In 2017, SEMrush published a study of the 17 most prominent SEO ranking factors. A year later, when Google announced page speed as a ranking factor in mobile search rankings, they released research on website performance. They found 82.89% of websites had issues that negatively affected page speed.
So our team decided to do some research.
Is site speed a prominent SEO ranking factor in 2020 if you want to rank well in desktop and mobile search results?
Are the usual SEO factors - authority and backlinks - still the most important in getting to number one?
Does it matter when you are competing for the local 3-pack, People Also Ask, or traditional search results listings?
To find out, we analyzed the top 10 web pages ranking in mobile search results for some of the most competitive keyword phrases. Using the top website speed tools, SEMrush and SEOquake, here is what we discovered.
The three keyword phrases we chose for this research were from the most expensive industries in paid search: insurance, loans, and law. The keyword phrases we narrowed it down were as follows.
"Cheap car insurance" has an estimated monthly search volume of 246000, keyword difficulty of 82.05%, and an average cost per click on Google Ads of $38.74.
"Business loan" has an estimated monthly search volume of 40500, keyword difficulty of 83.03%, and an average cost per click on Google Ads of $34.54.
"Personal injury attorney" has an estimated monthly search volume of 33100, keyword difficulty of 75.28%, and an average cost per click on Google Ads of $70.57.
Each keyword phrase's search results page has a local 3-pack, People Also Ask, and additional organic results. We focused only on the top ten search results in our analysis. We left out the top news items under the business loan SERPs as those were only relevant to current events and not typical search results.
The Raw Data
To analyze the top search results for each of our chosen keyword phrases, we reviewed metrics from the following tools.
Schema markup using the Structured Data Testing Tool from Google.
Content length using the Word Counter Plus Chrome Extension.
Mobile and desktop performance scores using Google PageSpeed Insights.
PageSpeed and YSlow performance scores using GTMetrix.
Page and domain authority, backlinks, and referring domains along with the overall number of site visits and bounce rate using SEMrush.
The number of pages indexed by Google and the age of the domain according to Archive.org using SEOquake.
Based on the numbers we pulled from the above tools, we found the following.
Does Schema markup play a role in organic rankings?
According to the analysis of our thirty search results, we found that Schema markup was present in all, but one of the local 3-pack and People Also Ask results.
Local businesses use the Organization or LocalBusiness Schema markup. Some correctly used their specific local business type, which is important when telling Google exactly what your organization does.
The People Also Ask web pages used a mixture of WebPage, Article, and FAQPage schema, based on the content. Be sure to look at your competitors in search if you want to rank in the People Also Ask results to see how they code their Schema. Simply grab the URL from search and enter it into the Structured Data Testing Tool.
It will give you the code AND tell you if there are any errors that need to be corrected. Update the code to match your website, organization, local business, and content, and you will be all set!
How much content do you need on your landing pages to rank well in search results?
The answer depends on a variety of factors.
Search results in the local 3-pack had an average of 714 words on their landing pages; this varied by industry, however. The three local results for "business loan" had an average of 289 words. On the other hand, "personal injury attorney" landing pages had an average of 1,272 words.
The average length of search results in the People Also Ask section jumps to 2,560 words. Most pages in these results were blog posts, even if they appeared to be standard or static pages. "Cheap car insurance" content had an average of 3,359 words, and "business loan" content followed closely with an average of 2,900 words.
Finally, the average length of landing pages appearing in standard organic search results was 1,036 words. Landing pages for "business loan" were the longest with an average of 1,496 words. Search results for "business loan" had the most content with an average of 1,995 words, compared to "cheap car insurance" (1,514 words) and "personal injury attorney" (1,285 words).
So while you may have heard there is a "minimum" number of words you need to have on every page, the truth is:
It depends on your industry/niche.
It depends on the keyword phrase you are targeting.
It depends on what type of organic search results you are trying to rank for (local, People Also Ask, standard, etc.).
It depends on what is ranking on the first page right now.
How fresh does your content have to be to make it to the first page of search results? For these competitive search terms, the results whose dates mattered most were the People Also Ask.
The results in the People Also Ask section were either blog posts or web pages with publishing dates and last modified dates. You can check this by right-clicking on any web page you are viewing an looking at the source code.
Local Ratings & Reviews
Does the number of reviews impact the position of a local business in Google search results? Yes, but it is not the entirety of the local ranking factors. If the search user is in a more competitive market, the businesses may need stronger reviews to appear in the local 3 pack.
For our research, this was only true in the case of cheap car insurance, which was extremely competitive in our area.
The same searches for "business loan" and "personal injury attorney" in larger markets like Toronto resulted in local businesses with higher reviews. Theoretically, the smaller your market is (neighborhood, zip code, target local audience, etc.), the fewer ratings and reviews you would need to rank well in the local 3-pack — this is assuming you are located close to your target audience.
Website Speed Tests
Google's industry benchmarks suggest that web pages load in under three seconds, have fewer than 50 requests, and have a page size under 500 KB. Here is what we learned after running the top three website speed testing tools on thirty web pages in search results.
The Fastest Results
The #6 result for "business loan" had a PageSpeed score of 100 (mobile) and 100 (desktop), Pingdom load time of 1.32s and 12 requests, and GTmetrix PageSpeed score of 90% (A) and YSlow score of 76% (C).
The #9 result for "business loan" had a PageSpeed score of 94 (mobile) and 100 (desktop), Pingdom load time of 218ms and 22 requests, and GTmetrix PageSpeed score of 98% (A) and YSlow score of 92% (A).
The #8 result for "personal injury attorney" had a PageSpeed score of 90 (mobile) and 100 (desktop), Pingdom load time of 657ms and 25 requests, and GTmetrix PageSpeed score of 99% (A) and YSlow score of 96% (A).
Although our testing was focused on mobile search results, it is important to note the fastest results were not the first results. The search results for "business loan" included four web pages that took over 3.99s or longer to load in the top five results above the Wikipedia entry that only took 218ms to load.
The Slowest Results
Another insight — web pages in the local 3-pack and organic search results tended to load under two seconds. Web pages in the People Also Ask area took anywhere from 464ms to 7.13s to fully load. We may be able to assume Google will trade page load times for quality content since increased load times coincided with increased word count.
The word count itself had nothing to do with the load times. The increased load times tended to come with media embedding — multiple images, videos loading from YouTube or other sources, and other content. While these make for more engaging content, they also make for higher load times. And while Google says they want faster load times, they don't seem to always penalize for it in mobile search results.
As with most marketing, you will need to look at your competition in search results with each keyword you are targeting to ensure competitive load times. If the top ten landing pages all load under two seconds, yours may not crack the first page at four seconds, even with quality content.
Each website speed testing tool will give you suggestions on how you can improve your web page's load times, page size, and the number of requests to the server.
While you may be focused on your score, focus on the important things: the actual load times, page sizes, and the number of requests. And, of course, the suggestions. The recommendations you get from Google PageSpeed Insights may differ from what you get from Pingdom, as shown below.
If you are not sure how to fix any errors or suggestions listed by Google PageSpeed Insights, Pingdom, or GTmetrix (below), you have a few options.
If you have an in-house web developer, they should be able to help you.
If you have a managed web hosting service, you can contact them to see if they can help you.
You can reach out to a company that has experience with both web development and SEO to solve your performance issues.
The goal isn't to get a 100 or an A, although those scores would be ideal. The goal is to be competitive with the web pages in search results for your target keyword phrases.
Helpful Guide: How to Improve Your Google PageSpeed Insights Score
Traditional SEO Factors
When it comes to ranking for highly competitive keyword phrases, one thing remains clear. You will likely have to have a strong domain authority to get to the first page of search results.
The exceptions are the local 3-pack, top stories, and video sections of organic search. Ranking factors for these sections focus on the user's location (local 3-pack) and recency (top stories and videos).
Of the 21 non-local organic search results, the average Domain Authority was 68. The average number of backlinks was 873 million, from an average of 627 thousand referring domains. They receive an average of 129 million visits per month and have an average bounce rate of 62.48%.
As for Page Authority, the scores are much lower. Thanks to several pages having a score of 0 of not available, the average Page Authority of non-local organic search results was 47. Pages had an average of 2,400 backlinks from 234 domains.
Applying the Study to Your Search Marketing
Instead of a typical "in conclusion," we wanted to tell you how to apply this study to your search marketing.
Are you planning to target a new keyword phrase? Or do you want to improve your current rankings for a keyword phrase your targeting? Start researching the top ten for that phrase using the above tools and metrics. Then compare your web page's metrics.
This should give you some insights into where you need to improve.
Do you need to add the appropriate Schema markup for your page?
Do you need to create some more in-depth content?
Do you need to get more local reviews?
Do you need to improve your website speed?
Do you need to get some more backlinks?
As you start to improve your current keyword rankings and target new ones using this research, you will get a better idea of what it takes to rank. And the higher you move in search results, the more qualified traffic you will begin to receive for your business.