How should you validate keyword suggestions and decide whether to use them?

Treat keyword suggestions as candidates, rather than adopting them outright. First, confirm that they address your audience’s questions and fit your website’s content. Then check search volume, ranking difficulty, and search intent. Before setting priorities, review the actual search results and which pages on your website can meet those needs. AI-generated keywords do not necessarily reflect demand; a tool reporting zero search volume does not necessarily mean there is no demand, either. Whether to use a keyword depends on relevance, evidence of demand, and your website’s circumstances—not a single score.

Check where suggestions came from and whether they address your audience’s questions

When gathering keywords from reviews, Reddit discussions, or sales calls, retain the original excerpts or summaries. This lets you check whether a suggestion reflects a problem customers actually raised or is a term that AI came up with that sounds plausible but is unrelated to your products or services.

• Check what problem the original wording describes and whether your website has relevant content.
• Confirm that the suggestion preserves the customer’s meaning rather than merely picking up similar words.
• Do not look only at keywords for which a Reddit thread already ranks. This approach depends on the thread already ranking well and may overlook needs expressed in the comments.

As part of our content services at Procrustes, we turn the questions customers genuinely care about into articles, with ongoing planning, writing, and updates.

Check candidate keywords against data; generation is not validation

Keyword tools such as Semrush can help you check search volume, ranking difficulty, and search intent to prioritize candidates. In a test described in a 2024 article, Semrush checked eight primary keywords suggested by ChatGPT and found that one had no recorded search volume, while several others were classified as difficult to rank for. These findings apply only to that test and do not establish that all AI suggestions are unreliable, but they illustrate why each keyword still needs checking.

Alongside tool data, review the search results and the actual state of your website’s pages. A 2023 article from awoo recommends considering Search Console data, the types of content favored in search results, the state of the client’s webpages, and keyword types when deciding which keywords each landing page should target. The decision should not be left to a generative tool alone.

Double-check zero search volume before discarding a keyword or creating separate content

A 2026 Semrush article notes that zero search volume or missing data may simply mean that the tool currently shows no search volume for that exact wording. If the phrase comes from prospective or existing customers, it may still be worth including in your content, but that does not justify claiming it will definitely generate traffic.

For example, the article starts with “best communities for beginner freelancers” and finds search volume data after checking “best communities for freelancers” instead. Checking variants like this can help determine whether separate content is needed or an existing article can answer the question. The goal is not to change the original need just to obtain a number.

Decide keyword by keyword, even when search results include an AI Overview

A 2026 Search Engine Journal article states that the presence of an AI Overview does not mean a keyword should automatically be abandoned. For informational queries at the top of the marketing funnel, results dominated by AI Overviews may generate fewer clicks. Your content goals should still guide whether you target the keyword or adjust the writing and how the content is presented in search results. An AI Overview is one consideration when deciding whether to use a keyword, not a reason to exclude all informational keywords.