searchmetrics email facebook github gplus instagram linkedin phone rss twitter whatsapp youtube arrow-right chevron-up chevron-down chevron-left chevron-right clock close menu search

How to Conduct SEO Experiments to Optimize Your Website

SEO optimization is like any scientific theory — the more you research, the more you realize how many questions you need to find answers to. And to get down to the bottom of things, they both require some good ol’ experimentation.

In reality, successful SEO specialists are just a different type of scientists. Nobody really knows how Google’s algorithm works, so we’ve all carried out experiments, discussed what’s worked and left the rest to chance.

From my perspective, here are the ingredients that enable SEO success:

Even though these points may sound simple, a truly successful SEO depends on so many things. That’s why testing out new SEO tricks and analyzing them is the key to continuously improving your rankings and getting better results for your business.

But just like scientific theories, you can easily fall into the rabbit hole. So I’m here to guide you on completing an experiment and the different types you might want to try out.

SEO experiments: what they are and why you need to do them

It’s likely that you’ve already been through the never ending montage of SEO blogs and videos on the internet and have seen the same broken record answer:

“It depends.”

Of course, you’re a bit frustrated and that’s because, like I just mentioned, nobody really knows how the algorithm works. All businesses are different, which means something that works for John may not work for Mary.

With that said, there are best practices we’ve all agreed on — just like we all agreed, pre-pandemic, it’s socially proper to shake hands when meeting for the first time. But that doesn’t mean you can’t experiment and give someone a hug, fist bump, or heck, even a kiss on the cheek when you meet for the first time.

So, SEO experiments will help you answer the question, “will this tactic work for my business?” Well the only way to find the answer is to test it, track it, analyze it and revise it. This of course takes time, so aside from implementing the usual SEO suspects, set aside some time and resources to carry these tests out. Then, aim for marginal gains over time instead of testing, getting impatient and giving up.

Ultimately, SEO experiments help you determine which tactics work for your business or website, and which ones don’t so get the most out of your SEO strategy.

Okay, let’s now walk the walk after all this talk.

How to conduct an SEO experiment

SEO Experiment Guide
Step Example
1.     Make an observation. Sharing a blog post on LinkedIn gave a boost to your website traffic.
2.     Ask a question. Will regularly sharing a blog post on LinkedIn increase your website traffic?
3.     Form a hypothesis or testable explanation. Regularly posting on social improves SEO.
4.     Make a prediction based on the hypothesis. Sharing a blog post on LinkedIn 3x a week for 3 months will increase your website traffic by 150%.
5.     Test the prediction. Commit to a posting and social sharing schedule and track your website traffic.
6.     Iterate and use the results to make new predictions. Post on LinkedIn more frequently. Post on other relevant social media channels. Test which types of content get the most engagement.


Alright, put your white coat and lab goggles on. It’s time to teleport to your 6th-grade science class.

Your science teacher, Mrs. Maury, mentioned a certain rhythm to experiments — a method if you will. This scientific method goes as follows:

  1. Make an observation.
  2. Ask a question.
  3. Form a hypothesis or testable explanation and make a prediction based on the hypothesis.
  4. Test the prediction.
  5. Iterate and use the results to make new hypotheses or predictions (for SEO experiments, use the results to improve your overall SEO strategy).

Now that you’ve got the general idea, let’s go over an example.

Use curiosity to build a foundation for questions

To start, let’s follow our curiosity. Say, after not posting anything on LinkedIn for months, you published a new blog post. And lo and behold, your traffic went through the roof. This is an intriguing observation that poses the question: What if you committed to posting 3x a week for 3 months on LinkedIn and shared a new or old blog post?

Form your hypothesis and related prediction

Now we have our question and we need to form a testable explanation. It’s quite possible your blog post rankings weren’t as high in Google because all social channels were dormant. If your SEO can be improved by simply posting one blog post, what would happen if you continuously made the commitment to post 3x a week for 3 months? Would organic traffic from those posts triple? We have our hypothesis and prediction, now it’s time to test it out.

Test it out

Keep track of your website traffic and rankings via Google Sheets after each LinkedIn posting until you’ve posted repeatedly for a set amount of time; and then analyze the data. If you notice your overall traffic and rankings have increased, then apply this concept to all your content. If it didn’t work, you might try testing this same experiment on a different social channel.

A particularly useful type of experiment: A/B testing.

Now flip your notebooks to a clean page for lesson two.

A/B testing

Not just for SEO but for all of marketing, A/B testing is a great way to run experiments and measure results using two different scenarios. It works best when you’re testing out small changes, like a tweak in your headline copy, for example.

How would adding emotional words (like “easy,” “proven,” and “powerful”) to your headline increase your click-through rate? Using A/B testing tools, you can find out whether page A with the old headline or page B with the new headline would generate more traffic. Once there’s a clear winner, you can then implement the winning headline.

Keep on testing

A successful scientist always records observations. So be sure to fill up your spreadsheets with notes and continue to reassess. And don’t shy away from testing out the same experiment six months from now. Maybe it’s a train wreck of an experiment now, but it’s quite possible your theory would render dynamite results later on down the line.

3 Types of SEO experiments to run on your website

1. SEO experiments on your website design

Imagine hours spent tinkering with font sizes and call-to-action (CTA) buttons just to find your site isn’t optimized for mobile. There are a variety of factors that contribute to user experience on your website and consequently play a role in your SEO performance. Because we humans are visual by nature, the visual components of your website make a massive impact. With that said, here are a couple of SEO website design experiments to consider:

  • Font changes: For example, changing to more readable font faces and sizes. The Honey Copy uses big and simple fonts so that users can easily digest the content and therefore stay on the website longer.
  • Color changes: For example, changing the color of link anchor texts. The original color most people associate with a hyperlink is blue, so changing the color to blue or to a different more subtle color could make a difference in your SEO performance. Read more about why hyperlinks are blue here.
  • CTA changes: For example, changing CTA formats. Instead of using text, try using buttons. If you decide to play around with buttons, there are also many routes you can go with that. For instance, making a button look visibly clickable could prompt more people to take action.

SEO Experiments clickable buttons

(Image Source- Just in Mind – learn more about best button practices here)


  • Visual changes: For example, adding images and infographics to your content. To illustrate, Designer Mikiya Kobayashi creates minimal pieces, and his style instantly shines through the moment you enter his website. In another experiment, try adding infographics to your blog posts to create a better experience for your visitors. By being considerate of the different ways your users consume content, you may notice your SEO rankings rise.

Those are just a few options to consider for the visual elements of your website, but now it’s time to move onto copy experiments.

2. SEO experiments on your website copy

You could have the best-designed website in the world, but if the words don’t resonate with your target audience, you’re toast.

Words are one of the most powerful mediums we have as humans. We can use them to express love, hate, fear or to persuade visitors to buy our products. Knowing this, here are some example copy experiments to try out for yourself:

  • Headline changes: Try formulating them as “how to … ”, “why … ”, direct-response and news. Then measure the user’s time on page, conversion rate, etc. To spice up your headlines, try out platforms like Coschedule or Contentrow that help you try out options before executing.
  • Metadata changes: Tweak your SEO title and meta description and measure the click-through rate and bounce rate. Maybe your title is missing a power word or your meta description is too long and doesn’t have a CTA. Craft different versions and do A/B testing to see what works.
  • Format changes: Add more bullet points or shorter paragraphs. It’s no secret that users want answers and they want them quickly. For example, the section you’re reading right now. Using bullet points is a fantastic way to easily break up longer pieces of content.
  • Content changes: Break down your copy into more sections. Who likes to read gigantic chunks of text? Yuck! Use more subheadings for blog posts or divide the text up with images to make it more digestible.

Now that we’ve gone over the experimentation details, it’s time to bring it back to the bigger picture.

3. SEO experiments on your overall strategy

Experiment 1

There are endless examples I could list here. But to make it more concrete, here’s one that I tend to lean on: why its worth targeting keywords with no to low monthly search volume. With this specific strategy, you identify longtail, low search volume keywords that have an easier chance of ranking, but also have high purchase intent (Bottom of Funnel).

This approach is particularly useful if you have a relatively new domain with low authority or in an industry that’s either very competitive or very niche. Let’s consider how a B2B SaaS company selling budgeting software can leverage this approach. Here’s the keyword research for “business budgeting and forecasting software:”

Business Budgeting and Forecasting Software
Keyword Monthly Volume
Mint 118, 505
You need a budget 5,278
Budgeting 3,053
Financial planning 2,572
Project management tools 2,388
Mint budget 1,779
Budgeting tools 798
Budgeting software 704
Adaptive insights 677

This table shows high search volume keywords related to “business budgeting and forecasting software.”

But instead of using these high search volume keywords, we’ll scroll all the way down to find low search volume (but high purchase intent/BoFu) keywords.

Business Budgeting and Forecasting Software
Keyword Monthly Volume
Business budget software 17
Envelope budgeting software 17
Forecast budget 17
Forecast vs budget 17
Free budgeting software 17
Planguru 11
Budget management software 11
Budget allocation software 11
Planning and forecasting software 11
Enterprise budgeting software 11

Low search volume keywords can boost your SEO strategy, especially if you’re working with a low-authority website.

When developing or optimizing your SEO content plan, focus on low search volume keywords and see how it affects your rankings and boosts your traffic.

Experiment 2

Another option is to update your content by adding LSI keywords (Latent Semantic Indexing, otherwise known as keywords that are semantically related to the focus keyword) in your content. You can find them on Google’s suggestions when you type your primary keyword or at the bottom of the SERP. This experiment helps your business be more user-focused and relatable, thus strengthening your SEO strategy.

Here’s an example:

SEO Experiments LSI Keywords 1

SEO Experiments LSI Keywords 2


Add LSI keywords to your content strategy by checking out Google suggestions for your focus keyword.

So maybe your content started out as being about business budgeting, but you get inspiration from LSI keywords and add a few more sections related to that.

Using LSI keywords to plan helps you write more comprehensive and useful content. Not only do you better serve your target audience, but this signals to Google that your page covers the topic in-depth. This helps your page become an authority on “business budgeting,” which brings in more users that might be interested in your software.

Run SEO experiments to continuously improve your ranking and your overall strategy.

There are more types of SEO tests to run (ex. technical SEO tests, etc.), but if you’re in 6th grade, you’re not going to jump to 8th grade (or at least not usually), so the experiments mentioned above are great places to get your feet wet. As you see the results of your first experiments, you’ll spot more observations and hypotheses to test and it will help you become a more analytic and strategic SEO expert.

In SEO and in life, it’s always best practice to take advice from others but then to try out some different tactics for yourself. Think outside the box, and see what works for your unique self. Start with these small baby steps, stick to your experiment consistently, and you’ll be making progress in no time!

Need some help to developing a winning SEO Strategy?

Get in touch with our team of SEO experts and discover they can support you in taking your SEO to the next level!