Bring your own Keyword difficulty score, No_cluster, Improved clustering algorithms +++

Bring your own Keyword difficulty score.

At Keyword Insights, we offer flexible options for clustering. A few months ago, we introduced a Keyword Difficulty score to help users find clusters that are easier to rank. Now, we’re taking it further by letting you import your own Keyword Difficulty scores from tools like Semrush, Ahrefs, or Moz.

Here’s how it works: Download your keyword data from your preferred provider, upload it to our clustering tool, and toggle “I want to use my own KD.” Then, map your columns and select Keyword Difficulty alongside keywords and search volume.

Once processed, we’ll apply your KD score to each keyword and average it for the entire cluster.

Cluster ranking at different domain levels

Introducing another flexible option for clustering: you can now track rankings for your cluster across different domain levels.

Subdomains - Default setting. Tracks everything.

Exact URL - Only tracks the exact URL.

Path - Tracks anything in the provided path.

Domain - Tracks only the specified domain. (Sub domains are not tracked)

No more dreaded “no_cluster

When we cluster keywords and find no SERP similarity between them, we used to place these keywords into a single cluster labeled “no_cluster” which wasn’t very helpful.

From now on, we’ll assign each of these keywords to individual clusters. We’ve also added a keyword filter that allows you to quickly identify clusters containing a single keyword.

Here’s how you might approach clusters with a single keyword:

1. Research Opportunity: If a keyword piques your interest, set it aside for further research. This could indicate that your current keyword list lacks depth.

2. Focused Content: If a keyword is highly specific, consider creating in-depth content focused on that single topic.

Significant improvements to the clustering algorithms.

After dedicating substantial time to research and development, we’ve made significant improvements to both of our clustering algorithms.

These enhancements deliver more accurate cluster outputs. Our analysis of existing orders and the improved algorithm revealed that the agglomerative algorithm offers better results for our customers, so we’ve made it the default setting.

Additionally, we’ve increased the SERP overlap from 3 to 4, which has greatly improved output quality.

You can still customize algorithms and settings under Advanced Settings.

For guidance on choosing the best algorithm, please read our documentation here.

• Fixed a bug with cluster relevancy filter.

• Fixed a bug with deleting all orders.