With AI, a content idea can be fully written before anyone decides whether it deserves to exist. Criteria for telling search intent apart from keyword variation, and a reminder that every publication requires maintenance.
Producing content has become much easier. Today I can ask for versions, titles and spin-offs of an idea before I have even decided whether it deserves to be published. This ease is useful, and I rely on it a lot. It also has a less obvious effect on editorial planning: production can arrive before the decision.
When a draft already exists, it seems simpler to use it than to throw it away. The cost of production stops working as a natural filter, and the question “is this worth publishing?” is replaced by “since it’s ready, why not?”.
Google addresses this explicitly. Its guidance on AI-generated content states that using these tools is acceptable, as long as the result is accurate, high quality and relevant, and warns that generating many pages without adding value for users may violate the policy against scaled content abuse (Google Search Central, Using generative AI content). The spam policy defines this abuse as creating many pages with the primary purpose of manipulating rankings rather than helping people (Google Search Central, Spam policies).
The criterion, then, is not the tool used. It is what the content adds.
A similar risk shows up in SEO when every keyword variation turns into an article proposal. Research tools show dozens of closely related phrasings, and AI makes it trivial to write a page for each one.
Before opening another page, I prefer to understand whether there is a different intent or just another way of asking the same thing. “How to do a content audit” and “content audit step by step” probably call for the same answer. “Content audit for e-commerce” may call for a different one, if the needs of the people searching are genuinely distinct.
A practical way to check is to look at the results that already appear for each query. If the same pages answer both searches, that is a sign the intent is close, and a single well-built page tends to serve people better than two pages competing with each other.
Beyond intent, I want to know what the publication brings of its own. Among the self-assessment questions Google suggests for helpful content are whether the text provides original information, research or analysis, and whether, when drawing on other sources, it adds substantial value instead of simply rewriting them (Google Search Central, Creating helpful, reliable, people-first content).
In practice, I like to ask whether the content brings any of these elements:
If the honest answer is just “we need to publish something”, it is worth revisiting the content idea. The same Google self-assessment list includes, as a warning sign, producing lots of content on many topics in the hope that some of it will perform well.
There is also a cost that ease of production hides. Every published page needs to be reviewed, updated and maintained. Information gets old, links break, recommendations change. A calendar that doubles in volume also doubles the maintenance work in the following months, even if the initial production took only a few minutes.
I factor this commitment in when approving a content idea. If there is no willingness to keep the content up to date, it may be better not to publish it, or to fold the subject into an existing page that already gets attention. A complementary path is to look first at what already exists: I wrote about how to find opportunities in published pages in High impressions, low clicks.
I like being able to produce more quickly. It lets me compare approaches, test titles and get to a good version faster. I just don’t want that ease to determine the volume of the calendar on its own. The decision about what to publish is still an editorial one, and it becomes more important precisely when production stops being the bottleneck.
A simple Search Console filter to find queries that already bring visibility but do not yet turn that visibility into clicks.
Asking for improvement and asking for an audit are different tasks. The workflow I use to find problems without outsourcing the editorial decision.
SEO is still the foundation, but a page can now appear as a source for a generated answer. Four questions to track this second competition without confusing a mention with a result.