I Build a Content Comparison Tool Powered by Google Cloud Natural Language API

I built a Google Colab tool that uses the Google Cloud Natural Language API to compare your content against top-ranking competitors.

Integrating keywords is still important, but it’s only part of how search engines evaluate content. Instead of just looking at keywords, this tool analyzes which concepts and topics carry the most semantic weight.

This is similar to how search engines and large language models interpret what a page is really about.

Here’s the link to the notebook so you can make a copy and try it yourself.

A quick note

This tool is fully functional, but it’s not as clean or polished as it could be. Feel free to make edits to your copy to improve it! I’d also love any feedback on how to improve it or ideas on where to take it next.

Getting started

First, make a copy of my notebook.

To use the tool, you’ll need to create a JSON credentials file for the Google Cloud Natural Language API. Instructions for enabling the API and generating your key are available in Google's setup documentation. This API costs a little bit of money but you can get a $300 credit for free when you first start out. The API also allows free usage up to a courtesy usage limit.

Once that’s ready, open the notebook, upload your credentials when prompted.

Then, paste in your content along with up to three competitor pages. The tool will analyze each one and return a comparison of the top entities, sorted by importance.

What you get from this tool

The output includes two main views. The first is a high-level summary that shows how your content compares to each competitor in terms of shared and unique high-salience entities. The second is a detailed table listing every entity found, whether it appears in your content or theirs, and how prominent it is.

Here's what the summary table looks like:

And then heres a preview table of the full data:

You’ll see which concepts your content is missing, which ones you cover that others don’t, and which are shared, along with how prominent each entity is. You can export either of these tables as a CSV:

Why does this matter?

This approach goes beyond traditional keyword research by helping you understand how your content is interpreted by search engines and large language models. It’s especially valuable for spotting gaps in topic coverage, improving semantic relevance, and ensuring your content aligns with how modern algorithms evaluate information.

Try it out and let me know what you think!