Skip to main content
What is the Search Ring?

The Reword search ring is a LLM (large language model) powered scoring system for articles, based on target search intents.

Charles Kelly avatar
Written by Charles Kelly
Updated over a week ago

What is the Search Ring?

SEO scores are a feature as old as search itself, often calculated by measuring a contents depth (word count) and combining it with a recognition of target keyword presence.

The problem is, that doesn’t resemble good content.

Instead, it incentivizes keyword stuffing and general bad practices. It’s one of Reword’s promises that we’ll always align with the readers over the algorithms, so that wasn’t an option.

That’s why we haven pioneered a proprietary search scoring system, which we are coining the search ring.

It’s designed to be qualitative (based on the strength of the content) instead of quantitative (based on the number of keywords) and correlates with a well-written, reader-centric article.

If your search ring is lighting up green, with a score of 80+, you know you have a good article.

How it works (in simple terms)

When you set out to create a new draft with Reword, you specify search intents that you are trying to satisfy for your readers.

For example, the article “How to make sushi in 5 easy steps” might have the following intents:

  • How to make sushi

  • Beginner sushi techniques

  • Easy ways to make sushi

  • Steps to make sushi

  • Tips for sushi beginners

Every time you make a change to your draft, our proprietary language model will review the content for each intent and determine if it is satisfying the given intent.

Each corner of your search ring will represent a single intent, with red equaling bad alignment and green representing complete alignment.

image

This makes it easy to quickly glance at a search ring and get a reputable impression of its performance.

While this is an explanation of the algorithm in its most simple form, it is representative of our fundamental approach to the search ring.

A mixture of quantitative and qualitative signals are then combined with this to determine the final search score.

Examples of different search rings

The best way to grasp the search ring quicker is by looking at examples of different scenarios.

Let’s walk through 3 of them:

Performing

Let’s start with an article that’s performing. In this example, each search intent is satisfied to a high degree. At a glance, we can tell this because the search ring is green.

image

There isn’t much scope for improvement here, unless we were to modify our search intent targets.

Opportunity

Now let’s look an example where there is room for improvement. If our search ring has patches of yellows (or reds), we know that there is more to be done to satisfy all of our search intents.

image

The search ring makes it effortless to scan through your articles and immediately see which one’s have performance to be unlocked.

Stale

Here’s an example of an article that really isn’t doing well. We’re not aligned with any of our intents, which means this article is not going to be well received by our readers or search engines.

image

As the issue is widespread, it likely means the content is thin and we need to write more content for each given intent.

Why should you trust it?

To an outsider looking in, the Reword search ring and generic SEO tool scores may seem familiar on first look.

It’s behind the scenes that the intelligence of the search ring (and score) are realized.

We believe we are the only provider to have a proprietary language model powering their search score algorithm.

More importantly, the qualitative nature of the search score is designed to replicate Google’s people-first approach to ranking articles.

The way we see it, as long as our search score always favours the reader and correlates with high-quality articles, we can’t go wrong.

Our search ring follows these core principles:

  • Led by qualitative analysis.

  • Cannot be tricked by keyword stuffing.

  • Prioritises people over algorithms.

  • Focuses on wider search intents, not individual keywords.

Did this answer your question?