point across. Basically, you can ask anything, whether it’s a recipe for a soup, the big questions about planting radishes or what Aristotle would think about mobile phones.
How can you use ChatGPT for search engine optimization?
Now let’s look at how you can gather the wind of ChatGPT in the sails of your SEO.
1. Create content
Content production is a problem in many places, as there is no capacity for copywriting. Now, even ChatGPT is not yet able to do this, but the signs are very encouraging. You need a person to proofread the text, as it is very often imprecise or monologues unnecessarily, but a skilled proofreader can remedy this.
Furthermore, it is not clear exactly what Google’s position is on AI-generated content. Last year, in 2022, Google spokesperson John Mueller said that the search engine considers AI-generated content to be spam and as such, a violation of the company’s policies.
But now news has emerged, also from Mueller, that Google is not penalising content created by AI. So it is better to be careful and rephrase the text generated by ChatGPT.
But the good news is that it’s great for brainstorming, whether it’s a headline idea or an entire year’s content plan.
It can also write great social media content and even write the script for you if you want to make a YouTube video.
ChatGPT is also a great tool for online shops with thousands of products, as it can help you write product descriptions as well as category descriptions.
But here again, the advice is to always check the text you receive with a professional.
2. Keyword research and analysis
In addition to developing a content strategy, another very useful feature of ChatGPT is keyword research.
All you have to do is open the chatbot, type in a few relevant, industry-specific keywords and ask it to provide you with additional, related keywords.
The results are quite astonishing.
The platform also saves us from monotonous work such as organizing keywords.
3. Developing a content strategy
It’s very interesting and useful (maybe even a bit scary) how the platform can give good advice on content strategy.
Of course, you may reasonably flash that red light in your head when you think about it: “what if my competitor is doing exactly the same thing?”.
This is why you should not rely solely and exclusively on ChatGPT. You can draw ideas from it and vary the order, but you should never implement suggestions without asking questions.
4. Generate schema code
Schema.org (or structured data) are special snippets of code that help search engines understand and display website content in search engine results pages (SERPs).
These codes can be used to structure different types of data, such as addresses, dates, products, articles, videos, recipes, people, etc. It is worth knowing that structured data is embedded in the HTML code of web pages and helps search engines to interpret the data more easily, so that better rankings in search results can be achieved, as well as various features such as rich snippets or AMP pages.
There is no lack of schema generators, but ChatGPT is one of the best because it requires few instructions and saves you a lot of time. Simply copy and paste the content and ask it to generate a schema code.
Note that it also offers help at the end. It’s amazing.
5. Create meta descriptions
Although there are industry rumours that Google ignores meta descriptions in organic rankings, it does display these descriptions in the snippets of search results.
So it’s actually in your interest that each page has a unique meta description, as this can improve click-through rates.
So you can use AI to write these short little descriptions that will save you a lot of time.
6. Generating heading ideas
ChatGPT can generate not only headlines, but also H2 H3 headlines for the content.
To do this, you should specify the main keyword, the main topic and the style and wording you want to use. It is best to write two or three examples on this.
7. Write Alt text
“alt text” or “alternative text” is a short description that you can write for images on web pages, and as such helps visually impaired people to get information from the image. It also helps search engines to better understand the content of images, which is an important SEO ranking factor.
If you have thousands of pictures, that means thousands of alt texts, which are no small amount of work and time to write each one. However, ChatGPT can help you with this too, if you enter the keywords and describe which images you want to text for.
8. Creating FAQs
For the FAQ, or frequently asked questions page, you can use the schema generator mentioned above. This is also highly recommended, as answering the most frequently asked questions by users is a really nice big red dot in Google’s eyes.
But you can take it one step further, because you can also use chatGPT to compile a list of the most common questions without having to do any research yourself.
And the answers are not just any answers, just look at them.
Beng! We’ve already got 10 great questions for you to answer (and of course, you can also use the program to help you.
9. Spell check with chatGPT
Once you’ve written the text (and you can now insert anything you like here, such as an email, blog article, product or category description, etc.), you can ask ChatGPT to check for typos.
It can’t correct semantic and stylistic errors, or clumsy sentences, but it’s perfectly capable of filtering out typos.
10. Writing emails to request backlinks
If you have already suffered staring blankly at the monitor and at the blinking cursor you should read this section in detail.
ChatGPT was put to a mini test and the results were quite interesting.
This one didn’t feel as strong, so we made another one with it, with more enthusiasm. This was a much more sophisticated template.
11. Creating product and category descriptions
Just as ChatGPT is perfect for creating meta descriptions or alt texts, it is also great for creating category and product descriptions.
This is especially useful if you have a webshop with hundreds or thousands of products, with lots of similar things. You can create different descriptions for 5-10 of the same product, although this can make even the best copywriters sweat, but for dozens it’s almost impossible.
ChatGPT is perfect for this: give him the source text and ask him to rewrite it. Then again, and again and again until you have the number you want.
Remember, it’s important for Google to have unique product descriptions, and it will reward you for the work you put into them.
It is also worth giving ChatGPT the main keywords, as well as the keyword density, how many times they should appear in the text. You can also give it different prompts for descriptions. For example: make the description emotional, or be objective and to the point, etc.
12. Creating regular expressions (i.e. regex)
Regular Expression (in short regex) is a set of characters used to search for patterns in text.
For example, the /b[aeiou]t/ pattern can be used to search for words like “bat, bet, bit, bot, but” on a given page.
Regular expressions can be useful for reporting, for example, as it allows Google Analytics or Google Search Console to look only at very specific groups of pages when collecting data.
To get a better understanding, let’s say you have an ice cream franchise network, all over the country. If you want to find out how ice cream franchises in a particular district or city are performing in search, you can use chatGPT to generate a regex code that you can use to easily query them.
This is especially useful if you would otherwise have to find dozens or even hundreds of URLs and paste them into the string. Instead, chatGTP does the lion’s share (99.9% of the work).
Let’s see how this looks in practice.
Here is my question:
“I have an ice cream franchise network and I want to know how the shops in the 9th district are performing. Can you write me a regex code for this? “
ChatGPT’s answer:
“Of course, I’m happy to help you write your regex code!
To search for shops in district 9, we suggest the following regex code:
^.*\b(9\. district|IX\. district)\b.*$
This code will find all text segments that contain the text “9th district” or “IXth district” before or after any string or space. The ^ indicates the beginning of the text, the .* indicates any string, the \b indicates the word boundary, i.e. that only the text ‘9th district’ or ‘IXth district’ will be found, and the $ indicates the end of the text.
You can use this regex code, for example, to list the names and addresses of shops, and then to analyze the data to determine how the shops are performing in the 9th district”
So you can paste this code into both Search Console and Analytics and retrieve the relevant stores. Quite simply genius.
13. Creating automations
With ChatGPT, you can also write simple code that you can use to make your automations more efficient. (This will not be easy, so it is recommended mostly for advanced users.)
We’ll show you this with an illustrative example, and let’s stick to backlink gathering.
So you have collected the URLs, domain names and first and last names of the authors you want to contact. Now you want to find their email addresses.
To do this, you’d have to browse the internet one by one, which is time-consuming. Instead, you can use ChatGPT. You will need App Scripts and Hunter.io.
To write the code, here is the command to ChatGPT:
“Using Hunter.io’s “Email Finder” API, write a function called findEmail in Google Appscripts that returns a person’s email address. If no email address exists, the response should be “no contact displayed”.
To do this, I enter my first name, last name, domain name and my API key [API key] is also needed.”
Then we take the generated code and open the App Scripts application, where we have to paste the code and return to Google Spreadsheets.
In the email column type “findEmail”, open the brackets, click on the first name, last name and domain. Then press Enter, and to continue displaying, simply drag down the cells and the formula will be inserted automatically.
Of course, this is just one idea of many, there is plenty more fun to be had with ChatGPT.
14. Making an outline/brief
This certainly gets most content producers salivating, and for good reason. In many cases, the making of the outline is the biggest problem, it’s the most time-consuming and vice versa: if you have a ready-made outline, you can work much more efficiently and quickly.
It is not worth relying entirely on ChatGPT, as it can make mistakes, so you should still treat the brief or outline with reservations. Think of it more as a helper who is there with you in brainstorming sessions, giving you inspiration and ideas for your work.
15. Creating a summary
The majority of users read by scrolling. They look at the H2 titles first, and then read the chapter summary if they find it interesting. This is why it is important to have a summary section in the post.
ChatGPT, while not being able to write long, complex and good texts, is a great tool for writing social media posts and summarising a larger text.
Thanks to NLP (neuro-linguistic programming), it can understand the text you have written and express its meaning in a simple, concise way.
16. Categorising content (semantic SEO)
Semantic SEO is a fairly recent term in search engine optimization, where the focus shifts from individual keywords to the meanings and themes behind keyword searches.
When a user enters a query, the search engine tries to understand the intent of the search to return more relevant results, rather than simply returning web pages containing the search keywords.
ChatGPT helps with semantic SEO in two ways: firstly, it categorises content so that search engines understand the main message of the website and let it appear in related searches.
Second, it helps you identify the search intent behind relevant keywords and adjust your content accordingly.
There are many words that can have different meanings in different contexts, and ChatGPT can help you identify them and use the right words. This will reduce the chances of your website appearing in searches that are absolutely not related to your website, simply because you used the keyword in a different context.
17. Synonyms
Keyword stuffing is a tactic where keywords are placed in the content with high frequency, often in their unchanged form (and without prefixes).
This, of course, results in very mechanical, often completely unreadable texts. A long time ago, in a galaxy far, far away, it was a very popular SEO technique, but nowadays it is not only extinct, but Google is actually penalising this type of content.
This is why you need to pay attention to keyword density and to create nice, varied texts with synonyms. Luckily, ChatGPT can help you with both.
Opinions differ on what the ideal key density should be. Some say that between 0.5-1.5% is ideal (for example, according to the RankMath WP SEO extension), while others say that it is above 2.5-3.5%. Somewhere in the middle is the truth: so it’s worth keeping this number below 2% in a given text.
But how can you use ChatGPT to find out what the keyword density is? Fortunately, the answer is very simple.
Simply copy the text and enter the keyword in the prompt.
Például volt egy cikkünk, amelynek fő kulcsszava stílszerűen a “kulcsszókutatás” volt. A cikk címe: Kulcsszókutatás a SEO-hoz.
A prompt:
What is the keyword density of the word “keyword research” in the following source?
Source:
“Algorithm updates come and go, SEO trends emerge and then disappear into the background… But in a sea of variables, one thing is always constant: keyword research.”
Good search engine optimisation is based on a well-developed content marketing strategy, and content is based on keyword research. (It’s not for nothing that SEO is often compared to construction; look at how the activities are built “brick by brick”.)
This is why it’s crucial (bad pun alert) that keyword research is thorough and, above all, tailored to you and your market. But let’s not go so far ahead, let’s start from the basics.
Keyword research for beginners
And what is keyword research?
Keyword research is actually a process where you look for phrases, words or phrases that your potential customers are searching for.
This sounds rather dry and uncharacteristic, so let’s look at it through an example.
If you have an online shop selling Korean cosmetics, this will obviously be your main keyword: korean cosmetics. (Although we consistently mention a keyword, the keyword will be more of a combination of words, but don’t be fooled.)
Of course, there are a lot of alternatives that fit your profile: Korean face creams, Korean makeup products, best Korean face creams, etc.”
ChatGPT’s answer:
The word “keyword research” occurs 5 times in the source, which is 294 words in total. This means that the keyword density (i.e. the proportion of occurrences of a keyword in relation to the total number of words) is about 1.7%.
If this number is too high (which is optimal), we can ask ChatGPT to write synonyms for us instead of keyword research.
Prompt:
Write some synonyms for “keyword research”
ChatGPT’s answer:
Some synonyms for “keyword research”:
- Keyword analysis
- Keyword optimization
- Search phrase research
- Search term analysis
- Keyword study
- Keyword investigation
- Search query investigation
- Search term analysis
- Keyword exploration
A teacher would give this a weak B-minus, but let us add that the main keyword itself is not easy. However, the words “keyword analysis” and “keyword optimization”, for example, can be incorporated into the text to improve keyword density.
18. Generating long tail keywords
You can also generate long tail (3 or more words) phrases from existing keywords. Let’s take a look at what our little friend has come up with based on the prompt.
Prompt:
Generate long tail keywords for “keyword research”
ChatGPT’s answer:
Of course, here are some long tail keywords that are related to the topic of “keyword research”:
- Effective keyword research strategies
- Advanced keyword research techniques
- Comprehensive keyword research tools
- Long tail keyword research tips
- Keyword research for SEO optimization
- Keyword research and analysis methods
- Targeted keyword research for content creation
- Local keyword research for location-based targeting
- Competitor analysis and keyword research
- Long-term keyword research for sustainable growth
As you can see, it’s put together quite useful phrases. You can incorporate them into your text, use them as H2, H3 titles, or give you more content ideas.
For example, you write a thousands-word Skyscraper article about keyword research, in which you mention free keyword research tools. You can then produce a smaller piece of content of 7-800 words about these tools.
You can link from the skyscraper article to the free tools content, and from there you can link back to the thousands of words of content.
19. Updating content
Google loves fresh content. However, if you have dozens or even hundreds of blog articles, updating your content can be extremely time-consuming. Luckily, we have AI to do the job for us quickly and efficiently.
If you have content where you want to update the year, for example, and rewrite the text slightly, simply copy the source and paste the following.
Prompt:
Rephrase the source given and change all the years in the text from “2015” to “2023”.
If there is a keyword that you want to include in the text afterwards, you can do so without further effort.
Prompt:
Rephrase the given source, insert the keyword “SEO marketing” logically into the text and change all the years in the text from “2015” to “2023”.
It’s okay if the prompt looks a bit like it was written by an AI, because you have to think with an AI brain: you have to create a text that ChatGPT understands as soon as possible and executes the prompt accurately.
The future is uncertain
In the first 5 days of its launch, ChatGPT has already reached 1 million users. It may not sound like much at first, but put in context, it’s a different story: it took Netflix around 3.5 years, Facebook 10 months, Spotify 5 months and Instagram 2.5 months.
Some SEO experts predict that in the next 2-10 years, many typical outsourcing or junior jobs such as legal, financial advice or accounting will be done by AI.
The real power of ChatGPT lies in its ability to provide direct answers to questions through a very simple user interface, and all in a few milliseconds.
But the future is still very uncertain. Bill Gates, for example, has stated that “ChatGPT will change the world as we know it today”. But what that means exactly, no one knows.