On Page SEO in 2016: A (2M+ Keyword) Data Driven Analysis

Google is changing their algorithm everyday! There are tons of articles saying on page SEO is dead, or is it?

Google is way smarter than just relying on your title, H1, meta description, content etc to rank your page.

We came across an interesting on page SEO research by AHREFS the other day. The study analysed 2 millions+ results returned by Google.

It is an in-depth article covers what is working and what is not working for on page SEO.

So we’ve put together an infographic for those who do not want to dive into tons of data in the original post.


On Page SEO in 2016: A (2M+ Keyword) Data Driven Analysis 1

On Page SEO Factors

Keyword in Domain Name

Has high correlation for ranking #1, but is predominantly caused by brand name search. Therefore, overall it is a weak signal.

Keyword In URL

No. Consider the use of a descriptive URL in order to help people understand the link before clicking.

Keyword In Title

It depends. As long as you can craft a title that informs the reader of the subject matter, you will rank well. The inclusion of the keywords in the beginning of the title also offers a very small positive correlation.

Keyword In Meta Description

Not at all, sometimes Google will even change your meta description. A total of 50% of websites that ranked well don’t have a target keyword in the meta description.

Keywords In headline (H1)

Again unimportant, because 70% of the 2M results don’t even have H1. However, SEO professionals are encouraged to use keywords in H1 because it provides a better user experience.

In subheading (H2)

Again unimportant. Some 93% of pages that rank in Google’s top 10 don’t have a target keyword in their H2 tag.

In Content

Correlation is near 0. Almost 75% of the pages analysed don’t have even a contain a single mention of an exact match keyword.

In first 100 words

As with keywords in content, not important at all. The focus is on providing answers and value instead of the inclusion of keywords in the first 100 words.

In “alt” tag

It doesn’t help in regular search results. However, it might help in Google image search.

General On Page SEO Factors

Age of the page

Showed the best correlation with ranking. New content takes time to rank well. You will obtain a better result if you update your older content to match searcher’s intent.


Yes, data clearly showed a higher percentage of secure pages in position: 1 – 3 is visibly higher than position 4 – 10

Page load time

Only shows a small correlation with ranking. Merely ensuring your site loads “quickly enough” to keep your visitors happy will be sufficient, cutting a few milliseconds may not help it to rank better.

Length of content

Data shows a high correlation. However, you merely need to make your content as long as necessary. Writing more words for the sake of making it longer is highly inadvisable.

URL Length

URL with fewer folders tend to rank better. Keep the URL short but descriptive to ensure a better user experience, not because it helps you rank.

Linking out to authority sites

Small correlation to ranking. Potentially, linking out will provide a better user experience, and because Google lacks other ranking signals, it will have no choice other than to use this as a signal.

Broken Links

A total of 2% of top 10 results have broken links. However, it is recommended to fix them for a better user experience.

Social Shares

Approximately 70% of pages ranked have zero shares. However, data nonetheless shows social shares correlated with Google rank.

Extra notes:

Relevance beats backlinks

There are cases where 19 referring domains can outrank pages with 1100 referring domains. Relevant content is therefore the key when there are no other articles more relevant to the search query.

Exact match keyword optimisation shows greater success when there’s a lack of content and backlinks

When a page targets long-tail keywords, and there is no other on-page content that Google can read, or there are no other on page factors, then Google can only rank pages based on all information they receive.

Related Posts:

Digital Marketing Scams: 9 Ways To Burn Your Marketing Budget

Digital Marketing Scams: 9 Ways To Burn Your Marketing Budget

“I hired a digital agency to do some marketing online, but it didn’t work.” We get this story a lot, and by probing deeper into what the previous agency did for them, we usually find one of the stories below. This is a highly controversial topic, one that digital...

[Case Study] How I Help To Enroll 114 students in 30 Days

[Case Study] How I Help To Enroll 114 students in 30 Days

I talk to a lot of small & medium business owners over the phone.  Many have the same 3 misconceptions “I need more exposure for my business” “I need a strong brand name to get more customers” “I need a website, it will generate a lot of business for me” This...

See How Other Businesses Use Their Website

See How Other Businesses Use Their Website

“I don’t think my business needs a website at this moment, and anyway, it’s too expensive for my business.” “What do I do with my business website?” We get asked these questions a lot before people engage us for website design services. So we’ve put together an...

How to Generate Life Insurance Leads Without Cold Calling

How to Generate Life Insurance Leads Without Cold Calling

"I just need one more deal to hit my quarterly bonus." This is what my friend told me, he is a full time insurance agent. I'm sure this is something a sales person - especially life insurance agents - can relate to. You're chasing sales targets, appointments, a payout...

5 Little Known Strategies to Use Retargeting The Right Way

5 Little Known Strategies to Use Retargeting The Right Way

Remarketing lists for search ads (RLSA) has been like a buzz phrase within the internet marketing community throughout recent years. Some ad companies like AdRoll and Perfect Audience are promoting the idea that if you use their service they can help by bringing...