A mix of ChatGPT, SEOZoom and strategic expertise: this is the “magic formula” for creating a perfect editorial plan according to Laura Copelli, SEO Specialist, who during her SEOZoom webinar gave her valuable tips on how to find new ideas for the editorial plan of an already heavily populated blog. And so, room for Artificial Intelligence, flanked by targeted use of SEOZoom but always with the eyes wide open of the professional, who must guide every step of this process.
Tips for a perfect editorial plan
Every editorial plan, for any type of site, needs strategy and goal differentiation; no less important are the consistent management of content planning and scheduling, constant creativity and analytical skills to adjust the strategy according to the results obtained.
And then there is another challenge: always being able to come up with new content ideas, especially when we are busy researching topics for a niche blog.
This is where the considerations of Laura Copelli, SEO Specialist and Digital Strategist with diversified skills in the field of “creativity” – from illustrator to art director, to web designer, then creative director and finally SEO Specialist – and long experience in training, with 18 years of courses in design schools and companies, start from here. Along this path she has developed technical, strategic and content skills and today collaborates with a team of 12 professionals thanks to whom she can deal with complex strategies on multiple clients.
On the strength of this experience, even on a daily basis, Copelli suggests various ways to exploit the potential of generative AI, a powerful tool that must be used with awareness to include it in the toolbox of copywriters and content managers. In fact, if we already know how to use Artificial Intelligence to write texts, the SEO specialist offers us interesting perspectives to make the work of editorial planning leaner, faster, more agile and to focus exclusively on writing, delegating precisely to AI some “boring” and repetitive tasks that risk taking away precious time.
New ideas for the editorial plan
Laura Copelli’s webinar is specifically dedicated to professionals who “inherited” an existing, full-bodied, monothematic blog with many articles (80/100) and are therefore grappling with the most common problem in these situations: exploding the central topic into a lot of content, which can become a tiresome task, especially after the first 30/50 articles.
As content creators, we need to identify and leverage every search intent in order to become authoritative in the eyes of Google and users. The starting point might be to analyze and catalog existing content, so as to be sure not to create duplicates, and then proceed to more direct interventions to find out what is missing, finding insights from competitors, People Also Ask questions panels, and data provided by SEOZoom tools.
As such, we have six ways to create an editorial plan effectively using generative AI (and SEOZoom tools) wisely:
- Finding new ideas from existing articles.
- Finding new ideas with SEOZoom and frequently asked questions.
- Finding new ideas from users’ needs.
- Finding new ideas with SEOZoom.
- Finding new ideas by analyzing online competitors.
- Finding new ideas by uploading a pdf to ChatGPT.
The concrete example: practical uses of ChatGPT to categorize blog articles
The expert then proposed a practical case to show the use (and usefulness) of ChatGPT to automatically generate and assist article titles and descriptions, facilitating the process of organizing and structuring content.
In this way, AI helps create a table containing article titles and descriptions in an orderly and consistent manner, streamlining editorial work and improving efficiency in managing the blog’s editorial plan.
Here are some steps and tips on how to use ChatGPT for this task:
- Collecting article titles and descriptions. Before using ChatGPT, collect all the titles and short descriptions of the blog articles we want to organize in a table, making sure they are clear, concise, and relevant.
- Defining the table structure. Decide which columns to include in the table to organize article titles and descriptions-the basics being the URL, article title, and short description.
- Interaction with ChatGPT. Use ChatGPT to generate table text based on the provided article titles and descriptions-we can enter a clear and detailed prompt to guide ChatGPT in creating the desired table.
- Review and optimization. After obtaining the text generated by ChatGPT, carefully review the table for correctness and consistency of information; if necessary, make changes or additions to improve the quality of the result.
- Tips for better results:
• Provide a detailed and specific prompt to guide ChatGPT in generating the table.
• Use clear and concise phrases in titles and item descriptions to facilitate ChatGPT’s understanding.
• If necessary, break the task into several smaller prompts to better handle the complexity of the task.
Writing with AI: ChatGPT vs. SEOZoom
The second part of the webinar focuses more analytically to the practical phase of creating content for the blog, entrusting the generation of the text to Artificial Intelligence and then intervening with the necessary revisions and changes by the professional.
According to Copelli, in fact, no one better than the copywriter “can transform a text written by artificial intelligence into an engaging, empathetic, emotional article, in which there is real value, in which the user really finds the solution to his or her problems.” Also because the flip side of the coin, and the risk to avoid, is producing and publishing “an article that is not read to the end and that is not liked, prompting the user to abandon it and causing Google to consider it of low quality.”
Instead, all it takes is a little practice to be able to create comprehensive and informative content, covering specific keywords and related topics related to the main topic, satisfying user needs and improving search engine rankings.
All this while always trying to make the most of the available tools, aware also of the limitations of AI and the potential SEO benefits of using SEOZoom’s AI tools. Indeed, Laura Copelli showed the differences between an article generated with SEOZoom’s AI Writer and one generated directly with ChatGPT, especially in terms of search intent satisfaction and topic “understanding,” and these are the central points of her experience:
- Consideration of the search intent. SEOZoom, through its Editorial Assistant, writes considering the user’s search intent: this means that the generated content takes into account the specific needs of users and aims to provide relevant and useful information.
- Less Redundancy. SEOZoom’s Editorial Assistant tends to write with less redundancy than ChatGPT; this can contribute to greater clarity and consistency in the content produced.
- Suggesting consistent concepts. SEOZoom suggests consistent concepts even if they are not requested; this means that the Editorial Assistant is able to generate content that maintains thematic consistency even beyond specific user requests.
- Keyword analysis and categorization. SEOZoom is able to analyze and categorize the keywords provided, suggesting others on the same topic. This feature can be extremely useful for optimizing content based on keywords relevant to search engine rankings.
In summary, SEOZoom’s Editorial Assistant offers advantages such as consideration of search intent, less redundancy in the text, suggestion of consistent concepts, and keyword analysis compared to a “simple” ChatGPT, which instead writes faithfully following the prompt, produces various redundancies, and does not seem to offer real user-side value.
The tricks for a perfect editorial plan: an interview with Laura Copelli
Laura Copelli also provided us with other “extra” considerations and tips for making an effective editorial plan (and for consciously and strategically using AI).
Here are her answers to our interview with her.
- How can the process of categorizing existing content on a blog be optimized to improve SEO strategy, and how can ChatGPT assist in this process?
In order to carry out a content strategy, you need to keep track of all the articles you publish. This allows you to identify all the possible search intent revolving around the brand you are writing for, without forgetting any of them.
If the blog has yet to be born you can already think about categorization, however I prefer to initially publish without organizing the articles into categories, which I prefer to go and create after 3/6 months, so after 12/24 articles have been written and published. At that point it makes sense to organize and categorize them.
This is because I find it uncool that in a category there is for example only 1 article published. In addition, categories can change as you write, they can transform, even depending on what content ranks best. In fact, I find that a content strategy adapts depending on the results it achieves. That’s why I prefer to create categories only after I understand how the blog is performing.
Clustering the articles means distributing the content, so having a balanced blog among all the different macro topics. This makes it easier to organize even the keywords we want to rank for.
If, on the other hand, we inherit the blog, it will be essential to organize the existing articles by clusters, to better understand the exploded macro topics and those that need new content.
ChatGPT is a very convenient tool for clustering.
But to be able to have all the titles and descriptions of each article ChatGPT is not suitable, because it does not follow blog pageviews and does not report all articles. It also struggles to report all articles; it always misses some.
For this initial analysis ScreamingFrog is ideal. Once you have opened the excel returned by ScreamingFrog with title and description of each article you can proceed with ChatGPT.
I do not recommend pasting more than 10 titles and descriptions per prompt: ChatGPT does not create tables longer than 10 rows. Or rather, it asks each time if it should continue making more, but if you convert the excel to pdf here is where ChatGPT works best.
You still need human control anyway.
The tables ChatGPT returns are convenient because they can be selected and pasted into excel. You can ask it to organize the various articles into clusters, and it will be interesting to see how many it creates and how it calls them.
- What is the most effective way to generate new content ideas with ChatGPT? Could you provide an example of a prompt that you found particularly effective?
ChatGPT is convenient for brainstorming: it’s neutral and not as focused as we are when we’re looking for inspiration. We are always too focused on the brand, or the keywords we want to reach, or the services and products. It’s as if we have blinders on, our heads are immobilized, and we only look at the dot in front of us because we want to hit the mark.
ChatGPT does not look at a single dot and returns all the topics that revolve around the prompt we have given it. Therefore, the topics it gives are relevant but much broader, and this “destroys” our blinders and “forces” us to turn our heads.
To provide new ideas we can ask it to:
- Start with existing articles, cluster them and tell us what is missing
- Starting with competitors, provide them with URLs and ask them for topics not yet covered
- Based on the topic I describe to him, I can ask him to find user questions
Remember that a prompt cannot just have the query, but needs context, any examples, and a prompt for completion. A prompt can be very long, in fact the better you explain the context and the better you describe the need, the more accurate the output ChatGPT will return.
I recommend creating ad hoc GPTs in ChatGPT 4 – I, for example, created one GPT from SEO Specialist and one from SEO copywriter. This allows you to start from a very specific and precise context without having to report it each time.
In the following I will only include a few prompts.
- Starting with an excel with all the blog articles already published I ask ChatGPT:
“With respect to all the articles you find in this pdf and taking into account the clusters, suggest me new titles for new articles. Organize them in a table.”
Or I can ask it to identify articles that are too similar:
“From this excel I upload tell me if there are articles that are too similar to each other and why.”
- Prompt to analyze competitors
I search in SERP for the keyword I am interested in. From SERP I open the first 10 pages that rank. Then at ChatGPT I ask, “Read the content of the following page:
https://www.ingigni.com/blog/differenza-marchio-marca/
Then create a table with the following columns:
- title H1
- page link
- short description
- most important keywords (5)”
ChatGPT prepares the table. I do the same thing with the other URLs. In the same chat, after the 10 tables, I ask:
“Now, taking into account the URL analysis you did and that I need to write an article on the topic [foo] and place the keyword [pluto], tell me what concepts and in formations are missing and that I can add in my article.
Prepare a table with the following columns:
- proposed title
- concepts that are missing
- keywords on which to base the content.”
- Prompt to find user questions
“On the topic [foo] locate user questions.”
Or even better:
“On the topic [foo] what are the needs that users might have? Set up a table with:
- need column
- brand solution column.”
- How does using SEOZoom integrate with ChatGPT to identify content gaps and SEO opportunities based on frequently asked questions?
ChatGPT is not an SEO tool, so it cannot provide important data for an SEO Specialist, such as traffic volumes, industry competition, related, seasonality, and more.
So I prefer to use SEOZoom to find user queries, which I find much more accurate, numerous and vertical on the keyword I am interested in, and only then use ChatGPT to cluster or organize them.
- How can competitor analysis influence the generation of content ideas and how can ChatGPT and SEOZoom facilitate this process?
Competitor analysis allows us to have a benchmark to surpass, or imitate, or exclude.
Knowing that it competes with us is critical to dosing effort and calculating time and resources.
But ChatGPT struggles to analyze a SERP: it returns few URLs and not always relevant.
SEOZoom, on the other hand, has a vertical tool on this, is precise, analyzes the SERP entirely, returns data such as position in the various features where it finds the domain, identifies search intent and other interesting data such as estimated traffic and seasonality.
All information that ChatGPT does not have.
Vertical analysis of a page is also too general with ChatGPT, whereas SEOZoom returns a lot of key data.
So on competitor analysis ChatGPT is not a tool I use.
- What is your approach in finding a balance between human creativity and artificial intelligence assistance in defining an editorial strategy? How do you ensure that the content maintains an authentic and engaging voice?
In my team of 12 people who collaborate with me almost daily, I have 6 copywriters whom I respect, train and pamper.
I have almost forced them to use ChatGPT, taken them to trainings, pushed them to experiment and share with others the discoveries they make.
But from the beginning I assured them that my fees would not decrease, and that for me their support would remain crucial.
The texts ChatGPT returns are as flat as a billiard ball.
Only a copywriter can transform them and make them engaging, empathetic, interesting, and enjoyable to read all the way through.
When ChatGPT is asked to create captivating text, the result is too fancy, exaggerated, marked up. There is no measure, no nuance.
He can’t even get the tone of voice right. It is easier to give him texts and have him adopt the tone of voice he evinces, but even then he will never be able to imitate the writing of a good copywriter.
I myself could never delegate the writing of my Linkedin posts: my followers follow me, my thoughts, my method, my tone of voice.
ChatGPT can help me find new ideas, new points of view, but then the writing of the content is up to me alone.