Google’s problem with AI-generated content is less the process than the lack of effort. The search giant’s spam policies don’t penalize AI content per se, but its guidelines for human quality raters (PDF) emphasize the importance of “effort” 120 times.
A lack of effort could explain studies that consistently show penalized AI content, presumably from the search algorithm’s spam component.
Low and high effort
The guidelines direct the quality raters, as in:
Consider the extent to which a human being actively worked to create satisfying content.
Effort is not necessarily manual composition, according to the guidelines. For example, an individual can create a layout that helps readers, such as a unique product page that includes customer reviews, specifications, comparisons with similar products, and applications.
“Low-effort” repurposes existing content without adding sources, expert opinions, or even the author and publication date.
Cyrus Shepard is the founder of Zippy Signal, a search optimization firm, and a former Google quality rater. In a post last month, he shared his experiences as a rater, stating the training repeatedly equated “effort” with “quality.”
He provided these examples:
| Low effort | High effort |
|---|---|
| Restates information | Includes original facts and perspectives |
| Lists well known options | Shares methodology on how a list was created |
| Includes commonly known facts | Includes detailed, lesser-known facts with expert opinions |
| Stock, AI-generated images | Original screenshots or photography |
| Summarizes others | Includes personal experiences |
“Low effort” and “commodity” content are often synonymous. I’ve addressed the latter, citing a Google Search Central post:
- “Non-commodity” content establishes authority and becomes a source for surveys, proprietary data, and whitepapers.
- “Content with effort” is a sitewide editorial rule for all published material: articles, comparisons, images, and more.
‘Effort’ impacts rankings
Content effort is likely a component of Google’s search algorithm.
Shaun Anderson, the founder of Hobo, an SEO agency and blog, cites an internal Google attribute called contentEffort in last year’s algorithm leak. Its existence is telling.
Many professional search optimizers, including me, believe contentEffort is now part of Google’s helpful content evaluation.
There’s no objective standard for “effort,” but here are two tools to help:
- Content Effort Grader from Eggnite, an agency, “reverse engineers” the contentEffort attribute to assign a page grade and to suggest ways to demonstrate effort.
