I thought I knew exactly what one of my articles was about.
Then I ran it through Google's Natural Language API.
Let's just say the result gave me something to think about.
The tool picked up several entities and concepts that I didn't consider important. When I looked back at the article, I could see why: I'd spent more time on a few side points than I realized.
That's when the experiment became useful.
What Did the Tool Actually Show?
Google's Natural Language API can analyze text in different ways, including entities, sentiment, syntax and content categories.
I wasn't interested in getting a magic “SEO score.”
I wanted to see whether the structure of my writing made sense from another perspective.
The Entity Surprise
Entities are basically the identifiable things or concepts mentioned in your content.
Seeing them listed made me notice something I normally wouldn't.
A supporting example can slowly become a major topic if you talk about it too much.
That's a useful editing lesson.
Before publishing now, I sometimes ask:
“Am I still answering the original question, or have I started writing about something else?”
The Sentiment Surprise
One of my casual, slightly sarcastic comments also came across as more negative than I intended.
That was a good reminder that tone isn't always obvious to the writer.
A phrase can feel harmless in your head and sound completely different on the page.
The Biggest Lesson
The biggest thing I learned wasn't about Google's tool.
It was about editing.
Clear writing usually starts with clear thinking.
If I know exactly what I want the reader to understand, the writing becomes simpler. If I'm trying to cover five ideas at once, the article usually feels messy.
I wrote about the broader lesson in NLP and Semantic Writing — What I Learned From Testing My Own Content.
And the connection with search language systems is interesting too, which I cover in Writing for BERT and MUM.
Need a hand with this? Our SEO content writing team can help you put it into practice. Talk to ClicZeo.
