Some search queries may return a "Low quality corpus" warning.
This warning typically occurs when search-results consist of webpages with thin-content, or very little article text.
This is a common issue with some types of webpages including storefronts, videos, download pages, and image-oriented websites, which have very little "article" content.
For example: If your search query is for "Stainless frying pan", it is unlikely that your search results will consist of long-form articles. Instead, the search results are more likely to include online storefront pages with large pictures and minimal text.
Without sufficient site text to analyze, Wordmetrics has very little content to work with.
Even with these thin content pages, Wordmetrics' AI will do its best to create a meaningful semantic index, but the quality of the generated index will only ever be as good as the quality (and length) of the content on the pages used to create it.
When a low-quality corpus warning is issued, there's an increased likelihood of less-relevant terms sneaking in to your index of semantic terms.
Does this mean you can't use this simple search term? No. But take it as a gentle hint to modify it to return a more knowledge-rich set of results.
In the case of a search query like "Stainless frying pan", try using a search query like, "Stainless vs. cast iron frying pans" which is more likely to return meaningful article content.
The more text Wordmetrics has to work with, the higher the quality of the semantic index.