Conversion Rate Optimization & Usability Testing

Are you looking to get more business from your website? Are you frustrated with the amount of money you’d spent on a beautiful design, but are worried that people are not turning into paying clients? Do you have a webpage that gets a lot of traffic, but doesn’t get customers to fill out forms or doesn’t encourage users to click to learn more about your company?

ASK YOURSELF THIS: Would I like to double the number of sales calls from my website without spending twice as much on advertising? If the answer is “YES!”, then read on.

If you’re experiencing any of the issues listed above then, you may be interested in conversion rate optimization. Also, known as split testing, a/b testing, bucket testing, split-run testing, or experience experiments; conversion rate optimization is one of my favorite tactics for improving a web page’s performance. In the following paragraphs, I’ll show you everything you need to know to get started turning your webpages into valuable business assets.

What is conversion rate optimization?

In a nutshell, conversion rate optimization is the practice of experimenting with a page’s design so you can find out which design choices get a user to take the desired action. As an internet marketer, I often use conversion rate optimization as a way to get around design arguments, to justify gut feelings/guesses, to know what words will get people to take action, or to find out if what a competitor is doing is actually working.

Regardless of why you’d want to get started, conversion rate optimization is simply a testing/experiment process that allows you to make design decisions that actually impact your business’s revenue. These experiments allow you to get real information about how a page’s design can positively or even negatively impact your website’s ability to turn website visitors into customers. Both positive and negative changes are very important to take note of. Ultimately the reports from your split tests will take away subjective beliefs that may be harming your business. If you pay attention to the reports from a split test you will be given a clear choice of which page design causes users to take action!

Why use conversion rate optimization?

I’d like to take a moment to advocate for your audience. From the perspective of your website visitor, if they cannot find what they are looking for on your site, then they may assume that if your website is hard to work with, then you are too.

Instead of risking the loss of sales, you can run experiments to see what will make the site easy to use. These experiments should be geared towards what makes your audience want to take the next step. The next step should be defined by the single purpose of a particular page. For example, an “About Us” page should build trust and encourage users to send a message or learn more about your services. If you don’t see users taking those actions often enough, then you will want to use conversion rate optimization.

Another example is your service page. This page’s job is to get people to request more information. If people are leaving the page without calling or without contacting your company, then it’s time to think about what may be blocking them from taking action. Are your contact details hidden? Is there something on that page that is offputting? Maybe you would get more leads if there was a big button sending people to your contact page or even a contact form directly on that page.. In either case, you don’t know. You need to split test to get the right answer.

The unexpected benefits of running split tests

As you test pages and get positive results you will actually begin to rank better in search engines. Why? This is because you are signaling to Google that you have a shared concern — giving website visitors a great user experience.

In a nutshell, the more people who spend time on your pages, click to other pages, fill out your forms, (etc.) the better Google will rank you. This is because it is a sign that your website is actually useful. Additionally, as your website ranks higher you will get more customers. The more customers you get, the higher your revenue. The higher your revenue the more you can market and advertise — which will, again, bring more revenue. We call that business growth!

What is important to measure: Conversions, Time one page, Clicks, bounce rate, video watched, etc…

If you don’t know what success looks like, then you’re just throwing money away. For conversion rate optimization experiments you have a few standard options which will influence the revenue. Your goal needs to align with which strengths and weaknesses you see in your data.

If you want people to spend more time reading your blog posts or watching your videos, then you’ll want to focus on the time-on-page and/or how far down the page someone gets. If you’re more concerned about making sure people who read your blog articles are moving to sales pages, then you’ll want to focus on “bounce rates”; which measures how many people look at just one page on your website, then leave. Both of those success indicators will eventually impact revenue, but if you have the opportunity to test how often someone actually reaches out to hear more about your offering or to reach out to become a customer, then that is where the real money is.

If you have a landing page or a sales page that has a ton of traffic, but not a lot of conversions (or a high bounce rate), then you really want to focus your split test attention there. Why? The reason you should focus on the sales and landing pages if they get a lot of traffic to them is that those people have landed on those pages and they are either making a decision about who to buy from or are ready to make a purchase now. If your page is missing something or has way too much information on it, then you can get real information about what changes you should make to your webpage.

Ask yourself this: If my sales page gets 5 phone calls for every 200 people that land on it, then how much would it be worth it to double the number of phone calls we get without paying twice as much in advertising?

What results can I get from a split test experiment?

Some tools will claim to increase conversion rates by 400% or more, but you really need to think that number through a little bit. More importantly, it is not the tool that will get those results. You get results by how you plan to use your chosen tool. They probably ran hundreds of tests before reaching a 400% increase. Some of their split tests possibly had a negative effect; which is also good because you can find out exactly what does not work for your audience. Other than an increase in conversions you may also see a lower cost per ad, increased awareness on social media, and more attention from your audience while they are on your website.

An example of a conversion rate optimization experiment

Let’s say you have a webpage with a “request a quote” form. You’d like people to give you information about their needs so you can create an estimate. On the page, you have a form and a paragraph of information an pricing details. As an experiment, you’d like to see if you could get more people to fill out the form if you remove the pricing information from the page.

Here’s an example of a conversion rate optimization test

You would create two versions of the page, one with pricing and one without, and use a tool (like Google Optimize) to send 50% of your users to the original page and 50% of your users to the page without pricing. After getting a lot of visits to the page (at least 500 each), then your conversion rate optimization tool will give definitive evidence as to which version of the page you should keep using. Simply put, if one design caused more people to fill out the form, then use that design.

How much does conversion rate optimization and split testing cost?

It is not as expensive as you’d think; especially when you factor in the additional revenue this will bring. With that said, each split test experiment is different and there are several factors that influence the price. For example, if your website doesn’t have Google Analytics installed yet, then that would need to be set up; which could cost you.

Another factor that could influence the price is the number of visitors your website receives. If it is too low or will take too long, then you may want to invest in SEO or PPC so that you can bring traffic to your website and be able to make real decisions. If you’d like a quick quote, then contact MediumInteractive.com for and we’ll let give you an estimate for your split test.

Why conversion rate optimization works?

This practice replaces gut feeling with real information about your audience. This is a hard pill to swallow so I’ll say it again, conversion rate optimization replaces GUT FEELING with REAL INFORMATION. Think about it like this, you may know your audience. You may have sat in a room with them. You may have paid a high-dollar marketing agency to generate demographic reports, but… when they are alone and looking at your webpage; none of that matters. Your people have experiences that cannot be captured until they are in the moment.

Who uses conversion rate optimization?

Microsoft, ebay, GAP, FOX, and Verizon all use split testing to figure out which version of a page will get people to download more – and that’s just a short list. Sure, big companies do test out which button will get you to click and tech companies will naturally use split test technology, but this is NOT only for big tech companies. This is NOT just for a specific industry either. Anyone with a website can do it. Conversion rate testing can be extremely useful for a small or growing business.

Try this: Start taking a look at the URL as you visit pages. If you notice unusual letters or numbers after the “/pagename”, then it is likely a split test. For example, mediuminteractive.com/services is different than mediuminteractive.com/services-b . To find out if an experiment is actually running, try to remove the “-b” from the URL, then reload the page. If you see a different result, then it is clearly a split test. (IMPORTANT NOTE: This was just an example. mediuminteractive.com/services-b does not actually exist.)

How to know what page I should run a split test on?

Many people will start experimenting with the homepage; which is not a bad place to start because a homepage typically does get a lot of traffic, but it would be unwise to neglect other pages of your website. If you’re serious about making your website a true business asset, then you will have more pages than just your website.

Other than the homepage, you will with to find other, high-traffic pages to optimize. If you know one of those webpages should be encouraging users to take a certain action (filling out a form, clicking a specific button, etc.) and you have the data to support that the same webpage does receive a lot of visitors, then that is the page you should optimize. If you have a lot of pages like this, then it may be difficult to choose which page to run the conversion rate experiment on. To make it as simple as possible, do not that A) you are able to run multiple experiments concurrently and B) if you only want to optimize a single page, then look in Google Analytics for a page that gets a lot of traffic and has a high bounce rate and/or a low conversion rate. This will be your key indicator for what page you should focus on.

Can you find out if your competition is running split tests?

The hard thing is that unless you are sitting in the marketing meetings with your competition you can’t easily find out what your competition is up to. Fortunately, there are paid tools, like Ispionage out there which will track landing page changes that could be a signal that their team is running experience experiments. Let’s be real about this. Would you rather be looking for competition’s tests, never knowing what their actual results were, or would you rather fix your own website? The answer is clear. If you waste time watching your competition that closely, then you will not have time to invest in your own innovation. I’d say you should worry more about improving your performance so you grow more quickly.

How long should my conversion rate optimization run before I can get actionable information?

The length of time your conversion rate optimization experiment should run for is based on two factors — how many website visitors does your webpage (single page) receive each month and how many different styles of the page you’d created. In the end, you want at least 500 visitors to each design variation you are testing.

Armed with that information you’re able to manage expectations by doing simple math. For example, if the webpage you are testing typically receives 1000 visitors each month and you are testing the current version of the webpage against one other design option, then it should take one month to receive 500 visits to each of those versions of the page. If you were to add 2 versions (4 in total) of the page and if your webpage has about 1000 visits per month, then you’d want to run that conversion rate optimization test for 2 months. Basically, in the end, you want to have at least 500 visits to each version of the page you’re testing. If you want to speed up the process, then Google Ads may be one option that will bring you a lot of traffic quickly.

How to set up a conversion rate optimization?

  1. Ensure you have google analytics (with goals)
  2. Find a page that gets a lot of traffic, but not a lot of conversions (or a blog post with high bounce rates)
  3. Install the Google Optimize tracking code
  4. Create a hypothesis (Ex. by adding 3 buttons to the page, then it will increase the number of conversions, or changing the background color from white to bright lime green will decrease time on page).
  5. Define your key performance metrics in Google Optimize
  6. Create multiple versions of the page within Google Optimize or on your website
  7. Run your experiment until there are at least 500 visits to each version of your page

Tools to use

Google Optimize: Free
Alternatives to Google Optimize: SEE ALTERNATIVES

How to find the right conversion rate optimization specialist

A conversion rate optimization specialist must have a certain set of skills. These skills will either make or break their ability to produce results that positively impact your business’s bottom line. First of all your conversion rate optimization specialist must have an unbiased mind. They need to think like a scientist. If they are only running tests to prove their point of view is right, then they will look for ways to skew the results in their favor. Though that may seem unethical, some people do act on a subconscious level; which will be harmful unless your conversion rate optimization specialist has the mindset that it is OK to be proven wrong when it benefits your business.

Secondly, your conversion rate optimization specialist should be hired because they respect hard numbers. If you are using a tool like Google Optimize, then your experiments will produce actionable reports. If your specialist understands the data and makes changes according to what will bring you more business, then they are the right person for the job.

Lastly, your conversion rate optimization specialist should understand your business’s goals. This is not less important than the other aspects, but certainly need to be considered when hiring a conversion rate optimization specialist. They should keep in mind whether or not you care more about getting phone calls versus getting more forms filled out. They should keep in mind if how much time a person spends on the page is important to you or not. They should keep in mind whether or not you simply want people to read more or sign up for your newsletter.

Ultimately, your business will have different needs at different times. Your conversion rate optimization specialist must be aligned with your current business objectives in order to make the biggest impact on your organization.

How many conversion rate optimization experiments should I do?

You should run as many conversion rate optimization experiments as you are able to handle. The trick is that you are able to try different things on the same page. Just because you’d run an experiment on one of your pages and that experiment had ended does not mean you should not try again. In fact, it is completely OK to run the same exact experiment again because you may want to eliminate any false data.

Also, it is completely OK to move on to a different experiment with the same page. What I have found to get the BEST results is starting with drastic design differences, then come back with smaller variations. For example, the first experiment will move entire sections of the page, include more content or videos, change the size of buttons, include more buttons, (etc). After that first experiment has ended and our data tells us which design direction to go in, then we can fine-tune the design with a second, third, or fourth round of experiments.

How drastic should my conversion rate optimization experiments be?

Have you ever heard the expression, “Go big or go home”? That is true with conversion rate optimization experiments.

If you want to see big results, then you must be willing to have your assumptions challenged. You may not personally like one of the design variations. You may feel one of the design variations is not aligned with your brand, but if it drives more action from your potential buyers, then I am assuming that you may be willing to make a few, small concessions for the sake of sales uptick.

Can conversion rate optimization improve SEO?

If done right, then conversion rate optimization experiments certainly will not hurt your SEO rankings. In fact, having a webpage that gets people to click or take action is something that search engines respond well to. The logic is that the more people engage with your webpage(s), the more Google will see that your website should be ranked higher. It is actually more complicated than that, but in a nutshell, yes, split-test experiments that make the user’s experience better will improve the SEO of your website.

What to do when conversion rate optimization results favor a design that looks bad?

Conversion rate optimization ideas

What could cause a Conversion rate optimization to go wrong?

Not enough visitors
Not testing the variations for mobile
Not believing the data
Not having enough variation in the first round of experiments
Not scheduling your test at the right time
Not testing on critical pages
Not testing the language that is on the page

Is conversion rate optimization only for web pages?

No. Conversion rate optimization and split tests are not only for web pages. The reason webpages get more attention for conversion rate optimization tests is because webpages are easy to get data on. With just a tracking code and a tool (like Google Optimize) you can get rapid results and make changes quickly. Not only that, but webpages are easier to change than an item you’ve mailed out.

Regardless of what you’re testing, do keep in mind that you should not spend money on a major promotion without knowing your goals and testing so you capitalize on the maximum number of conversions possible. Here are some other marketing channels you can split test and optimize:

  • Bulk Email Messages
  • Bulk Email Subject Lines
  • Cold Call Scripts
  • Paid Ad Images/Copywriting
  • Billboard Location
  • LinkedIn Messaging
  • Responses to client objections (Face-to-Face Meetings)
  • Your Elevator Pitch At Networking Events