The Complete ChatGPT Content Writing Workflow
- Tarasekhar Padhy
- Sep 24, 2024
- 4 min read
Updated: Sep 28, 2024
Below you will find how to complete each task in a content writing workflow with the help of ChatGPT. Tasks include generating content ideas, creating outlines, writing the draft, and obtaining the meta details for publications.
Adopting this workflow will help you save time and improve the quality of your drafts.
1. Generate content ideas
Consider you are the owner of a plant-based protein powder company aiming to educate your audience about the benefits of going vegan, and ultimately plugging in your product as a healthy protein supplement.
You can use the Blog Title Suggestions GPT for this:
There are different ways to prompt here. You can simply drop the link to your website homepage and ask for title suggestions for a particular use case. Additionally, you can ask it to search the internet for the latest vegan trends and suggest topics based on that.
It is possible that GPT’s suggestions are total trash. You can continue the conversation by giving it feedback to get better options.
For the sake of this workflow example, let’s consider the second title "How a Vegan Diet Can Transform Your Health: Essential Tips for Beginners." This title will make it easier to explain the advantages of going vegan and promote the plant-based protein powder organically.
2. Create an outline
There are multiple ways of creating an outline. The simplest one is to simply mention the word count, a bit about the target audience, and the objective of the article. For this example, our target audience can be health-conscious individuals and the objective is to promote a product.
Let’s call the plant-based protein powder V-Protein (creative, I know) so the Article Outline generator includes a strong mention.
Note that there are many ways to prompt here. This was the most minimalistic method which can be leveraged for most general-purpose use cases. You can also prompt it in two other ways:
Ask GPT to analyze the top articles with a similar title and create one that is more thorough and detailed than them all.
Mention the topics you wish to cover personally to get an outline that is more personalized for your needs.
Anyway, the GPT did include an organic mention for V-Protein:
It is a good idea to use the GPT-generated outline as reference material while creating the outline yourself. This makes the process faster and less cognitively demanding while protecting the quality and integrity of your article.
For the sake of this example, I kept it as is.
3. Write the draft
So far, you have a detailed outline with all the talking points sequentially mentioned. Rather than converting the entire outline into a draft at once, do it one section at a time.
First, tell the Content Writer GPT to generate the first section. And then the second, and so on.
This retains the quality of each individual section. There are more details and things are tied up nicely. When you generate the entire article at once, GPT-4o produces a blurb of text that is destitute of details.
As you can see, there is an “additional instruction” section in the above prompt. This is an easy and effective way to customize your request. You can add keywords, set word count, include a bit about perspective or style, and more.
Earlier, I created that GPT to generate drafts for articles in the SaaS, technology, and IT realm (as you can guess from its name) but this technique makes it easy to write for any niche and audience. It is programmed to prioritize the immediate instructions, if given, over the default setting.
I highly recommend using the machine-written draft as research material and writing the draft yourself. The biggest decelerator in the content writing process is consolidating research and converting talking points to sentences that make sense. More on this in the subsequent chapters of this book.
4. Obtain the meta details
After you are done writing your draft, inserting internal and external links, and adding necessary visuals, it’s time to put it out. Here, there are many tiny, administrative things that take too much of your time.
I am talking about the meta description, H1 tag, and URL slug. Sometimes you may even need a short social media post content.
Well, the Meta Details GPT does just that. All you have to do is paste your entire draft in its entirety. You will get ten meta descriptions, ten alternate title suggestions, five URL slug suggestions, and three distinct social media posts.
This gives you enough options to pick the right one and save a bunch of time. Most importantly, this reduces cognitive fatigue, considering these are routine things that are simply tedious.
Of all the GPTs I’ve created, the Meta details GPT has saved the most of my time. Even if you are just updating an article’s metadata, this one is your savior. If the article is already published on the internet, just paste the URL and you’re done.
Conclusion: It’s all about the nuances
There will be writers who will double their output with the above workflow while maintaining quality while others will churn out shit that nobody reads. The secret lies in how well you understand how things occur under the hood.
You need to break the content production process down to individual action items and identify the ones that can be automated. And then automate those action items with an LLM of your choice.
But how would you know which action items can be and should be automated?
Well, for that you need to have a firm grip on the ins and outs of the art of writing (outside the scope of this book) and how AI works. The rest of the chapters of the book are about the latter.
Furthermore, I have also shared certain techniques and principles that will help you integrate various AI-powered tools like Perplexity to elevate your content writing game.
---
Next Chapter: Best Practices While Using ChatGPT to Write Content
Previous Chapter: Content Writing With AI: Principles, Strategies, and Prompts
Index (with Prologue): Content Writing With AI: Principles, Strategies, and Prompts