Platform Posts

AI-Powered Tag Generation

illustration of AI neural network

At RebelMouse, we’re always striving to have the latest and greatest technologies implemented into our CMS platform, and the current boom of artificial intelligence functionality has not passed us unnoticed. We’ve recently integrated with OpenAI to analyze content to detect potential content tags for your articles.

This integration is actually quite simple to set up. All you need is an account with OpenAI and to send your API key to your account manager or the RebelCare team. You can create an account on the OpenAI site and access your API key very easily. Once you have that, just send an email to your account manager or RebelCare letting them know that you’d like to have this feature enabled.

Once enabled, you’ll be able to very quickly have the AI analyze your article and provide suggested tags. As an example, let’s look at an article on my site that is about the history of Mt. Everest. In the Channels tab in the right-hand side bar, you can see that I’ve manually added the tags “climbing history” and “mt everest.”

At this point, I have not yet analyzed my content for additional suggested tags. To do that, I simply click the ANALYZE CONTENT button and OpenAI will return my suggested tags:

We can see that it suggests “Mt Everest” and “History” (though not “Climbing History”) as tags, which matches my editorially selected tag, and it also offers some very interesting suggestions. For instance, the article I wrote does indeed touch on tragedies and triumphs, which could be very interesting tags for my site. I’ve written many articles on my site that discuss various tragedies and triumphs, but I’ve never thought to have those as tags. In my case, I decided that the new tags would indeed be important content areas for my site, so I began to use those tags. Beginning with this article, I simply need to click the plus button (+) next to “Tragedies” or “Triumphs” to add the tag.

This is of course just the beginning of our expected plans for AI and taxonomy. We’re also working on filtering AI-suggested tags to tags that are already used on your website. Additionally, we’re looking at approaches to help find tags that are very similar to each other and cluster them into one winning tag to keep taxonomy clean at a massive scale through the power of AI.

Not on RebelMouse and interested in combining AI with your publishing workflows? Reach out today and let’s start working together.

What Is RebelMouse?
Request a Proposal
Meet the RebelMouse Platform: The Highest Performing CMS on the Web
Rebel Insights

Meet the RebelMouse Platform: The Highest Performing CMS on the Web

Make sure your site is set up for success in 2022.

In the spring of 2020, Google let the world know that its Core Web Vitals would become the new benchmark for measuring a site's performance in its search results, known as the page experience update. Fast forward to more than a year later in August 2021 when, after much anticipation, Google's page experience update became official.

Since its rollout, developers have felt the impact of how their publishing platforms stack up against the new standard. Important decisions around the architecture of your site can now make or break your site's performance in the eyes of Google.

HTTP Archive, a tracking platform that crawls the web to identify trends and record historical patterns, frequently reports on how top content management systems (CMS) have weathered the page experience update through the creation of its Core Web Vitals Technology Report. RebelMouse has consistently outperformed major CMS platforms on Google's most critical metrics throughout the year:

Getting superior scores on Google's performance benchmarks isn't easy, either. The Ahrefs blog analyzed Core Web Vitals data from the Chrome User Experience Report (CrUX), which is data from actual Chrome users, to see how the web stacks up against Core Web Vitals. Their study found that only 33% of sites on the web are passing Core Web Vitals.

data from Ahrefs tracked on a line chart finds that shows only 33% of sites on the web pass Google's Core Web VitalsFrom Ahrefs.

Luckily, performing well on Core Web Vitals is possible with thoughtful, strategic changes to your site’s codebase. Here's what you need to know and how we can help.

Keep reading...Show less
Multivariate Testing: An Introduction to Data-Driven Site Design
Rebel Insights

Multivariate Testing: An Introduction to Data-Driven Site Design

Understand the differences between multivariate testing and A/B tests

The modern digital landscape is founded on one critical element — data. From content creation to site design, there’s no reason to take chances on what will resonate with your audiences. Adopting a data-driven mindset means you can take the guesswork out of your business strategy and focus on the methods that are actually moving the needle.

And one of the best ways to figure out what strategies are moving the needle for your website is through multivariate testing.

What Is Multivariate Testing?

Multivariate testing is the process of testing one or more components on a website in a live environment. These components can be anything from a CTA button, headline formatting, or even an entire page design. The beauty of multivariate testing is that you can test each one of these individual features on a page to see what performs well among your users.

Think about it for a moment. Creative teams with great ideas are most successful when they have an environment where ideas can easily be tested against each other instead of trying to find total agreement on one idea. Multivariate testing allows teams to cherry-pick each idea to create an end result that works best, backed by the data to prove it.

multivariate testing allows for various layout designs and element placements to be tested live to see what attracts the most readershipSee which elements and layout designs attract the most readers with multivariate testing. Graphic from Invesp.

Multivariate Testing vs. A/B Tests

Traditional A/B testing is the process of creating two different layouts and splitting the traffic between the two to see which one performs better. It’s possible to test more than just two layouts, of course, and there’s no issue with creating A/B/C/D/etc. tests depending on how many layouts you have to try.

A/B tests can produce great results, but they are limited since they test an entire layout at once. Remember, multivariate testing allows you to test the different components of a layout individually. Think of multivariate testing as running multiple A/B tests at one time. Here’s a good illustration of the differences between A/B testing and multivariate testing from HubSpot:

A/B testing compares two layouts as a single page, while multivariate testing allows for multiple elements to be tested simultaneouslyAn illustration of the more complex testing available through multivariate testing. From HubSpot.

Multivariate testing is a great way to help creative environments stay focused. However, it’s vital that all ideas get measured, because one idea might sound awesome to the group or a team member, but it may not always perform.

How Do I Know When to Use Multivariate Testing?

If you are looking for fast results, it’s best to use A/B testing. However, multivariate testing is the preferred choice if you have the time to analyze and review multiple data points. Combined, the testing on each one of your site elements will help you curate the highest-performing page possible. It’s also recommended that you use multivariate testing on your pages with the highest traffic because there will be more data to analyze to determine which site elements are garnering the most engagement.

Keep reading...Show less
Interested in a Free Website Health Check?Check Your
Website's Health
Get Your Free Analysis Now