In September 2026, SEO researcher Cyrus Shepard published results from a survey of 131 SEO practitioners who rated more than 100 potential Google ranking factors, producing 13,665 individual data points. The findings are not a Google-confirmed list of ranking factors, but they are a real, methodologically transparent snapshot of what experienced practitioners currently believe moves rankings, and several of the results line up directly with practices this site has followed throughout this blog series, particularly around content quality, internal linking, and the risks of low-effort AI content. This post walks through the survey's findings across four categories, relevance, content quality, internal linking, and user signals, and translates what each means for a law firm's own content decisions.
What Was This Survey, and How Was It Conducted?
The survey asked 131 named SEO practitioners to rate over 100 potential ranking factors on a seven-point scale, from strongly negative to strongly positive, and separately asked each participant an open-ended question about which three factors they believed mattered most. Combined, this produced 13,665 data points across the full set of factors (Cyrus Shepard, Zyppy Signal, 2026 Google Ranking Factors Expert Survey). It is worth being precise about what this data is and is not. It is not Google's own confirmation of what its algorithms weigh, and it is not a controlled experiment. It is a structured aggregation of expert opinion, disclosed with its methodology, which makes it a legitimate data point to reference and a poor source to treat as definitive proof of any single mechanism.
Across the open-ended "top three factors" question, the most commonly named categories were relevance, named by 57.1 percent of respondents, backlinks at 54.8 percent, content quality at 47.6 percent, authority and trust at 36.5 percent, and behavioral or click signals at 29.4 percent, with brand signals, user satisfaction, technical SEO health, topical authority, and internal links rounding out the top ten. It is worth noting what did not make this list as prominently as some industry discussion might suggest: individual technical tweaks and narrow on-page tactics were rated far less influential by these experts than broad, substantive qualities like relevance and content quality.
What Did the Survey Find About Content Relevance?
Relevance was the single most commonly named top-three factor in the entire survey, and the seven-point ratings back that up clearly. Search intent match rated highest of any individual factor measured, at plus 2.60, followed by page semantic relevance at plus 2.34, site topical authority at plus 2.29, and title relevance at plus 2.26. Passage-level relevance, language match, and geographic relevance all rated strongly positive as well, each in the plus 1.85 to plus 2.00 range.
Keyword-based signals showed a clear hierarchy worth noting. Keywords appearing in an H1 heading rated plus 1.67, meaningfully stronger than keywords in subheadings at plus 1.27 or keywords in the URL at plus 1.07. An exact-match domain rated only plus 0.94, weaker than any of the on-page keyword placements, which cuts against the older assumption that owning an exact-match domain name carries outsized weight.
One result worth flagging specifically for a smaller firm with limited resources: meta description rated only plus 0.22, the weakest factor in the entire relevance category by a wide margin. Experts in this survey do not see meta description content as a meaningful ranking lever, which suggests time spent obsessing over exact meta description wording is generally better spent on the content itself, even though a well-written meta description can still matter for click-through rate in ways this particular ranking-focused survey was not designed to measure.
What Did the Survey Find About Content Quality?
This category produced some of the sharpest contrasts in the whole survey. Original research and first-party data rated highest at plus 2.19, followed closely by sitewide content quality at plus 2.07 and content originality at plus 2.02. Information gain, content freshness, demonstrated experience, content depth, page answers related questions, content effort, and creator expertise or authority all rated strongly positive as well, generally in the plus 1.6 to plus 1.9 range. Factual accuracy rated plus 1.52, and matching expert consensus on a topic rated plus 1.23, both meaningfully positive but notably lower than the content-effort and originality signals above them, suggesting raters weigh how much genuine work went into content somewhat more heavily than whether it perfectly mirrors existing expert opinion.
Two results stand out as directly relevant to how this site approaches content. Content length alone rated only plus 0.52, meaningfully weaker than nearly every other quality factor measured, which is a real data point against the common assumption that longer content automatically performs better. Image and video quality rated plus 1.02, and image alt tags rated plus 0.77, both solidly positive but well below the substantive content-quality factors at the top of the category.
On the negative side, ordinary AI-generated content rated negative 0.50, and scaled AI-generated content, produced in volume with minimal editorial oversight, rated negative 1.82, the single most negative factor found anywhere in the entire survey. Stock photos rated negative 0.66 and pages with affiliate links rated negative 0.53. The gap between ordinary AI-generated content and scaled AI-generated content is itself a notable finding: the raters did not treat AI assistance as inherently disqualifying, but treated volume production without real editorial oversight as a serious liability.
What Did the Survey Find About Internal Linking?
This category has direct, practical implications for site structure. Authority of the internal pages linking to a given URL rated highest at plus 1.92, followed by internal anchor text relevance at plus 1.66, the number of internal links pointing to a URL at plus 1.49, and links from sitewide navigation at plus 1.47. Click depth from the homepage rated plus 1.06, and internal anchor diversity, using varied but still relevant anchor text across different links to the same page, rated plus 0.91. On the negative end, an orphaned page, one with no internal links pointing to it at all, rated negative 1.79, one of the most negative results in the entire survey outside the AI-content category.
This finding validates a specific, concrete priority: a page's own content quality is not enough if nothing on the rest of the site links to it, and where the links come from, not just how many exist, carries real weight according to this data. A single link from a high-authority page on a site appears to matter more than several links from low-authority pages, based on how far ahead "authority of internal linking pages" rated compared to raw link count.
What Did the Survey Find About User Signals?
Satisfaction and task completion rated highest in this category at plus 2.14, followed by long clicks, a visitor spending real time on a page rather than immediately leaving, at plus 1.95, and organic click-through rate at plus 1.83. A user selecting a result as their preferred source rated plus 1.48. On the negative side, a user returning to the search results page shortly after clicking through, a pattern often described as pogo-sticking, rated negative 1.04, the only clearly negative factor in this category.
Taken together, these results describe a coherent picture: content that actually answers the question well enough that a visitor stays, rather than content engineered to capture a click and nothing more, is the pattern experts associate with positive ranking outcomes. A firm optimizing purely for click-through rate, at the expense of actually satisfying the visitor once they arrive, is optimizing for a factor these experts rated as meaningfully less important than what happens after the click.
(Source: 2026 Google Ranking Factors Expert Survey, Zyppy Signal)
Category Strongest positive Rating Strongest negative Rating Content Relevance Search Intent Match +2.60 None reported strongly negative N/A Content Quality Original Research / First-Party Data +2.19 Scaled AI-Generated Content -1.82 Internal Linking Authority of Internal Linking Pages +1.92 Orphaned Page -1.79 User Signals Satisfaction / Task Completion +2.14 Return to SERP -1.04 This table summarizes selected findings from the cited survey; all figures are the survey's own reported ratings, not Lawways Marketing data, and are included with attribution to the original source above.
What Does This Data Mean for a Law Firm's Content Specifically?
Several of these findings connect directly to guidance covered elsewhere in this series, and it is worth being explicit about that rather than letting the connection stay implicit. The strength of original research and first-party data as a content quality factor supports the broader case for genuinely useful, citable reference content, the same logic behind the glossary series this post belongs to, rather than thin, templated pages. The weakness of content length as a standalone factor supports writing to the length a topic actually needs rather than padding content to hit an arbitrary word count. The strong showing for real experience and expertise signals lines up directly with the case made in our glossary entry on E-E-A-T for law firms, and the emphasis on search intent match and clear relevance signals reinforces the answer-first, question-structured approach covered in our glossary entry on GEO for law firms.
The internal linking findings carry a specific, practical implication worth stating plainly: a page with strong content but no internal links pointing to it from elsewhere on the site is working from a real disadvantage according to this data, not a minor one. For a law firm with dozens of practice-area, location, and blog pages, deliberately maintaining internal links, rather than treating them as an afterthought once new content is published, is one of the more concrete, low-cost actions this survey's data supports.
The scaled AI-generated content finding is worth a direct word of caution. A firm tempted to generate a large volume of thin AI content quickly is, according to this survey's respondents, working against one of the most negatively rated factors measured anywhere in the data, roughly as negative as an orphaned page with zero internal links. The alternative, fewer pages produced with real editorial effort, fact-checking, and genuine expertise, aligns with nearly every positively rated factor in the content quality category. Taken together with the internal linking findings, the data points toward a specific operating principle: publish less, link deliberately, and make sure every page reflects real effort rather than treating either volume or link count as a substitute for the other.
If you want help auditing your firm's content and internal linking against this kind of data, you can see our full law firm SEO services or talk to us about your firm's content strategy.
Frequently Asked Questions
Is this survey Google's official list of ranking factors?
No. It is a survey of 131 SEO practitioners' opinions on what they believe influences rankings, not a confirmation from Google of how its algorithms actually work. It is a useful, transparently sourced data point, not a definitive technical specification.
What was the single highest-rated factor in the survey?
Search intent match rated highest of any individual factor measured, at plus 2.60 on the survey's seven-point scale, and relevance overall was also the most commonly named top-three factor in the open-ended portion of the survey.
What was the most negatively rated factor?
Scaled AI-generated content, meaning AI content produced in volume with minimal editorial oversight, rated negative 1.82, the most negative rating of any factor in the entire survey.
Does content length actually matter for rankings, according to this data?
Not much on its own. Content length rated only plus 0.52, notably weaker than most other content quality factors measured, which suggests writing to the length a topic genuinely needs matters more than hitting a specific word count.
Why does internal linking matter so much according to this survey?
Two internal linking factors rated among the strongest and weakest in the entire survey: the authority of pages linking to a URL rated plus 1.92, while having no internal links at all, an orphaned page, rated negative 1.79. Together they suggest internal link structure carries real weight independent of a page's own content quality.