I must confess. I have a love/hate relationship with keyword search volume. I get why people use it. Unfortunately, keyword volume is often employed in the worse way possible. Content strategists frequently use it to slice and dice their data to build small scale, highly complex strategies. But there are problems with relying on keyword monthly search volume.
1. Traffic is an Estimate
Monthly search volume data found in keyword research tools are only estimates. Those estimates are prone to error. Historical data is gathered and used to make predictions about future search volume. Those predictions are not infallible. Having ‘X’ number of searches for a particular term last month does not guarantee there will be that same number the following month.
2. Modeling Traffic is Difficult
It’s notoriously difficult for keyword research software to model traffic for low search volume keywords. Some do this more successfully than others. Machine learning algorithms help to make these predictions more precise, but it’s still a challenge when dealing with low volume traffic.
3. Traffic Figures are Hazy
Keyword research tools may group close variants together and often bucket traffic into different ranges, further obscuring true values. Close variants can include misspellings, singular or plural forms, stemmings (for example, cat and catlike), abbreviations, and word reorder (such as stemming example and example of stemming). Often you’ll see nice round numbers like 100, 150, or 80 instead of something more accurate, like 102, 149, or 76.
4. Traffic is Based on Different Data
Different tools rely on various publicly available data sets, none of which are 100% complete. There is no single source of truth. Many keyword research tools get their data from Google. Since they don’t disclose all search volume information, some tools may additionally use third-party clickstream data. Third-parties collect this data via browser extensions, plugins, and other software installed on your computer, with your permission. Yet other tools may use a combination of data and predictive algorithms.
5. Different Tools Take Different Approaches
Depending on how the data is obtained, a keyword tool may have its own minimum volume threshold for a keyword phrase to be included in its database. Think about it from an efficiency standpoint. There are billions of possible keyword phrase combinations with little search traffic and, therefore, little commercial interest.
6. Not Every Keyword Can Be Tracked
Significant traffic to a page often is the result of untraceable keyword variants. 90% of search volume comes from phrases with fewer than ten searches per month. Here at MarketMuse, we refer to this as the term pool multiplier. The term pool multiplier is the volume potential of a piece of content beyond the trackable volume of ranking keywords. It’s the organic search traffic a page gets from untrackable variant keywords (the term pool) it also ranks for, in addition to the totality of low-volume terms that add up to significant traffic.
7. One Page Can Rank for Hundreds of Keywords
A well-written and topically-rich page can rank for hundreds of keywords, bringing in far more traffic than the volume for one keyword alone would indicate. To get a real understanding of the traffic potential of a topic, here’s what you need to do. Combine that topic’s monthly search volume along with the volume of other closely related topics for which the page could rank. Keyword research tools require a lot of manual effort on your part to accomplish this feat.
Keyword search volume is not all bad news. It can be quite useful when paired with other metrics in a more holistic approach. The problem arises when it’s used as the guiding metric for a content strategy. Since search engines no longer focus primarily on keywords, it makes sense to use keyword search volume along with other evaluation metrics.
Consider pairing it with MarketMuse’s Opportunity Score, Authority Score, Difficulty Scores, and Topical Variants (topical breadth) to further build and refine your strategy.
Written by Stephen Jeske