Ranking

Factors
 Ranking Factors Study 2024

Table of
Content
03. About Semrush
04. Ranking Factor Study
06. Executive Summary
22. Introduction
23. All Ranking Factors
Content-Related Metrics
Backlink Metrics
URL and Domain Metrics
On-SERP Factors
User Experience
User Signals

89. Key Learnings

94. Methodology
95. Limitations
96. Benchmarks

02
 Ranking Factors Study 2024

About
Semrush
Semrush empowers marketers and business owners worldwide with the reliable data
and actionable insights they need to drive measurable results from their marketing campaigns.
We’re continually investing in and working on delivering the very best products and data-from
building the largest keyword database on the market to the fastest backlink discovery tool.
Our databases contain search data spanning 190 regions and countries, 25.4 billion keywords,
43 trillion backlinks, 808 million domain profiles, and more. Hundreds of thousands of users in
142 countries, including 30% of the Fortune 500, already know they can rely on us for the
freshest data to help them build winning marketing strategies.
Our expert engineering and product teams are continuously building and expanding the platform
to help our customers stay on top of developments in Search and identify new opportunities.
With more than 50 products, tools, and add-ons in a single platform, Semrush makes it easy
to build, track, and optimize marketing campaigns across various channels, including SEO,
advertising, social media, digital PR, and content marketing.
Whether you’re a seasoned industry leader or a small business owner just getting started,
Semrush can help you grow your online presence in ways you had never imagined before.

03
 Ranking Factors Study 2024

Ranking
Factors Study
We know that measuring the relevance of individual ranking factors and the potential impact
of a single factor is difficult, if not impossible.
We don’t have access to the same data as Google, which is able to track user engagement
on the SERPs. Nor do we know how Google applies certain weights to these factors.
Therefore, this study should be an inspiration, as SEO is often more challenging than people
think. Our insights, and examples that influence rankings, should spark a conversation about
the many different facets of SEO and help SEOs think outside the box.
Achieving consistent top rankings in Google is more than just “great content” or a “great user
experience.” The job of an SEO is often like a detective investigating a crime scene, trying
to make sense of the clues and anticipate patterns.
That’s why, motivated by the Yandex code leak, which brought a lot of insight into how Yandex
works as a search engine, we were driven to create this ranking factor study.
We use traditional ranking factors along with some new ones, like measuring the relevance
of a page to the query, using word embeddings, which is similar to how Google deconstructs
content. And we use features to determine how strong a domain is using direct traffic (branded
traffic) and correlate that with rankings.
Our goal is to show that SEO is not just about SGE (search generative experience), and that
reverse engineering what makes a good ranking is still a core part of SEO.
SGE is currently the hottest topic in SEO-with good reason. The breakthrough success
and usefulness of ChatGPT will affect Google and SEO in the short and long term.

04
 Ranking Factors Study 2024

SERPs will change as Google will be able to answer even more queries directly on the SERP
and keep the user in the Google ecosystem (see also our Zero Click study). But at the same time,
we doubt that SGE will influence every query a user makes. In general, SGE is useful for more
informative queries, and less so for commercial ones. At the same time, we see Google changing
its SERPs significantly for commercial queries.
More implications could change SGE:
• Using data from other sites to keep them in the Google ecosystem could break the unspoken
contract with site owners, as Google sends traffic to sites and in return could use the “content”
from sites to place ads on top of them.
• The cost associated with generating results using LLMs is at least 10 times more expensive
than “linear” search, which will have a massive impact on Google’s costs if it uses SGE for every
query.
We hope that our new study provides insights that inspire you, that you can test and apply some
of the findings, and that your conversations about SEO with your colleagues and stakeholders
can be influenced by some of our findings. Please feel free to share our study and give us
feedback on what we can improve next time.

Marcus Tober,
SVP Enterprise Solutions at Semrush

05
 Ranking Factors Study 2023

Executive Summary
Factors Influence, Top-20 Results
Correlation of a factor with the position on SERP

06
Ranking Factors Study 2024

07
Ranking Factors Study 2024

08
 Ranking Factors Study 2024

Content Factors
Content Factors Correlation with the Position on SERP
Top-20 Results, Based on Averages

09
 Ranking Factors Study 2024

Content Factors Strength

Top-20 Results

10
 Ranking Factors Study 2023

Factor Top-10 average Top-10 median

Text relevance 90.6% 93.9%

Content quality score 76.9% 84.8%

Keyword coverage 67.8% 75.1%

Images alt tags count 31 17

Images count 34 18

Schema.org count 0.2 0.0

Percentage of pages with Schema.org 19.9% -

Page word count 1,451 1,069

Meta description word count 23 23

Video count on page 0 0

Percentage of pages with author Schema markup 5.2% -

Sentence count 53 31

Lists count (<ul> tags) 18 12

Paragraphs count 36 18

Tables count 1 0

Percentage of pages with tables 17.1% -

Readability score 51 58

Content age (in days) 714 378

Title word count 8 8

Focus on creating relevant, quality content

Content relevancy and quality showed the strongest correlation with higher rankings. Don’t waste
your time obsessing over keyword count or updating content dates for the sake of making your
content appear fresher. Focus on the things that matter, which is creating content that covers its
topic in a way that meets readers’ needs.

11
 Ranking Factors Study 2024

Backlink Factors
Backlinks Factors Correlation with the Position on SERP
Top-20 Results

Backlinks Factors Strength

Top-20 Results, Based on Averages

12
 Ranking Factors Study 2023

Factor Top-10 average Top-10 median

Domain authority score 67 72

Page authority score 20 18

Number of referring IPs (domain) 1,9611,256.4% 37,438

Number of referring domains

1,083,945 58,131
(domain)

Number of referring IPs (URL) 114 7

Number of referring domains (URL) 112 6

Number of backlinks (domain) 2,335,555,799 9,683,696

Number of backlinks (URL) 2,418 13

Earning backlinks remains important in an SEO program

Earning backlinks from unique domains is still important, both at page and at domain level. You
may not always need a big number of backlinks, but they can help you build your topical authority
over time and drive traffic to your pages.

13
 Ranking Factors Study 2024

URL & Domain Factors

URL & Domain Factors Correlation with the Position on SERP
Top-20 Results

14
 Ranking Factors Study 2024

URL & Domain Factors Strength

Top-20 Results, Based on Averages

15
 Ranking Factors Study 2023

Factor Top-10 average Top-10 median

URL's organic traffic 10,678 735

Domain's organic traffic 398,282,102 214,715,033

URL's organic positions in top-20 76 16

Domain occurrences in Reviews 743,089 34,195

Domain's organic positions in top-20 1,436,906 95,346

Domain occurrences in Images 817,526 31,176

URL occurrences of Star Ratings 38 10

Domain occurrences in Featured

5,668 506
Snippets

URL occurrences in Images 44 13

Domain's age 20 22

URL occurrences in Featured

13 3
Snippets

Is subdomain 0 -

If nTLD 1 -

URL parameters count 0 0

https 98% 100%

URL length subfolders 2 2

URL length characters 65 60

Focus on improving your organic rankings

Invest in improving your rankings. It can lead to a ripple effect where you might prove your
content to be ideal, so that you might also rank for other highly related queries.
Consider this: Moving from position 2 to position 1 can lead to an increase of +50% in organic traffic.
Start with pages on your website with which you’re already ranking on the first page of Google.
Invest time into updating and improving them in order to improve their organic performance.

16
 Ranking Factors Study 2024

On-SERP Factors
On-SERP Factors Correlation with the Position on SERP
Top-20 Results

On-SERP Factors Strength

Top-20 Results, Based on Averages

17
 Ranking Factors Study 2023

Factor Top-10 average Top-10 median

URL-keyword similarity [BERT] 1 1

Title-keyword similarity [BERT] 68.6% 70.0%

Percentage of pages with keyword in title 22% -

Number of words from a keyword in title 3 2

Percentage of pages with keyword in description 17% -

Number of words in description 25 25

Number of words in title 7 8

Include relevant terms in your titles and descriptions

For this whole category, we found minimal or no correlations between the factors we investigated
and higher rankings.
What became clear is that including exact match keywords in the title or the meta description
does not seem to play a role when optimizing your pages to achieve higher rankings.
On the contrary, we found slightly higher correlations when we analyzed terms’ inclusion in titles
and description based on their similarity to the main keyword (using embeddings). This means
that naturally including keywords that are relevant to your content and the users’ intent is more
important.

18
 Ranking Factors Study 2024

User Experience Factors

User Experience Factors Correlation with the Position on SERP
Top-20 Results

User Experience Factors Strength

Top-20 Results, Based on Averages

19
 Ranking Factors Study 2023

Factor Top-10 average Top-10 median

URL-keyword similarity [BERT] 1 1

Title-keyword similarity [BERT] 68.6% 70.0%

Percentage of pages with keyword in title 22% -

Number of words from a keyword in title 3 2

Percentage of pages with keyword in description 17% -

Number of words in description 25 25

Number of words in title 7 8

Across all the factors in this group, we found minimal or no correlation to higher rankings.
However, each of them separately and all of them collectively can impact user experience on your
site. This, in turn, can shape users’ decisions like whether to stay and interact or bounce off.
The key takeaway here is that you should monitor your performance for these metrics not in order
to hit an “ideal” score, but as an indication of how technically sound the user experience you offer
is.

20
 Ranking Factors Study 2023

User Signals
User Signals Correlation with the Position on SERP
Top-20 Results

User Signals Strength

Top-20 Results, Based on Averages

Factor Top-10 average Top-10 median

Direct traffic share (domain-based) 31.2% 29.1%

Time on site (domain-based) 0:09:42 0:08:34

Branded search traffic share

23.8% 28.5%
(domain-based)

Bounce rate (domain-based) 60.5% 59.5%

Build your brand

Almost 1 out of 3 visitors to domains of top-ranking pages goes to these websites directly. And 1
out of 4 does so by typing a brand query in Search. This demonstrates the importance of raising
awareness around and establishing your brand.
21
 Ranking Factors Study 2024

Introduction
How does Google rank content?
This is the top question for anyone working on building a website’s organic visibility.
Over the years, Google has become more transparent in sharing information and guidance
with webmasters and SEOs. We now know more than ever before about what its algorithm looks
to reward, the different systems in place for different types of content or tactics, major updates
rolling out and what they’re addressing, their guidelines around quality content, developments
in its machine learning, and AI capabilities.
But conflicting or generic advice online, regular algorithmic and system updates,
and discrepancies between official guidance and the content that is rewarded on the SERPs can
often lead marketers to confusion.
We wanted to use the one thing that can help dissolve confusion: data.
So we set off to investigate the correlation between known (confirmed and unconfirmed) factors
and rankings.
One thing needs to be clear from the start: Correlation doesn’t equal causation. It would be
a mistake to interpret these findings as “do this and you’ll rank”.
We analyzed a sample of 16,298 English keywords, all of which have more than 100 monthly searches. For
each keyword, we collected the search engine results pages (SERPs) and analyzed the top 20 positions
and their performance across a number of factors, ending up with a total of 300,000 positions.
Our goal was to understand how specific factors correlate with search engine rankings. To do this,
we calculated a correlation score for each factor.
You’ll find our discoveries and observations in this report.
From the importance of content relevance and quality to our findings around the degree
of correlation with direct traffic share, we hope this report opens up new avenues for your
marketing strategy and helps you challenge some long-standing ideas.
P.S. We want to thank Mordy Oberstein for doing a peer review of this report. His ideas,
observations, and feedback helped shape this document into what you see.

22
 Ranking Factors Study 2024

All Ranking Factors

We initially looked into over 90 potential ranking factors. For the study, we kept 65 that either
showed correlation with search rankings or presented a special interest for further discussion:
• Content-related metrics: Metrics around content structure and quality, including text
relevance, content quality, content length and age, content elements (images, lists, tables,
and Schema elements), and more.
• Backlink metrics: Backlink-focused metrics, such as the number of referring domains,
Authority Score, total number of backlinks, and more.
• URL and domain metrics: Metrics around URL and domain performance (organic
performance, appearance in SERP features, domain age, etc.).
• On-SERP factors: Metrics related to elements you can see in search engine results pages
(titles, descriptions, and page URLs-as seen on the SERPs vs. as defined by webmasters).
• User experience metrics: Metrics related to site speed, such as Cumulative Layout Shift, First
Contentful Paint, First Input Delay, Largest Contentful Paint, Total Blocking Time, Time To
Interactive, and Server Response Time.
• User signals: Metrics related to user signals, such as direct and branded search traffic share,
time on site, and bounce rate.
We analyzed the correlation between all factors and SERP positions utilizing a Spearman
correlation metric between a position and a factor’s average value per position. This helped us
assess how strong the connection is between the numerical value of a factor (e.g., the number
of backlinks) and its position on the SERP.
We use an “inverse” correlation (add minus to each resulting coefficient). We do this in order
to show a more natural direction: e.g., if our value increases with every decrease of position, it is
a positive correlation for us (the larger the value, the higher the position).
We have put all factors’ correlation values below.

23
 Ranking Factors Study 2024

Factors Influence, Top-20 Results

Correlation of a factor with the position on SERP

24
Ranking Factors Study 2024

25
Ranking Factors Study 2024

26
 2024
Ranking Factors Study 2023

In this study, we also introduce “factor strength” by calculating the percentage difference
between the average for the top 3 positions and positions 16-20.

How to use factor strength:

You’ll notice that for some factors, the strength score is higher compared to others. Factors
with a high strength score are those for which the difference between lower and higher SERP
positions was the highest. It represents the overall factor value volatility within the top 20 results.
Below, we’ve listed all ranking factors based on their strength (from highest to lowest).

27
 Ranking Factors Study 2024

Factors Strength
Top-10, Based on Averages

28
Ranking Factors Study 2024

29
Ranking Factors Study 2024

30
 Ranking Factors Study 2024

Content-Related Metrics
We also examined content-related metrics, including:
• Text relevance: We used embeddings (numerical representations of text) for this factor.
By using BERT model embeddings to compare content on a page to content on the SERPs,
we were able to assess the similarity (or “relevance”) between all content on the SERPs
and content on specific URLs. This allowed us to explore content factors beyond the usage
of exact match keywords but also based on semantics and context.
• Keyword coverage: A score that reflects the extent to which terms, specific to the search query,
are covered. For example, for the search query “Iphone 14 review,” keywords like “camera,” “display,”
and “price” will appear a lot on SERPs. If we use these terms on our page (with corresponding
occurrences), we get a higher Keyword Coverage score. Here we use TF-IDF to calculate the metrics.
• Content quality: A combination of the keyword coverage, embeddings, word count,
and sentence count scores. All of these factors are weighted to allow us to assess not just
relevant term usage but also density (spotting any keyword stuffing cases in the SERPs). This
metric is useful when you need to estimate overall page performance in terms of content.
• Content elements on the page: The number of images, lists, tables, and Schema elements
on the page.
• Content length: Length of content on the page in terms of word, sentence, and paragraph count.
• Authorship: Whether the page lists an author.
• Readability score: The Flesch Reading Ease score of the page.
• Content age: How old the page is (in days).
We found that text relevance and content quality correlate the most with SERP positions.

31
 Ranking Factors Study 2024

Content Factors Correlation with the Position on SERP

Top-20 Results

32
 Ranking Factors Study 2024

Text Relevance
Our data shows that top-ranking pages are more relevant to the search query compared to lower-
ranking pages. This perfectly aligns with how Google describes the process behind ranking
content and listing content relevance as one of the main factors it uses.
Notice the high level of relevance for URLs ranking in the top 5 positions, as well as the steep
downward trend for pages from position 6 to position 20.

Text Relevance

Here’s an example: Looking at the top-ranking pages for “icebreakers for adults” (2,900 MSV,
S Semrush Edit Page AM
KD: 68%), we can see that a page from Teambuilding.com (“17 Fun Icebreaker Games for Adults”)
ow
ranks atAnalysis
Content the top, with a Relevancy score of 89%.
Details

ormance icebreakers for adults teambuilding.com Australia 11/03/2023, 11;32 AM Refresh data

lysis Total Score +BCA/ Relevancy Score +BCA/ Keyword Coverage +BCA/ Word Count +BCA/

78.0% 89.0% 68.0% 2,766

Insights

0.0% 78.0% 0.0% 89.0% 0.0% 68.0% 0 2,766

Deviation Avg. Top 5 Deviation Avg. Top 5 Deviation Avg. Top 5 Deviation Avg. Top 5

All terms for content analysis Export

Let’s compare it to the page on elfster.com that currently ranks in position 14 (“5 Truly Unique
Term Translation Term Group Your Relevancy Average Relevancy Delta Relevancy Your Keyword Cover… Average Keyword Co… Delta Keyword Cover…

meeting n/a must have 30.1% 45.5% 15.4 3 16.8 14

Adult Ice Breaker Games and Discussion Questions”):

virtual n/a good to have 7.7% 22.9% 15.2 1 7.7 7

group n/a must have 23.5% 35.7% 12.2 9 23.5 15

participants n/a must have 32.4% 43.7% 11.3 0 14.6 15

33
room n/a good to have 15.3% 26.7% 11.3 0 5.5 6

things n/a must have 24% 34.1% 10.1 0 7.4 8

 Ranking Factors Study 2024

S Semrush Edit Page AM

Content Analysis Details

w
icebreakers for adults atlassian.com Australia 11/02/2023, 06;36 PM Refresh data

mance

Total Score ?BCAA Relevancy Score ?BCAA Keyword Coverage ?BCAA Word Count ?BCAA
sis

sights
53.0% 74.0% 59.0% 1,354
G32.05% 78.0% G16.85% 89.0% G13.24% 68.0% G51.05% 2,766
Deviation Avg. Top 5 Deviation Avg. Top 5 Deviation Avg. Top 5 Deviation Avg. Top 5

All terms for content analysis

The page on Teambuilding.com has a much higher Relevancy score than the one on Elfster.com.
Export

Term Translation Term Group Your Relevancy Average Relevancy Delta Relevancy Your Keyword Cover… Average Keyword Co… Delta Keyword Cover…

Although the word count is not significantly different between the two articles (2,766 words
ask n/a must have 8% 27.7% 19.7 4 10.6 7

for the Teambuilding.com article vs. 2,424 for the elfster.com article), the former lists many more ideas
questions n/a must have 14.8% 34.1% 19.3 2 12 11

random n/a good to have 7.5% 23.9% 16.4 0 2.9 3

for its readers to get inspired from, whereas the latter only lists five and goes in-depth for each of these.
favorite n/a good to have 14.2% 28.1% 13.9 1 5.1 5

play n/a good to have 23.4% 37.1% 13.8 0 7.2 8

With this approach, the first article manages to cater for user intent better and to include queries
fun n/a must have 34.1% 46.3% 12.3 2 13.8 12

that are more relevant to the main term naturally.

person n/a must have 15.9% 28.1% 12.2 9 20.9 12

pick n/a good to have 13.2% 24.8% 11.6 2 3 1

adults n/a good to have 16.3% 27.8% 11.5 0 5.5 6

Content Quality
story n/a good to have 23.7% 34.7% 11 0 5.6 6

ion Tool

1"10 of 70

We also found that top-ranking pages have a higher content quality score, on average, compared
to lower-ranking pages.

Content Quality Score

Google offers guidelines on what constitutes quality content here. Keep in mind that quality
doesn’t just refer to the text on the page, but also to the overall quality of the user experience
on your website (including the design, layout, and other aspects of user experience).
34
 Ranking Factors Study 2024

Keyword Coverage
The correlation between this factor and higher Search rankings is small (0.08). However, we found
that top-ranking pages cover more terms related to the main topic compared to lower-ranking
pages.

Keyword Coverage

Look at this example of two of the top-ranking pages for “nonverbal communication” (MSV:
S Semrush Edit Page AM

18,100, KD: 82%).

Content Analysis Details
The page on BetterUp.com (currently ranking #5) has 100% keyword coverage.
w

mance nonverbal communication betterup.com United States 11/02/2023, 02;49 PM Refresh data

sis Total Score -BCA1 Relevancy Score -BCA1 Keyword Coverage -BCA1 Word Count -BCA1

100.0% 100.0% 100.0% 2,041

sights

67.53% 93.0% 61.01% 99.0% 616.28% 86.0% 0 2,041

Deviation Avg. Top 5 Deviation Avg. Top 5 Deviation Avg. Top 5 Deviation Avg. Top 5
S Semrush Edit Page AM

All termsAnalysis
Content for content analysis
Details Export

On the other hand, the page on Forbes.com (in position #9) presents 71% keyword coverage.
w

Term
nonverbal communication Translation
forbes.com Term Group02;31 PM
United States 11/02/2023, Your Relevancy
Refresh data Average Relevancy Delta Relevancy Your Keyword Cover… Average Keyword Co… Delta Keyword Cover…
mance

nonverbal cues n/a must have 59.3% 56.6% I2.7 2 4.2 3

sis Total Score +BCA/ Relevancy Score +BCA/ Keyword Coverage +BCA/ Word Count +BCA/
touch n/a must have 48.9% 36.3% I12.7 7 7.9 1

83.0% 98.0% 71.0% 1,385

sights
verbal n/a must have 44.9% 45.2% 0.3 2 5.3 4

conversation n/a must have 49.5% 45.9% I3.7 6 4 I2

510.75% 93.0% 51.01% 99.0% 517.44% 86.0% 532.14% 2,041
nonverbalDeviation Avg. Top 5 n/a Deviation
must have Avg. Top 5 68.3% Deviation
58.5% Avg. TopI9.9
5 6 Deviation 5.9Avg. Top 5 0

gestures n/a must have 49.2% 51.5% 2.3 3 9.9 7

speaking n/a must have 48.1% 46.7% I1.3 5 6.3 2

All terms for content analysis
eye contact n/a must have 62.9% 48.7% I14.2 7 7.4
Export
35
1

Term Translation Term Group Your Relevancy Average Relevancy Delta Relevancy Your Keyword Cover… Average Keyword Co… Delta Keyword Cover…
facial expressions n/a must have 51% 52.9% 1.9 6 5.9 0

nonverbal communication n/a must have 83.3% 71.9% J11.3 18 22 4

communication n/a must have 54.3% 50.3% I4 13 17.2 5
 Ranking Factors Study 2024

Content Elements on the Page

Next, we looked at the number of content elements (images, lists, tables, and Schema markup)
on the page across the top 10 and top 20 ranking pages.
Although the correlation we found between each of these elements and high rankings is minimal
(0.01 to 0.03), our data shows that:
• The top-ranking page has 21% more images than the page ranking in position 20.
• The average number of images for the top 10 pages is 34. The average number of alt tags is
31. This means that top-ranking pages use alt tags for most of their images-which is something
Google recommends everyone should do.

Images Count

The top-ranking pages have Schema.org mark-up more often. On average, they utilize it across
17% of their pages.

36
 Ranking Factors Study 2024

Percentage of Pages with Schema.org

Our next finding concerns the number of lists on URLs. We found that the URLs in position #1
utilize on average 23 lists. This number drops to 21 for URLs in position #2.
The average list count for organic positions 2 to 20 remains relatively close (ranging from 20 to 21),
but the noticeable difference in the top 2 results is definitely something that caught our attention.

Lists Count

37
 Ranking Factors Study 2024

Content Length
Google has confirmed there isn't a "preferred" word count for ranking. And the correlation we
found also seems to confirm this; it is minimal at 0.02.
When it comes to content length, the data shows that, on average, pages ranking in the top 10
positions have a higher word count compared to those ranking on page 2 of SERPs.
The average word count for the top 10 pages is 1,451 words.

Page Word Count

Authorship
E-E-A-T is a set of criteria Google’s human Search Quality Raters use to determine
the helpfulness and relevance of pages or websites. The Search Quality Raters reports do not
directly impact rankings on SERPs; they are used to improve Google’s ranking systems.
E-E-A-T stands for:
• Experience: Does the author have first-hand experience with the topic?
• Expertise: Is the author or website an expert on the topic?
• Authoritativeness: What is the overall reputation of the author or website in the industry? Are
they considered an authority on the topic?

• Trustworthiness: Is the author or website trustworthy?

38
 Ranking Factors Study 2024

We explored the correlation between demonstrating authorship on the page (through the use
of author Schema.org markup) and top rankings. We found none.
And we detected the relevant markup on only 5% of the pages ranking within the top 10 results.

Percentage of Pages with Author Schema Markup

Due to the nature of our sample (random keyword dataset, all ranking content, which includes
various page types), we believe that our findings largely reflect a big part of the ranking pages
being homepages, ecommerce pages, product/service pages, etc. For this type of content, it is
common not to include an author.
Therefore, these results should not be interpreted as “authorship is not important.”

Readability Score
The Flesch Reading Ease score shows how easy it is to read a specific text. It’s scored from 0
to 100, with higher scores indicating easier readability.
A score between 50 and 60 classifies your content as “fairly difficult” to read in the FRE scale
and makes it suitable for 10th-12th graders. Meaning, it’s harder to read than content hitting 90
or above, which is suitable for a 5th grade student.
Based on our data, the top 5 ranking pages have a considerably lower readability score than
the rest.
39
 Ranking Factors Study 2024

For the pages ranking in the first position, for example, this score was 48.5. For position 14, it was
52.5-which is a whole 4 points higher (and would classify this content as “fairly difficult”).
The average readability score was 51.2, which means pages ranking in the top 3 all fell below this
score.

Redability Score

According to John Mueller, Google doesn’t use readability as a ranking factor. Our findings
confirm this, as we found no correlation here.

Content Age
Over the years, Google has put out a number of updates that tackle content freshness, including
“query deserves freshness” systems, which favor newer content for queries where it makes sense
(e.g., news, trending topics, etc.).
We wanted to investigate whether we could find a general correlation between content age
with rankings.
We found none.
However, what we found is that on average, the pages in the top 3 positions are older compared
to those in positions 4 to 10. The average content age for pages in the top 10 positions is 713
days.
40
 Ranking Factors Study 2024

Content Age (in days)

Caveat: Our keyword sample is not split into categories based on trending vs. evergreen queries.
To this extent, our findings on content age offer an overview of the age of ranking pages,
with the limitations this carries.

Key Takeaways
Focus on creating relevant, quality content
Content relevancy and quality showed the strongest correlation with higher rankings. This is
on par with Google’s consistent guidance to create helpful, reliable, people-first content.
Don’t waste your time obsessing over keyword count or updating content dates for the sake
of making your content appear fresher. Focus on the things that matter, which is creating content
that covers its topic in a way that meets readers’ needs.

Look beyond keywords

Writing with your audience and topic front of mind should lead to including relevant terms
naturally.
Place your efforts in covering your topic in a comprehensive way instead of counting exact match
keyword occurrences in it.

41
 Ranking Factors Study 2024

Images can bring value

Images can help improve engagement, increase the helpfulness of your content, and make it
easier for readers to digest the content on the page.
They are also a great way to rank organically in image search results and features (especially when
optimized for search).

Readability should still be a consideration

Ensuring your content is easy to read, digestible, and clear can lead to more engaged
and satisfied users-which in turn can also impact conversions.

Demonstrate E-E-A-T sitewide

Talk about the things you are an expert in or have extensive experience in. Take care of trust
signals across your website pages, from clearly demonstrating authorship where it is relevant
to showcasing why you are entitled to talk about the topics you do.

42
 Ranking Factors Study 2023

Backlink Metrics
Next, we looked at the following link-related metrics:

• Authority Score: The Authority Score for the page (PAS) and the domain (DAS).

• Number of referring domains: The total number of unique referring domains for the page
and the domain.

• Number of backlinks: The total number of backlinks for the page and the domain.

• Number of referring IPs: The total number of referring IP addresses for the page and the domain.
Based on our data, higher Domain and Page Authority Scores correlate with higher SERP
positions the most.
Similar to previous parts of this study, many of the metrics we will examine here can arguably
be the result of pages ranking higher. For example, is the number of backlinks a URL has simply
correlated to higher rankings or the outcome of these?
Same as before, we want to acknowledge this perspective on the results and continue with our
analysis based on our findings.

Backlinks Factors Correlation with the Position on SERP

Top-20 Results

43
 Ranking Factors Study 2024

Authority Score
Authority Score is Semrush’s proprietary metric which aims to grade the authority of a page
(Page Authority Score or PAS) or website (Domain Authority Score or DAS).
In principle, domains with a higher Authority Score are considered more trustworthy. Their
outgoing links are also considered more valuable.
It’s calculated based on three main factors:
• Link power: The number and quality of backlinks that a website or page has.
• Organic traffic: The average number of monthly visits a website or page receives from
Google.

• Spam factors: Indicators of spam or link manipulation in the website’s or page’s link profile.
Google has stated that they don’t use a sitewide authority signal or score.
Our data showed that top-ranking pages tend to be on domains with a high Domain Authority
Score, with the average DAS among the top 10 being 67.42 (maximum is 100).

Domain Authority Score

The same is true for Page Authority Score, with higher-ranking pages having a higher PAS.

44
 Ranking Factors Study 2024

Page Authority Score

Number of Referring Domains

We also found that higher-ranking pages have a higher number of referring domains. This is true
at both domain and URL level.

Number of Referring Domains (domain)

45
 Ranking Factors Study 2024

Number of Referring Domains (URL)

Note the big difference between the average and the median values in the number of referring
domains in both cases.
While the average values may seem daunting to most of us, if you look at the median values, these
are much more feasible.
This is due to some huge websites and their pages taking up the top-ranking spots in certain
cases-sites like Amazon.com or Apple.com that have accumulated enormous volumes
of backlinks over the years.

Number of Backlinks
When it comes to the number of backlinks, domains of top-ranking pages tend to have more
backlinks compared to domains of lower-ranking pages. The same is true for the pages
themselves, with higher-ranking pages having more backlinks compared to lower-ranking ones.

46
 Ranking Factors Study 2024

Number of Backlinks (domain)

Number of Backlinks (URL)

On average, top-ranking pages have 2,418 backlinks.

However, the median number of backlinks for top-ranking pages is just 13, which demonstrates
that you don’t necessarily need a huge number of backlinks to rank on the first page of search
results.
47
 Ranking Factors Study 2024

There are still cases where a page in a higher position has fewer backlinks than a page in a lower
position. For example, let’s look at the search results for the keyword “accountant salary” (MSV:
27,100, KD: 62%).

Indeed’s and Salary.com’s pages outrank U.S. News’ and even the U.S. Bureau of Labor Statistics’
pages while having fewer backlinks and lower PAS.

However, if we examine the four domains closer, we see that Indeed has a higher Domain
Authority Score compared to the rest and many more backlinks at domain level.
48
 Ranking Factors Study 2024

Google has said for years they don’t look at links at a domain level. However, when it comes
to adding a new page on a website, they have said they might look at content and context site-wide
to get a better understanding of what the new page is about and where it can be shown in Search.

Number of Referring IPs

We looked at the number of referring IP addresses at domain and page level. This type of analysis
allows us to spot signs of black hat SEO tactics (e.g., links from private blog networks).
In general, you want to have backlinks from a variety of IP addresses in your link profile.
Our data confirms this, with the number of referring IPs at both domain and URL level being
higher for top-ranking pages. The correlation was also strong at 0.18 in both cases.
On average, domains of top-ranking pages have 196,112 referring IPs, while the pages themselves
have an average of 114 referring IPs.

Number of Referring IPs (domain)

49
 Ranking Factors Study 2024

Number of Referring IPs (URL)

Key Takeaways
Monitor your domain’s Authority Score
Monitoring your own website’s Authority Score, as well as that of other domains linking to yours,
can give you a good overview of whether you’re headed in the right direction when it comes
to building your authority.
Our data around backlinks and rankings showed the strongest correlations between PAS
and DAS and higher rankings.

Earning backlinks remains important in an SEO program

Earning backlinks from unique domains is still important, both at page and at domain level. You
may not always need a big number of backlinks, but they can help you build your topical authority
over time and drive traffic to your pages.

Stay clear of black hat link-building tactics

Tracking the number of IP addresses linking to your domain would not add value to your SEO
program. As long as your link-building tactics follow Google’s guidelines and you earn links
naturally, there should be no issues with the referring IPs.

50
 Ranking Factors Study 2024

URL & Domain Metrics

Next, we looked at URL and domain metrics, including:
• URL organic performance: The volume of organic traffic the ranking page gets.
• Domain’s organic performance: The volume of organic traffic the domain the ranking page
belongs to gets.
• Domain and URL occurrences in SERP features: How many times does a page and a domain
appear in SERP features.
• Domain age: The age of the domain (in years).
• URLs on subdomains: The percentage of top-ranking pages that are on subdomains.
• HTTP vs. HTTPS: Whether a page’s domain uses the HTTP or HTTPS protocol.

• URL length: The length of the URL, measured by the number of subfolders in the URL.
We found that a URL’s organic traffic, the domain’s organic traffic, and the number of organic
listings a URL has in the top 20 positions have the biggest correlation with higher SERP positions.
All three factors, by their nature, can also demonstrate what happens after pages achieve higher
rankings-constituting also the outcome of higher rankings on the SERPs.

51
 Ranking Factors Study 2023
2024

URL & Domain Factors Correlation with the Position on SERP

Top-20 Results

52
 Ranking Factors Study 2024

URL Organic Traffic

Higher-ranked pages have more organic traffic compared to lower-ranked pages.
The correlation for this factor is amongst the highest ones in this study (0.33).
What’s interesting to look at here is the big drop-off in organic traffic from position 1 to position 2,
and then to position 3.

URL's Organic Traffic

We previously examined the significant variation in the organic traffic URLs get based on their
organic position. Together with Backlinko, we analyzed 4,000,000 SERPs in order to calculate
the average click-through rate (CTR) for top-ranking pages.
We found that the URL in #1 has an average CTR of 27.6%.
The page in #10? Only 2.6%!
However, this doesn’t mean there are no exceptions.
For example, if you look at the search results for the keyword “business name,” (MSV: 8,100, US
desktop) you’ll see that the page in the top position receives less organic traffic than the pages
in positions 2 and 4.

53
 Ranking Factors Study 2024

In this particular example, what stands out is that the pages ranking in positions 2 and 4 serve
a similar intent-one that is very different from the page ranking in position 1.

URL’s Organic Positions in Top 20

We discovered that higher-ranking pages tend to have more top 20 spots for other keywords
compared to lower-ranking pages.
The correlation (0.22) is among the highest in this study.

URL's Organic Positions in top-20

54
 Ranking Factors Study 2024

Domain’s Organic Performance

We found a strong correlation between domains’ organic traffic and rankings. Domains
with higher-ranking pages tend to also have higher organic traffic levels.

Domain's Organic Traffic

We also found that domains with top-ranking pages tend to also have more top 20 organic
positions.

55
 Ranking Factors Study 2024

Domain's Organic Positions in top-20

Domains and URLs in SERP Features

We found various degrees of correlation between higher rankings and occurrences of/in SERP
features.
This seems to be true across star ratings, images, and featured snippets.
Among all three features, the strongest correlation at both domain and URL level was found
between higher rankings and occurrences of star ratings in the SERPs.

56
 Ranking Factors Study 2024

Domain Occurrences of Star Ratings

URL Occurrences of Star Ratings

57
 Ranking Factors Study 2024

Domain Occurrences in Images

URL Occurrences in Images

58
 Ranking Factors Study 2024

Domain Occurrences in Featured Snippets

URL Occurrences in Featured Snippets

59
 Ranking Factors Study 2024

Domain Age
We found that higher-ranking pages tend to be on older domains, with lower-ranking pages
usually being hosted on relatively newer domains.
The correlation isn’t a very strong one here, though (0.11). In fact, Google’s John Mueller has
stated that domain age doesn’t play a part in ranking pages.
There can be many explanations for why some of the top-ranking pages tend to be on older
domains: these domains may have accumulated more authority over the years, or they have built
their topical relevance and trust over time, and so on.

Domain's Age

URLs on Domain or Subdomain

According to Google, subdomains and subfolders are both fine to use. Google’s official
recommendation is that you use whichever makes more sense for your specific setup.
Our data seems to confirm this, too. We found minimal correlation between whether URLs sit
on domains or subdomains and higher rankings.
Our data also showed that the majority of top-ranking pages (97%) are hosted on root domains.
(Note: This dataset includes all TLD types except for ccTLDs).

60
 Ranking Factors Study 2024

Non-country level TLDs (e.g. .com, .agency. etc)

average top 10 average 11-20 correlation

97% 95% 0.055

How many top-ranking pages are on subdomains? Within the top 10 results, it’s only 11% of the URLs.

Percentage of Pages Hosted on Subdomains

Want to find out more about subdomains and subdirectories and which one is best for you? We've
written a guide for this.

HTTP vs. HTTPS

HTTP (Hypertext Transfer Protocol) is a protocol that is used to load webpages using hypertext
links. HTTPS (Hypertext Transfer Protocol Secure) is an extension of that protocol, which uses
TLS or SSL encryption, making it more secure.
Google is a big proponent of a secure web, and back in 2014 confirmed it's using HTTPS
as a ranking signal. Since then, more webmasters have adopted HTTPS.
In 2023, Google’s guidance is that although HTTPS alongside other page experience aspects
(like Core Web Vitals, mobile-friendliness, etc.) may not be directly used to inform ranking, they
“align with success in search ranking, and are worth attention”.

61
 Ranking Factors Study 2024

Our data shows that over 98% of top-ranking pages use HTTPS nowadays.

average top 10 average 11-20 correlation

98.2% 98.4% 0.007

This doesn’t mean that you can’t rank without HTTPS. We also encountered some HTTP pages
ranking high.
In the example below, the top-ranking page is on a website that uses HTTP (what’s even more
interesting is that this page also outranks a page on Pinterest).

62
 Ranking Factors Study 2024

URL Length
The length of URLs showed minimal to no correlation to better rankings. This aligns
with information previously shared by Google about URL length not being important.
To investigate this factor, we analyzed:
• The number of subfolders (based on number of “/”) in the URL.

• The URL character length between top-ranking and lower-ranking pages.

We didn’t find any correlation between the number of forward slashes in URL and rankings,
and we found only a minimal correlation between URL character length and rankings.
Google has only mentioned one exception regarding URL length. This is when it comes
to canonicalization-if you have multiple copies of a page on your website, Google tends to display
the one with the shortest and cleanest URL in the SERPs.

URL Length Subfolders

63
 Ranking Factors Study 2024

URL Length Characters

Key Takeaways
Focus on improving your organic rankings
All top three ranking factors in this category fall under the “chicken and egg” situation.
For example, organic traffic (to both URL and domain) is a factor with a strong correlation
to higher rankings. Whether this is the result of being at the top of the SERPs or not, the nature
of our study doesn’t allow us to say.
Invest in improving your rankings. The rewards are just too great to pass up. Consider this: Moving
from position 2 to position 1 results in a +50% organic traffic jump on average.
And, even for lower positions, the benefits can be great. For example, simply climbing from
position 7 to position 6 can give you a 20% traffic boost, on average.
Plus, being in that higher spot might also mean ranking for many more keywords and taking up
more space thanks to SERP features.
Start with pages on your website with which you’re already ranking on the first page of Google.
Invest time into updating and improving them in order to improve their organic performance.
Tip: You can use Semrush’s Organic Research to discover which pages on your website rank
between positions 2 and 10 (or 2 and 20) and for which keywords. Then prioritize work based
on the level of opportunity each of them presents.

64
 Ranking Factors Study 2024

Follow best practices when it comes to security and user experience

Providing a good user experience is paramount-not just to your SEO program, but to earning
users’ trust and encouraging conversions.
If you haven’t already, implement HTTPS on your website.

Do not spend resources on trivial SEO signals

When optimizing your website pages, there are so many things you have to consider and take care of.
Avoid wasting your resources on tackling things that have minimal to no impact on your organic
performance. Based on our data, URL length is one of them.

65
 Ranking Factors Study 2024

On-SERP Factors
This group includes factors such as:
• URL-keyword similarity: Whether relevant keyword(s) appeared in the URL (using
embeddings).
• Title-keyword similarity: Whether relevant keyword(s) appeared in the page’s title (using
embeddings and the title as shown on SERPs, whether it matches the specified title by
the webmaster or is written by Google).
• Keyword occurrence in the URL: Whether the target keyword appears in the page URL.
• Keyword occurrence in the title: Whether the target keyword is included in the page title
(as shown on SERPs, whether it matches the specified title by the webmaster or is written by
Google).
• Keyword occurrence in the description: Whether the target keyword is included in the page’s
meta description (as shown on SERPs, whether it matches the specified title by the webmaster
or is written by Google).
• Title length: Title length (in characters, based on the title shown on SERPs, whether it
matches the specified title by the webmaster or is written by Google).
• Description length: The description length (in characters, based on the description shown
in the SERPs, whether it matches the specified description by the webmaster or is written by
Google).
The URL-to-keyword and title-to-keyword similarity showed the biggest correlation to higher
SERP positions. To calculate similarity we used embeddings, allowing us to assess similarity
and relevance beyond exact match keywords.

66
 Ranking Factors Study 2023

On-SERP Factors Correlation with the Position on SERP

Top-20 Results

Title-Keyword Similarity
The title of your page should be relevant both to the user’s search query and to the content
of the page.

Here, it’s important to note that Google often rewrites title tags, seemingly to match user queries
and user intent better.

Nonetheless, it’s still important to include either your main keyword or other relevant keywords in
the title. This goes beyond utilizing exact-match keywords; it refers more to including queries that
are relevant to the page’s content and target user intent efficiently.

67
 Ranking Factors Study 2024

Title-Keyword Similarity [BERT]

URL-Keyword Similarity
While Google has never claimed that this is a major ranking factor, in a 2021 Google Webmaster
Central office-hours hangout session, Google’s John Mueller stated that it’s a small ranking
factor indeed.
In our study, we found that higher-ranking pages in search results tend to include relevant terms
in the URL more frequently than lower-ranking pages, with a higher correlation being seen
for pages in the top 3 positions compared to the pages in positions 16-20.

68
 Ranking Factors Study 2024

URL-Keyword Similarity [BERT]

Keyword Occurrence in the Title

The title of your page, defined by the title tag, is shown as a part of the preview snippet for your
page in Google’s search results. Google had confirmed title tags as a ranking factor back in 2016.
We found that higher-ranking pages tend to include the primary keyword in the title, with pages
on the first page of search results including it more frequently compared to pages on the second
page of search results.
However, note that the correlation we found is really low and that keywords included in the title
tags we analyzed may well have been the result of Google rewriting these on the SERPs (vs.
webmasters specifying these in the page’s code).

69
 Ranking Factors Study 2024

Percentage of Pages with Keyword Title

Keyword Occurrence in the Description

According to our findings, there is a higher frequency of primary keyword occurrence
in the description among top-ranking pages, but the occurrence itself is fairly low-only 17%
of pages in the number 1 spot use it (whether it is webmasters or Google adding these).

Percentage of Pages with Keyword in Description

70
 Ranking Factors Study 2024

Description Length
We also looked at the average length of the description shown in the SERPs for top-ranking
pages. We found no correlation between this factor and rankings.
It also seems that the number of words in the description is lower, on average, for top-ranking
pages.

Number of Words in Description

Title Length
According to our data, there’s no correlation between title length and higher rankings.
We found that for the top 3 results in the SERPs, shorter titles tend to be shown compared
to lower-ranking pages. Pages in positions 4-10 have increasingly longer titles.
Finally, there seems to be a relatively flat trend for positions 10-20.

71
 Ranking Factors Study 2024

Number of Words in Title

Key Takeaways
Include relevant terms in your titles and descriptions
For this whole category, we found minimal or no correlations between the factors we investigated
and higher rankings.
What became clear is that including exact match keywords in the title or the meta description
does not seem to play a role when optimizing your pages to achieve higher rankings.
On the contrary, we found slightly higher correlations when we analyzed terms’ inclusion in titles
and description based on their similarity to the main keyword (using embeddings). This means
that naturally including keywords that are relevant to your content and the users’ intent is more
important.

You can still use the target keyword in the page URL, title, or description
You don’t have to use exact-match keywords in URLs, titles, and descriptions.
At least, not for optimization purposes (as demonstrated by the results of our study).
However, if your content is centered on the topic, including your keyword should come naturally.
And it’s a great way to instantly demonstrate to users and search engines alike what your content’s
main idea is about.

72
 Ranking Factors Study 2024

Note that your content-including your title tags-always needs to be natural. Don’t try to stuff
the keyword (or multiple keywords) in just for the sake of doing it. Keyword stuffing explicitly
violates Google’s spam policies.
Titles and meta descriptions are the gate to your content on the SERPs. Make them enticing
and to the point, and ensure they convey your main message clearly so users are more inclined
to click on your result.
URLs are often created automatically by the CMS. If your keyword is included in your title, it will
probably end up being part of the URL, too.

Don’t worry too much about title tag and meta description length
But try to follow best practices for title tag and meta description content and length, so you
minimize the chances of Google truncating them on the SERPs.
Our suggestion: Keep title tags under 60 characters and meta descriptions under 105 characters.

73
 Ranking Factors Study 2023

User Experience
We also evaluated different metrics that affect site speed to discover their correlation with SERP
positions.
Google doesn't use page speed as a metric, but it does use page experience signals (partly
grouped under Core Web Vitals) to assess the quality of a website’s user experience
and the helpfulness of its content.
In our study, we looked at both field metrics (anonymized, collected from real users) and lab
metrics (simulated in a controlled environment). This allowed us to get a clearer idea of the user
experience real users get when visiting these pages as opposed to just relying on lab data.
Out of the five metrics we analyzed across lab and field, only on three occasions did we find
a slight correlation with higher SERP positions (First Input Delay (FID), Total Blocking Time (TBT),
and Time to Interactive).
This aligns with Google’s stance on the matter: “While not all of these may be directly used
to inform ranking, we do find that all of these aspects of page experience align with success
in search ranking, and are worth attention”.

User Experience Factors Correlation with the Position on SERP

Top-20 Results

74
 Ranking Factors Study 2024

Field Metrics
We looked at the following field metrics:
• Cumulative Layout Shift (CLS)
• First Contentful Paint (FCP)
• First Input Delay (FID)
• Largest Contentful Paint (LCP)

CLS (Cumulative Layout Shift)

Cumulative Layout Shift is a metric that measures the unexpected shifts of website elements
on a page. It’s one of the three Core Web Vitals metrics, a set of metrics that Google uses
to measure websites’ user experience.
A good CLS score is considered to be 0.1 or less (the lower, the better). Our data shows that
the average for the top 10 ranking pages hits that threshold (0.09).
We found no correlation between CLS and ranking higher in the SERPs.

CLS [field]

75
 Ranking Factors Study 2024

FCP (First Contentful Paint)

First Contentful Paint measures how long it takes from when a page starts to load to a part of its
content being rendered on screen. A good FCP score is considered to be between 0 and 1.8
seconds.
When looking at top-ranking pages, we see that they have a lower FCP score compared to lower-
ranking pages, with the average FCP score being 2.02 seconds.

FCP [field]

FID (First Input Delay)

First Input Delay measures the response time from when a user interacts with the page
for the first time (e.g., clicks a button). A good FID score is 100 milliseconds or less.
In our tests, we’ve found a minor correlation for this factor (0.08) and that the FID score tends
to be higher for higher-ranking pages.

76
 Ranking Factors Study 2024

FID [field]

It’s worth noting that Interaction to Next Paint (INP) will replace FID as a Core Web Vital in March
2024.

LCP (Largest Contentful Paint)

Largest Contentful Paint (LCP) is a metric that measures the time between a user starting to load
a page and its main content being rendered. A good LCP score is considered to be 2.5 seconds
or less.
The top 10 ranking pages have an average LCP score of 2.68 seconds, and they tend to have
better LCP scores overall compared to lower-ranking pages (the average for the pages
in positions 10-20 is 2.72 seconds).

77
 Ranking Factors Study 2024

LCP [field]

Lab Metrics
We looked at the following lab metrics:
• Total Blocking Time (TBT)
• Time To Interactive (TTI)
• First Contentful Paint (FCP)
• Largest Contentful Paint (LCP)
• Cumulative Layout Shift (CLS)
• Server Response Time

78
 Ranking Factors Study 2024

Total Blocking Time (TBT)

Total Blocking Time shows the total amount of time that a page is blocked from responding
to user input (e.g., mouse clicks or keyboard presses). A good TBT score is 200 milliseconds
or less.
Our data shows that the TBT average among top 20 positions is 20+ times higher than
the recommended score. And it is higher for higher-ranking pages!

Total Blocking Time [lab]

Time to Interactive (TTI)

Time to Interactive is a metric that measures the time it takes between a page starting to load
and it becoming fully interactive. A good TTI score is considered to be between 0 and 3.8
seconds.
According to our data, higher-ranking pages have a higher TTI score.

79
 Ranking Factors Study 2024

Time to Interactive [lab]

FCP (Lab)
We also tested First Contentful Paint in a controlled environment. Again, it showed that the FCP
score is lower for higher-ranking pages, with the average value (2.93s) being a bit higher compared
to FCP scores in the field (2.02s). This factor presented no correlation to higher rankings.

FCP [lab]

80
 Ranking Factors Study 2024

LCP (Lab)
Largest Contentful Paint scores tested in a controlled environment are lower for higher positions,
but the average value (7.18s) is quite a bit higher compared to LCP scores in the field (2.68s).
No correlation was found between this factor and higher rankings.

LCP [lab]

CLS (Lab)
Lab-tested Cumulative Layout Shift scores are lower for higher-ranking pages, with the average
value (0.08) being lower compared to field-test CLS scores (0.097).

81
 Ranking Factors Study 2024

CLS [lab]

Server Response Time (Lab)

Server Response Time (also known as Time To First Byte) is the time it takes for a user’s browser
to receive the first byte of page content from the page’s server. A good server response time is
800 milliseconds or less.
Top-ranking pages have a lower server response time compared to lower-ranking pages,
with the average server response time for pages in the top 10 positions being 0.66 seconds.

82
 Ranking Factors Study 2024

Server Response Time [lab]

Key Takeaways
Across all the factors in this group, we found minimal or no correlation to higher rankings.
However, each of them separately and all of them collectively can impact user experience on your
site. This, in turn, can shape users’ decisions like whether to stay and interact or bounce off.
The key takeaway here is that you should monitor your performance for these metrics not in order
to hit an “ideal” score, but as an indication of how technically sound the user experience you offer
is.

83
 Ranking Factors Study 2024

User Signals
We used data from Semrush .Trends and the Semrush Organic Research tool to analyze
the following metrics (at domain level):
• Direct traffic share: The percentage of direct traffic a domain gets.
• Time on site: The average time users spend on a domain/website.
• Branded search traffic share: The percentage of branded search traffic a domain gets.

• Bounce rate: The percentage of users that leave the website after visiting only one page.
We didn’t find any strong correlation between these metrics and SERP positions.

User Signals Correlation with the Position on SERP

Top-20 Results

Whether Google uses behavioral metrics, including clicks on search results, as a ranking factor
has been a topic of debate for years. During the recent hearings amid the Google search antitrust
trial, it was revealed that Google does indeed use user clicks, among other user behavioral data,
to determine rankings-even if not as a direct ranking factor.
Below, we have included our findings about the user signals we investigated.

84
 Ranking Factors Study 2024

Direct Traffic Share

Top-10 ranking pages in search results tend to be on domains with a higher share of direct traffic.
The average direct traffic share for these domains is 31%.
Conversely, the average for domains of pages ranking in positions 10-20 is 29%.
The degree of correlation is quite low, though, as it stands at o.09.
Attempting to draw a clear line between higher rankings and direct traffic share would be
a mistake, especially bearing in mind the blind spots of different attribution models or cookie
policies.

Direct Traffic Share

Branded Search Traffic Share

We also looked at branded search traffic share on domain level. This refers to the percentage
of traffic coming from searches that include a branded keyword.
Here, we found that not only was there no correlation between branded search traffic share
and rankings, but also this: the share of branded traffic was at its highest for positions 4 and 5.
For positions 1 and 2, this figure was lower than for positions 3, 4, 5, and 6.
85
 Ranking Factors Study 2024

We found similar “atypical” behaviors for domains of pages ranking on the second page
of Google. So, positions 14, 17, and 19 had a higher share of branded search traffic compared
to the top positions on page 2 of SERPs (positions 11, 12, and 13).

Branded Search Traffic Share

Time on Site
Our data shows that users tend to spend more time on websites with high-ranking pages: around
8% more on the page ranking #1 vs. on the page ranking #10.
Keep in mind that Google has repeatedly claimed it doesn’t use time on page, dwell time, or other
similar metrics as part of its ranking systems (here and here).
But, since Google tries to rank high-quality pages that address search intent at the top, it makes
sense that users are spending more time on these pages because they find them useful.

86
 Ranking Factors Study 2024

Time on Site

Bounce Rate
Bounce rate refers to the percentage of visitors that leave your website after visiting just one
page. These visitors do not take any action, such as visiting other pages on your website, clicking
a link, signing up for an account, or making a purchase.
In the SEO community, it’s been speculated for years that Google uses bounce rate as a ranking
factor, even though Google has denied this is the case. In fact, Google’s Gary Illyes had stated
that Google doesn’t use bounce rate as a ranking factor as far back as 2015.
Our data showed a minor correlation between site-wide bounce rate and rankings, meaning that
top-ranking pages tend to have lower bounce rates.
The average bounce rate for the domains of top-10 ranking pages is 60%.

87
 Ranking Factors Study 2024

Bounce Rate

Key Takeaways
Build your brand
Almost 1 out of 3 visitors to domains of top-ranking pages goes to these websites directly. And 1
out of 4 does so by typing a brand query in Search. This demonstrates the importance of raising
awareness around and establishing your brand.

Address user intent

Capturing user intent, covering topics in a comprehensive way, and offering a stellar experience
to users on your website should form a priority in your strategy.

Monitor bounce rate on your pages

What is an “optimal” bounce rate varies from one industry to another or for different page types.
Monitor the bounce rate on your site pages to spot abnormally high rates: a high bounce rate
can indicate issues in areas of your website that could ultimately impact revenue and rankings
negatively.
These can include a slow website, excessive pop-ups, unhelpful content, or an overall poor user
experience, all of which can make users leave your website without taking any action.

88
 Ranking Factors Study 2024

Key Learnings
Here is a selection of our key learnings from the study:

Text Relevance is the Top-Ranking Factor

The data clearly shows that text relevance (the content of the page covering relevant queries
and/or subtopics to the main query/topic) has the strongest correlation to high SERP positions.
This indicates that the content Google surfaces in the top positions is doing a better job than
content ranked lower when it comes to answering the main user intent behind a search query
in a comprehensive way, covering relevant subtopics.
It’s of paramount importance that you consider and address user intent with your content every
single time.

URLs Ranking at the Top for One Term Tend to Rank High for Many More Terms
We found that pages ranking at the top for one keyword tended to rank for four times as many
terms than pages in position 20.
This may indicate these top-ranking pages do a good job of covering the main topic in great
breadth and depth. By doing so, they often naturally cover relevant topics and queries.
This can also be partly attributed to acquiring higher authority levels by ranking at the top.
The key takeaway is that it’s worth auditing your content on a regular basis. Discover low-hanging
fruit-pages ranking in positions 2-20. Then improve them to try and push them to higher
positions on the SERPs.
Here’s how to do it using Semrush’s Organic Research tool:
Type in your domain and hit the “Search” button.

89
 Ranking Factors Study 2024

Next, click on the “Positions” tab.

Here, you’ll be able to see all the organic search positions for your domain. Click the “Positions”
drop-down and use the “Custom range” option to filter the list to only show keywords where your
domain is in positions 2 to 20. Hit “Apply.”

You’ll then get a list of all the keywords you’re ranking in positions 2-20 for.
For these keywords, Google probably already sees you as a relevant and trustworthy source, but
your content needs improvements. Update and upgrade your content.
Ranking #1 means 1.5 more traffic than ranking #2.

90
 Ranking Factors Study 2024

Your Reputation and Customer Experience Matter

There’s a strong correlation between the number of star ratings in a domain’s pages’ organic
listings and rankings. While this is dependent on the type of query, it’s still crucial that you pay
attention to your business’s reputation.
You should encourage customers to leave reviews and always respond to the reviews you get. If
you’re a local business, get your Google Business profile in order and tackle reviews there.
For Google, trust is the central piece in E-E-A-T, so it’s super important to strengthen
trust for your website at all times. One of the many ways to do this is by featuring reviews
and testimonials on your website.

Authority is Key in Higher Rankings

We found a strong correlation between higher Page Authority Score (PAS) and Domain Authority
Score (DAS) and higher rankings (0.19 and 0.21 respectively).
Building a strong backlink profile with links from authoritative and relevant pages and domains is
key to a successful strategy.
Yes-backlinks still matter, both as a site-wide and a URL signal.
Gary Illyes saying they’re no longer among the top 3 ranking factors at Pubcon resonates-
yet, they’re still important. The web is still based on the idea that good content gets votes
of confidence. And these votes often come in the form of backlinks.
Our data clearly shows that top-ranking pages have a higher number of backlinks compared
to lower-ranking pages. But there’s also a “chicken or the egg” situation here-the higher you rank,
the more backlinks you may acquire.
What this means for you:
• Create content that catches attention and attracts links and mentions.
• Promote your content and amplify it through marketing channels.
• Monitor your progress regularly-Semrush’s Page and Domain Authority Scores are good
metrics to monitor for knowing if you’re on the right track.

91
 Ranking Factors Study 2024

Focus on Creating Quality Content

Ignore talks around content length or word count-we didn’t find any correlation with rankings
when it came to content length. On the other hand, the correlation with content quality was
strong.
Keep in mind that “quality content” is a much wider notion than one or two metrics.
What you should do: Ensure your content is readable and easy to digest-use content formatting,
images, and videos (when these make sense and improve the user experience).
Avoid fluff, complicated phrasing, and syntax, and write naturally. By doing so, you’re also likely
to cover relevant subtopics and queries, which we also found a correlation for with higher
rankings.
Ultimately, “quality” has to do with bringing substantial value to your readers. Review your content
with this as your North Star.

Pay Attention to Technical Optimization and User Experience

Yes, correlation for Core Web Vitals, HTTPS, and more technical elements in the study was lower
than for other factors.
All things being equal, it’s been argued that these signals act as a tie-breaker.
Even if this isn’t the case, all these factors contribute to an overall better user experience on your
pages. They can influence user decisions to purchase or not, to stay or bounce back.
So, it’s important to maintain a good technical site health. You can use Semrush’s Site Audit
to regularly audit your website for technical issues and to get actionable tips on how to solve
them (including Core Web Vitals and HTTPS).

Report on Metrics that Matter

Often, metrics like bounce rate or time on site are featured in SEO and content reporting. For
many professionals, these are core metrics; for others, they are regarded as “vanity” metrics.
Our study showed minimal or no correlation between user signals and rankings.
Depending on the focus and goals of your SEO strategy, these may still be things you want
to track to understand how people engage with or react to your content.

92
 Ranking Factors Study 2024

Don’t Go Overboard with Schema

We already saw Google removing the majority of FAQs and How-Tos from search results earlier
this year. Our study also showed very little correlation between Schema and rankings.
Use it when and where it makes sense. Remember, don’t abuse it; it can bring on a penalty if you’re
seen as using it to manipulate rankings.

Build Author Trust Signals

Yes, correlation was low for authorship on the page (and for Person schema).
But an author’s or website’s authority, experience, and expertise are explicitly mentioned
in the Search Quality Evaluator Guidelines as factors affecting content rating as lower or higher.
Note that an “author” doesn’t have to be one person. It can also be scientific teams, multiple
authors, company staff, or editors.
Here’s what you can look into to improve your content’s E-E-A-T:
• Audit your content for trust signals in the context of authorship.
• Attribute your content to authors with authority and reputation in the field (where relevant).
• Invite influencers and experts to participate.
• Leverage schema to make it easier for Google to connect the dots.

93
 Ranking Factors Study 2024

Methodology
For this study, we aimed to get a sample of ~16,000 keywords of three out of four main intents.
From our US database we took a sample of 150,000 keywords with a volume of 100+ monthly
searches.
We cleaned it up to get the resulting sample size of 16,298 keywords. We excluded the following
keywords:
• Keywords in languages other than English.
• Keywords with navigational intent: Queries indicating a strong intent toward discovering
content on specific pages or domains only.
• Branded keywords: For these types of keywords, brand and brand/product-related signals
would get priority, meaning a more targeted analysis would be required.
• Adult keywords.

• Misspelled, mistaken, or “duplicated” keywords.

For each keyword, we collected the mobile SERPs (US). We analyzed the top 20 positions from
each SERP (both SERP features and organic links).
We excluded positions featuring Wikipedia and YouTube to avoid skewing the results. After doing
this, we ended up with 300,000 organic positions.
For the study, we focused on calculating three main metrics:
• Factor Influence: A Spearman correlation metric between a position and a factor’s average
value per position. This helps us assess how strong the connection is between the value
of a factor (e.g., the number of backlinks) and position on the SERP. We use an “inverse”
correlation (add minus to each resulting coefficient). We do it in order to show a more natural
direction: e.g., if our value increases with the decrease of position number, it is a positive
correlation for us (the larger the value, the better the position).
• Factor Strength: Shows how much the value varies for the first 3 positions vs. the last 5.
We compare averages to see how large the difference is (as a percentage). The higher
the percentage, the greater the difference between the factor values for top and bottom
positions within the range of organic positions we analyzed, and vice versa-the lower
the percentage, the smaller this difference is (so you can assume that top and bottom
positions have more or less the same values of the factor).
• Benchmarking: We calculate and show how the average and median values of a factor look.
94
 Ranking Factors Study 2023

Limitations
We’ve provided details about our methodology; now it’s time to acknowledge some of the core
limitations facing a study of this nature:
• US mobile data: For our study, we focused on US mobile data only. We understand factors like
language, culture, and user behavior differ across markets and can have an impact on SERPs
(along with Google’s capabilities in different markets). We chose the US as it is the market
Google receives most of its monthly traffc from (+18B monthly traffic, 18% global traffic share).
• Impossible to confirm and analyze all potential factors: Different ranking factors come
at play at different stages in the process (as confirmed recently during the Google search
antitrust trial). With no way to positively identify and analyze all of Google’s ranking factors, we
focused on factors included in Google’s documentation, Google’s spokespeople statements,
and/or longstanding conversations in the SEO industry. From those, we kept the ones that was
feasible to quantify and collect meaningful and reliable data for.
• Personalized and query-type results: Our study aims to look at the full picture.
Personalization alongside factor-weighting and ranking systems coming into play (depending
on the query type) means our study can’t offer deeper insights on a query-type basis.
• Chicken or the egg: Ranking higher in the SERPs can result in increased organic traffic, number
of backlinks, and more. We understand a number of factors we’ve explored in the study are often also
“inflated” as a result of ranking higher. Our study focuses on exploring the correlation between each
factor and organic positions, not in analyzing each factor as a potential outcome of each position.
• User behavioral data: Whether or to what extent behavioral data (such as user clicks) are
used by Google for ranking purposes has long been a point of debate between Google
and SEOs. Recent witness testimonies and documents surfacing during the Google search
antitrust trial confirm that user signals have been used in rankings. Our data was collected
before this evidence came to light; at the time, we chose to investigate some core metrics
typically used to assess user engagement on a page.
• Search Generative Experience (SGE): SGE is not part of our study. When the data for our
study was collected and processed, SGE was still in test mode for only a small group of people.
Including a small-scale, early-stage live experiment in our data would skew our results.

95
 Ranking Factors Study 2023

Benchmarks
We used benchmarking for all the metrics analyzed in our study. These were based on the top 10
and the top 20 positions. You can find all benchmarks in the Appendix.
Below, you can find benchmarks for all the factors and metrics analyzed in the study.

average average median median

factor_name factor group
top-10 11-20 top-10 11-20

Page authority score Backlinks Factors 20 14 18 8

Domain authority score Backlinks Factors 67 57 72 58

Number of backlinks (URL) Backlinks Factors 2,418 1,628 13 4

Number of backlinks
Backlinks Factors 2,335,555,799 1,923,408,023 9,683,696 1,992,337
(domain)
Number of referring
Backlinks Factors 112 68 6 2
domains (URL)
Number of referring
Backlinks Factors 1,083,945 880,086 58,131 22,329
domains (domain)
Number of referring IPs
Backlinks Factors 114 72 7 2
(URL)
Number of referring IPs
Backlinks Factors 196,113 148,056 37,438 16,239
(domain)

Content age (in days) Content Factors 714 769 378 400

Text relevance Content Factors 91% 81% 94% 85%

Readability score Content Factors 51 52 58 58

Percentage of pages with

Content Factors 5.2% 4.6% - -
author Schema markup
Percentage of pages with
Content Factors 19.9% 18.0% - -
Schema.org
Percentage of pages with
Content Factors 17.1% 18.5% - -
tables

Images alt tags count Content Factors 31 29 17 15

Images count Content Factors 34 31 18 16

Keyword coverage Content Factors 67.8% 64.1% 75.1% 69.3%

Lists count (<ul> tags) Content Factors 18 18 12 12

96
 Ranking Factors Study 2024

Meta description word

Content Factors 23 22 23 22
count

Paragraphs count Content Factors 36 34 18 17

Schema.org count Content Factors 0.2 0.2 0.0 0.0

Sentence count Content Factors 53 51 31 30

Tables count Content Factors 1.5 1.5 0.0 0.0

Title word count Content Factors 8 9 8 8

Content quality score Content Factors 77% 72% 85% 78%

Video count on page Content Factors 0.01 0.01 0.00 0.00

Page word count Content Factors 1,451 1,378 1,069 1,005

Number of words from a

On-SERP Factors 2.5 2.4 2.0 2.0
keyword in title
Percentage of pages with
On-SERP Factors 16.9% 15.9% - -
keyword in description
Percentage of pages with
On-SERP Factors 21.7% 19.4% - -
keyword in title
Number of words in
On-SERP Factors 25 25 25 25
description

Number of words in title On-SERP Factors 7 8 8 8

Title-keyword similarity
On-SERP Factors 0.69 0.67 0.70 0.68
[BERT]
URL-keyword similarity
On-SERP Factors 0.67 0.64 0.69 0.66
[BERT]
URL & Domain
Domain's age 20 18 22 20
Factors
URL & Domain
https 98% 98% - -
Factors
URL & Domain
Is subdomain 11.4% 10.8% - -
Factors
URL's organic positions in URL & Domain
76 46 16 7
top-20 Factors
Domain's organic positions URL & Domain
1,436,906 806,297 95,346 21,354
in top-20 Factors
URL & Domain
URL's organic traffic 10,678 5,439 735 151
Factors
URL & Domain
Domain's organic traffic 398,282,102 295,685,389 214,715,033 214,715,033
Factors
URL & Domain
If nTLD 97.2% 95.4% - -
Factors

97
 Ranking Factors Study 2024

URL occurrences in URL & Domain

13 9 3 2
Featured Snippets Factors
Domain occurrences in URL & Domain
5,668 3,767 506 184
Featured Snippets Factors
URL & Domain
URL occurrences in Images 44 33 13 8
Factors
Domain occurrences in URL & Domain
817,526 410,354 31,176 6,389
Images Factors
URL occurrences of Star URL & Domain
38 26 10 6
Ratings Factors
Domain occurrences of URL & Domain
743,089 401,226 34,195 3,456
Star Ratings Factors
URL & Domain
URL length characters 65 68 60 63
Factors
URL & Domain
URL length subfolders 2 2 2 2
Factors
URL & Domain
URL parameters count 0.11 0.09 0.00 0.00
Factors
User Experience
CLS [field] 0.097 0.103 0.04 0.04
Factors
User Experience
FCP [field] 2,023 2,054 1,844 1,849
Factors
User Experience
FID [field] 40 35 22 20
Factors
User Experience
LCP [field] 2.68 2.72 2.42 2.44
Factors
User Experience
CLS [lab] 0.083 0.097 0.007 0.012
Factors
User Experience
FCP [lab] 2.93 3.02 2.61 2.65
Factors
User Experience
LCP [lab] 7.18 7.47 4.96 5.36
Factors
User Experience
Server response time [lab] 663 727 435 474
Factors
User Experience
Time to interactive [lab] 19.14 17.88 15.42 14.18
Factors
User Experience
Total blocking time [lab] 4.71 4.09 2.43 1.88
Factors
Branded search traffic
User Signals 23.8% 23.4% 28.5% 28.5%
share (domain-based)
Bounce rate
User Signals 60.5% 61.9% 59.5% 62.0%
(domain-based)
Time on site
User Signals 0:09:42 0:09:06 0:08:34 0:07:44
(domain-based)
Direct traffic share
User Signals 31.2% 29.4% 29.1% 26.4%
(domain-based)
98