Release Note: NLP Analysis with Sentiment, Entities Coverage, and Context
Since the BERT update (October 2019) and the rise of natural language processing (How NLP affects SEO in 2020), we’ve been working hard to improve our product in this very direction.
During the last weeks, a few Surfers had an opportunity to test our NLP Analysis. They saw some stellar results by optimizing their pages for NLP and so they assured us, we’re ready to push it to all our customers.
Today we released NLP Analysis in Surfer.
This major update opens up new opportunities for writing and optimization, such as:
- Sentiment Analysis—find out if top-ranking pages have a positive or negative context
- Entities Coverage—get better suggestions in True Density. Find out which entities are the most relevant for Google for your specific topic
- Usage Context—figure out how your competitors use relevant words and phrases
The NLP Analysis will be available only with Pro plans and above, but until February 29th, it will be available for Hobby, Basic, and Trial users. Therefore you’ll get a chance to test NLP and its impact on your keywords no matter which plan you’re subscribed to.
Besides the NLP, we added a URL as a new factor to SERP Analyzer and changed the look of the main input.
NLP Analysis in Surfer
We hooked Surfer with Google (for True Density entities and entities sentiment) and IBM Watson (for the general sentiment) NLP APIs. The connection is fully automated and will give your content the edge of AI and Machine Learning.
Article Sentiment with IBM Watson
IBM Watson does a great job recognizing the sentiment of page content. We tested it against Google’s and it works much better distinguishing a negative sentiment. We decided to use it for the whole article sentiment analysis.
You will find the Sentiment Analysis as a ranking factor in SERP Analyzer so you can analyze it just like any other factor on the graph view.
Surfer takes into consideration all pages from the top ten to provide you insights on what their sentiment is. Compare your website to find out if you’re not an outlier. Remember to add your page as Custom URL if you are below the top ten!
Entity sentiment with Google NLP API
Sentiment for a single entity is another important part of the analysis. With this feature, you can check if the entity is mentioned in a positive, neutral or negative context. Open an Audit and go to the True Density section to see it in action.
Surfer sends content from the top ten ranking pages to the Google NLP API. That way we obtain a set of terms that are important to rank for a given search query. The whole topical segment fetched this way is crossed with terms used on pages that you picked for the audit.
You don't have to plug in your Google NLP API account—Surfer covers it for you.
This feature affects the True Density in Audit as well as the Content Editor recommendations.
The ultimate goal for implementing the NLP is to provide you even more relevant terms to use within your content.
After analyzing many search queries, we figured out to combine prominent words and phrases (which were the base for True Density and Content Editor prior to this update) with the Google API NLP Entities.
From now on, Surfer will show you more relevant terms that are essential entities on the top-ranking pages.
Use the advantage of integrated NLP API to write and optimize your content now! Log into your account and run your first NLP Analysis.
Google NLP API provides analysis for the following languages:
- Simplified Chinese
- Traditional Chinese
Implementing NLP resulted in a pretty big bonus feature, which is the context of use for each Entity in Audit and Content Editor.
From now, you can check how your competitors used a given phrase. That way, you can figure out how to place it within your content and make a much more informed decision about your on-page optimization. The feature will be available from the PRO plan after the testing period.
We hope that adding the examples will be a massive improvement for the overall experience with Surfer. From now on, you don't have to browse your competitors' pages bur you get the data in one place.
Secondly, if you’re not familiar with some niche terms, the context will tell you a lot about it so you get a better chance to understand it correctly.
Usage limits and availability
NLP analysis will be available with Pro subscription and above. Each—sentiment and entities analysis costs one credit. You can customize your query within the new input in SERP Analyzer and Content Editor.
Until February 29th, we decided to give access to NLP Analysis to ALL our subscribers. We wholeheartedly encourage you to test it with your phrases and see yourself the impact of this new feature.
You can still use SERP Analyzer and Content Editor the same way, without NLP—no extra credit required.
New Input with Bulk Analysis
With the NLP Analysis, we had to rethink the look of the main dashboard. As the bulk analysis was one of the most frequent feature requests, we add it to this update.
You can add multiple keywords at once by copy-pasting them from your Excel file or by separating phrases with shift+enter. Keywords will be added to the queue.
Use the new input to turn on the NLP Sentiment and Entities analysis.
URL as a factor
Putting a target keyword to your URL is a common practice among SEOs. From now you can check how relevant it is to rank high in your SERP.
Find the URL as a new segment in the right menu. You have three options from here:
- whole URL analysis
- path analysis (URL without domain name)
- domain analysis
This factor is also included in the Audit so you know if there is anything else you can do with the URL to improve your rankings. Missing exact in the URL may be a hand brake for your page.
Looking forward to hearing your thoughts on the update—let's meet in the SEO Surfers group!