Only 22% of companies are satisfied with their conversion rates.

You can find a lot of different experiments on the internet explaining how it’s possible to boost your conversion rates by changing, say, the color of your CTA.

However, these often quoted tests have one major flaw – they do not apply to every business.

Their results depend on so many factors and situations unique to that particular testing process. In other words, to get the same or similar results, you’d have to replicate the conditions in which the said tests took place.

Naturally, it’s a much better idea to use optimization tests tailored to your needs.

To do that, you need a well-thought-out CRO plan.

What’s Conversion Rate Optimization?

Your new website is polished and packed with useful content, and now you’re waiting for your visitors to convert.

And yet, that rarely happens. Your audience seems to be leaving in a rush, without subscribing to your newsletter, downloading your e-book, or purchasing your product.

The whole process of conversion rate optimization boils down to

  • Identifying the root of the problem
  • Creating a strategy for overcoming it
  • Doing a lot of targeted and planned testing
  • Implementing the things you learned.

But, many marketers simply start randomly fixing different elements of their website because they guessed that something like long forms, complicated checkup process, or cluttered landing pages might be the problem.

They might indeed, but if you don’t have a precise CRO plan, you’re just throwing darts in the dark as best practices, and other people’s metrics can’t be used in a one-size-fits-all manner.

1. Collect Data

The first step of your CRO plan should be avoiding guesswork and collecting different kinds of data.

Only then will you be able to get to the bottom of the problem and identify the obstacles which stand in the way of your conversions.

When it comes to the data itself, make sure to split the data gathering process into three categories:

  • The company.
    As Simon Sinek said, “People don’t buy what you do; they buy how you do it.” This means that you need to unearth the “why” of your business which is deeply buried in its DNA and find out what makes your company different from your competitors and why loyal customers love it. Focus on your goals and test how they can be achieved – investigate micro-conversions and possible patterns in your customers’ purchasing process. Coming up with an effective USP is crucial, especially if you can offer something that your competitors can’t. Talk to your CSRs and find out what the main complaints and objections of your customers are.
  • The website.
    Analyze your sales process and identify what the obstacles might be. Determine drop-off points and use heat maps to get a clear picture of what areas of your website generate the most or least engagement. Also, monitor browser and device usage to spot potential technical issues – if the number of visitors from mobile devices is low, it’s a sign that you should check how your website displays on mobile. Given that nowadays more than     50% of all traffic comes from mobile devices, it’s essential to optimize your site. As your website needs to run like clockwork, only the best-known web design companies should be trusted with the job of building or redesigning it.  
  • The customers.
    Naturally, you need as much direct feedback from your customers as possible. There are different ways of feeling the pulse of your audience and gaining valuable insights into what they like and don’t like about your company, products, and website. Remember that     only one out of 26 unhappy customers complain, while the other 25 simply churn, so you need to be proactive if you want to get some really accurate data.

2. Create a Hypothesis

After careful consideration of all the information you acquired, it’s time to create your hypothesis about the possible issues which are driving your prospects away.

So, if people who come to your site complain that they find your checkout process too complicated and unsafe, then you should obviously shorten it and create some trust signals.

The hypothesis should have the following structure:

If (the variable), then (the result), because (the data-backed reasoning).

If we organize our checkout process into three simple steps, our conversions should increase by 30%, because this percentage of our visitors complained about the currently used model.

The variable refers to a website element that can be modified in order to get the desired outcome.

The result is the outcome of the changes which have been made.

The rationale explains why you expect a certain outcome, based on what you know about your audience and their preferences.

What’s extremely important about the hypothesis is that you should be able to test it and measure your results, and see whether something will improve after you change your variable.

The process mentioned above talks about the “what” part of your hypothesis.

To make sure that your hypothesis can be properly tested, you need to consider the following:

  • Which segments of your audience will you test?
    Demographics isn’t the only way to segment your audience, especially when it comes to optimization. For example, make sure not to include new and returning customers in the same checkout process test. Otherwise, you can’t expect reliable results, because loyal customers apparently don’t have too many objections and including them in the test with one-time customers could be confusing and send you on the wrong track.
  • What pages are you testing?
    When testing multiple pages at the same time, it’s worth noting that you shouldn’t put product pages from different price ranges in the same test. It’s self-explanatory that the buying cycles for them aren’t the same and this fact can affect the results.

3. Run Your Tests

It’s best to use designated tools in order to ensure that this step is properly executed.

Optimizely or Unbounce can lend you a hand and help you with the A/B testing process.

There are a couple of things to bear in mind during this phase:

  • Test your design across different browsers, because otherwise, the design might not display properly.
  • If your traffic is high, it’s essential not to send all your visitors to the pages which are being tested. In other words, if the hypothesis is wrong, then your numbers and conversions might plunge, and you can lose your profit. It’s a good call to perform the test on a 10% sample and prevent a potential disaster.
  • The average order value matters too. Don’t be upset if your conversion rates don’t go up. Instead of that, check the average order value, and if there’s an increase, this means that your hypothesis actually worked.

4. Analyze Your Results

Numbers don’t lie, but you still shouldn’t forget a tiny detail called statistical significance.

In layman’s terms, this concept has the purpose of ensuring that the difference between the two variables isn’t the result of chance.

But you don’t have to be a statistics expert to figure this out, because most A/B testing tools calculate this value automatically.

If your hypothesis is correct, then you should make permanent changes to your website and send all your traffic to the more effective page. Of course, this doesn’t mean that your work is done, as there’s always a way to improve things.

So, test every element that possibly causes friction and fine-tune every single part of your website.  

In case your hypothesis is incorrect, then another round of tests is ahead of you. Learn from your mistakes and use your findings when creating a new batch of tests.

Sometimes you’ll fail a couple of times before you succeed, so persistence is one of the most important ingredients of the entire process.

Creating a CRO plan is a detailed task, and it requires a lot of analyzing and data processing, but this 4-step procedure can give you a sense of direction as to what you should do.