Tone Analyzer FAQ

Modified on Tue, 22 Nov 2022 at 03:19 PM

The Tone Analyzer is intended to help Contently customers understand their brand voice and tone so they can be consistent without stifling creativity or editorial quality. The tool brings data to an area that used to rely more on intuition and subjectivity. Ultimately, though, we aim to present it as a useful guide rather than a strict rule. The Tone Analyzer should be one part of a larger conversation about brand governance, voice, and style.


How has the Tone Analyzer evolved over the years?

The core IBM Watson technology is still similar to when we launched this feature in 2016. However, there have been a few minor tweaks to the natural language processing meant to improve accuracy. And since the Tone Analyzer uses machine learning, it's supposed to automatically improve over time as it digests more data.

On the Contently platform, we've also made a few product enhancements over the years:

  • You can see a timestamp for the most recent changes.

  • You have the ability to refresh the Tone Analyzer immediately while editing rather than needing to submit a draft.

  • There's a dropdown menu titled “Sentence Breakdown" that highlights examples for a given tone.

The tone for a given story is off. What should I do?

Try to have a broader conversation about the brand's tone and editorial direction with your managing editor and your CS lead. Examining the “Sentence Breakdown" may highlight opportunities to make a few tweaks in the text. But keep in mind that tone analysis is more of a checkpoint or a guideline, and we recommend framing it around an entire publication rather than an individual story.


The most prestigious and successful publications around the world often employ a variety of contributors who bring different skills, voices, and interests. These same publications also tackle topics from different angles and tones. The point is not to have every article sound exactly the same. It's more beneficial for the customer to focus on delivering as much value as possible to the audience.


The more important consideration is that you're satisfied with the overall editorial quality of the story, aside from any discrepancies in tone. We recommend using the “Requirements" and “Review" tabs on the story page to make your case by:

  • Ensuring the story is free of broken links, double words, passive voice, etc.

  • Optimizing SEO elements for keywords, internal links, character counts, etc.

  • Confirming the content aligns with the story brief and any agreed-upon details.

The tone analysis can come off as abstract. What actionable steps can I take from the data?

We recommend setting up the brand's target tone by analyzing a handful of pieces you've already published. This will help with expectations and scope of work.


If you're interested in exploring more options, you can reach out to your customer success lead so we can help with the conversation. We also recommend that you only focus on 1 or 2 traits that are really important to the brand (at first!) to keep things in scope. Once enough data has been collected, one good step you can take is to review the publication tone every month or quarter.


As far as actionable steps to take on the story page, you can use the “Sentence Breakdown" tool to find strong examples of a given trait. These examples should provide some guidance on changes you can make to the text.


But remember that revising a few words in a full article will only affect the tone ratings to a point. As we mentioned throughout this FAQ, we usually don't advise making drastic changes to an otherwise good piece of content just to satisfy the target tone.


Do you have any customer testimonials or success stories that help show how the Tone Analyzer made a positive impact?

Check out this case study on Braintree, one of our financial services customers. While going through a brand overhaul, Braintree used the Tone Analyzer to find expert contributors who could write for its blog. The decision led to a 4x boost in efficiency for producing content because Braintree's team saved time on revisions and approval with contributors who already matched a desired tone.


Have additional questions about the Tone Analyzer? Please reach out to your CS lead.

Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select atleast one of the reasons

Feedback sent

We appreciate your effort and will try to fix the article