This article is for content writers who have started using generative AI tools to write. With the rise of automation, it’s easy to lose sight of what we’re all trying to do: create valuable information for real people, not churn out content by machines for algorithms.
An SEO article is successful when it thoroughly covers a topic and provides insightful analysis or interesting information beyond the obvious. A good article will include a clear headline, some first-hand expertise, and clear references or evidence to support its arguments.
The most reliable way to make sure you are writing for people is to follow Google’s guidelines for creating helpful, reliable, people-first content. I’ve extracted all the questions that apply to individual articles, rather than your overall website or blog content strategy, and put them into a GPT-4 prompt anyone can use to evaluate their writing:
Please evaluate the following blog post against the following questions. Assign 0 points to a question if it’s poor, 1 point if it’s decent, and 2 points if it’s great. Please make suggestions for improvements where you can. Then tally up my score at the end and express the total as a percentage. The final output should be a table, with all of the suggestions listed out at the bottom.
QUESTIONS:
Does the content provide original and insightful analysis or interesting information beyond the obvious?
Does the content provide a substantial, complete, or comprehensive description of the topic?
Does the main heading or page title provide a descriptive, helpful summary of the content?
Does the main heading or page title avoid exaggerating or being shocking in nature?
Is this the sort of page you’d want to bookmark, share with a friend, or recommend?
Would you expect to see this content in or referenced by a printed magazine, encyclopedia, or book?
Does the content provide substantial value when compared to other pages in search results?
Does the content present information in a way that makes you want to trust it, such as clear sourcing or evidence of the expertise involved?
Does your content clearly demonstrate first-hand expertise and a depth of knowledge (for example, expertise that comes from having actually used a product or service, or visiting a place)?
After reading your content, will someone leave feeling they’ve learned enough about a topic to help achieve their goal?
BLOG POST:
[Insert your blog post]
I got a 55% on my first draft of this article (11/20 points). It’s not like I set out to write a spammy SEO post; it’s just so easy to fall into the trap. I blame the economics of search engine content production and poor industry standards.
Having a set of guiding questions (and a concrete score you can wrap your head around) makes it easier to beat back at the tide of mindless automation. Using this prompt at the end of a workflow helps you snap out of machine mode and remember you’re writing for a human being.
I have no issue with writers using AI. I encourage it on both the projects I’m working on. I also understand how easy it is to get carried away. Now I ask the writers I work with to run this prompt once they have a first draft of whatever they are working on.
Based on the suggestions from the prompt for this piece, I added some examples to my first draft , made the headline a bit more descriptive, and explained how I apply these principles in the real world…and managed to bump the score up to 75%.
A perfect 20-point score every time is unrealistic, especially if you’re creating content on the regular. But knowing where the goalposts are, and being able to track your progress toward them, makes a world of difference. This is more of a formula for not polluting the internet with mindless machine-generated junk.
Here is a rubric for each of the questions. The questions come from Google, the examples are mine. Spelling it out like this makes it easier to do a great job with each question, and also highlights what goes into a terrible example of each:
Does the content provide original, insightful analysis or interesting information that is beyond the obvious?
- Poor: The content is copied verbatim from another source with no original thought or analysis. The content merely repeats common knowledge or clichéd ideas without adding any new insights. The content mainly consists of summaries or paraphrases of other sources, with little or no added value.
- Passable: The content rephrases information from other sources and adds a few original ideas, but it still relies heavily on existing materials. The content offers some interesting information or analysis, but it doesn’t go far beyond the obvious.
- Great: The content draws on other sources for support but provides a unique perspective, new information, or in-depth analysis that adds significant value and originality beyond what is available from other sources. The content presents thought-provoking analysis or uncovers fascinating information that significantly deepens the reader’s understanding of the topic.
Does the content provide a substantial, complete, or comprehensive description of the topic?
- Poor: The content only provides a superficial or partial description of the topic, leaving out important details or context. The content is insufficient or unhelpful, leaving readers feeling the need to seek out additional information elsewhere.
- Passable: The content covers the topic reasonably well but may be missing some key details or depth. The content provides some useful information, but readers may still want to consult other sources for a more complete understanding.
- Great: The content thoroughly covers the topic, providing a comprehensive understanding of the subject matter. The content is comprehensive and informative, leaving readers feeling satisfied and well-informed without the need to search further.
Does the main heading or page title provide a descriptive, helpful summary of the content?
- Poor: The main heading or page title is vague, misleading, or unrelated to the content.
- Passable: The main heading or page title gives a general idea of the content but could be more specific or engaging.
- Great: The main heading or page title accurately and effectively summarizes the content in a way that entices the reader to continue.
Does the main heading or page title avoid exaggerating or being shocking in nature?
- Poor: The main heading or page title uses clickbait, exaggeration, or shock value to attract attention, without accurately representing the content.
- Passable: The main heading or page title is slightly sensationalized or exaggerated, but it still relates to the content.
- Great: The main heading or page title is clear, accurate, and avoids sensationalism, while still being engaging and informative.
Is this the sort of page you’d want to bookmark, share with a friend, or recommend?
- Poor: The content is of low quality or uninteresting, and you wouldn’t consider sharing or recommending it.
- Passable: The content is decent but not exceptional; you might share it with a friend who has a specific interest in the topic.
- Great: The content is high-quality, engaging, and informative, making it something you’d want to bookmark, share, or recommend to others.
Would you expect to see this content in or referenced by a printed magazine, encyclopedia, or book?
- Poor: The content is not of a high enough quality or relevance to be included or referenced in a reputable publication.
- Passable: The content might be included or referenced in a publication, but it would not stand out as exceptional.
- Great: The content is of a high enough quality, depth, and relevance that it could be included or referenced in a respected publication.
Does the content provide substantial value when compared to other pages in search results?
- Poor: The content does not stand out in comparison to other search results and offers little or no additional value.
- Passable: The content is similar to other search results but has some unique aspects or information that provide value.
- Great: The content stands out from other search results, providing significant value and a better user experience.
Does the content present information in a way that makes you want to trust it, such as clear sourcing or evidence of the expertise involved?
- Poor: The content lacks clear sourcing or evidence of expertise, making it difficult to trust.
- Passable: The content provides some indication of expertise and trustworthiness, but it could be improved with more clear sourcing.
- Great: The content presents information in a way that inspires trust, with clear sourcing and evidence of expertise.
Does your content clearly demonstrate first-hand expertise and a depth of knowledge (for example, expertise that comes from having actually used a product or service, or visiting a place)?
- Poor: The content does not demonstrate any first-hand expertise or depth of knowledge on the topic.
- Passable: The content shows some evidence of first-hand expertise or depth of knowledge, but it could be more convincing or comprehensive.
- Great: The content clearly demonstrates first-hand expertise and a deep understanding of the topic, providing insights and information that can only come from personal experience.
After reading your content, will someone leave feeling they’ve learned enough about a topic to help achieve their goal?
- Poor: The content does not provide enough information or insights to help the reader achieve their goal or learn something new, leaving the reader feeling unfulfilled or frustrated.
- Passable: The content provides some information or insights that may help the reader achieve their goal, but it could be more comprehensive or helpful. A decent reading experience, but it could be more engaging, informative, or enjoyable.
- Great: The content provides a wealth of valuable information and insights that help the reader achieve their goal or gain a deep understanding of the topic. A satisfying, engaging, and enjoyable reading experience for the reader, leaving them feeling fulfilled and informed.
I’ve linked to Google’s entire list of self-assessment questions below. This rubric merges some overlapping questions from the original guidelines. I’ve also left out the more obvious ones, such as avoiding spelling mistakes and don’t blatantly lie in your blog posts and articles. If you need to be reminded of that stuff, these questions will not help you.
Use this rubric as a hurdle to prevent yourself from over-automating the work you do. Consider it a gentle reminder to produce better, more humane content.
What to read next…
Creating helpful, reliable, people-first content
A more in-depth look at all of Google’s self-assessment questions, and if you want to go even further, check out Google’s search quality rater guidelines.
121 Examples of ridiculously good content
A fantastic collection of content to aspire to put together by Rand Fishkin. I found this in a recent post he wrote where he talks about AI-generated content as the new floor and some of the biggest advantages human creators have over AIs (for now).
Clickbait is Unreasonably Effective
Headlines are tricky. Play it too safe and no clicks on them. Push it too far, and end up with spam. This video does a great job of explaining what the boundaries are here and how to think about type-1 and type-2 clickbait.
Why every Johnny Harris video goes viral
If you don’t have original research to work with, then a lot of your points are going to come from thoughtful analysis. Johnny Harris’ framework for layering analysis into his work is brilliant. The only thing better is a short video about his framework that demonstates how it works as it’s being explained to you.
How to write lukewarm content for search engines that no one will care about
The checklist in this article is something you run after you have a first draft. Here is a content scorecard for putting a piece together in your head before you start writing it. Helpful for making sure a post actually helps someone solve the problem they’re dealing with.
What about keyword research, internal links, and your meta description?
Everything so far has been about substance, so here is a primer on technical SEO and a beginner’s guide to keyword research that covers everything else you need to think about if you’re thinking about SEO.
Recent Posts
I'm two weeks into my 30-day AI content challenge, and it's harder than I thought. Most people have dropped out at this point. My big takeaway so far is that while AI can produce content, the...
I went to Japan for the first time in my life. I put on a nice little roll of fat from the trip. The food, at every meal, was just incredible. I mean, even their 7-Elevens have great food. This is...