For marketers or business owners who prioritize growth, a recurring thought is:
What can I do to optimize my conversions?
To those not in your world, a 10% increase in conversion might not look like a lot—but for your eCommerce business bringing in $17,000 monthly, that’s an extra $1,700.
And if it were the other way round, you’re at a $1,700 loss.
Ask an average CRO specialist how they’d increase your website’s conversions in the next 6 months.
Their answers will reveal who’s going to waste your time, and who will increase your conversion.
How’s that possible? Let’s start with this.
A quick search on Google on “What is Conversion Rate Optimization?” gives you about 154,000,000 results.
You’d get roughly about a thousand tactics you can use to optimize your conversion in those results.
If the CRO specialist is right off the bat suggesting about 100-200 tactics you should try —you probably shouldn’t listen to that person.
Why?
Well, imagine trying out 50 different “proven CRO tactics.”
Where would you start and how would you implement every one of them?
If it takes you 6 weeks to get quantitative data on a test for each tactic, then 50 tactics give you 300 weeks.
That’s about 5.7 years to try them all out —assuming you don’t give your team a rest.
That means it’s impossible and unscalable for your business—or any other business for that matter.
On the other hand, if a specialist outlines the processes it takes to optimize your conversions, better to stick with them.
It shows they understand that conversion rate optimization is not a tactic, it’s a process.
A process that takes on user experience and engagement as a focal point in determining how a website converts.
In chapter one of this series, you learned about CRO and how you can use it for your business. Now, learn about the steps to optimize your conversions.
The 5 Steps of Conversion Rate Optimization
The latest technique in CRO for 20….
The best trick to optimize your conversion…
The latest growth hack to increase your conversion…
And many more like these are the complete opposite of what conversion rate optimization stands for. The concept behind CRO is more about what factors influence people to take action.
Now, these factors are ever-changing —which means the CRO process is never stagnant.
However, regardless of how things might change later on, the 5 important steps involved in optimizing your conversions remain the same.
1. Data Gathering
The what, who, where, when, and how is what makes up data —and gathering data for conversion rate optimization is not different.
Identifying the areas you need to optimize can be tricky if you’re in a guessing game with your gut.
And no, don’t do that.
Rather, use comprehensive tools that provide actual data to help you paint a picture of where to optimize.
There are many ways to gather enough data to get you started; below are 3 of the best ones I’d recommend.
2. Heatmaps
Heatmaps are your best approach to understanding areas of your website where visitors show more interest.
Hotjar defines a heatmap as a visual representation of user interaction data where values are depicted by a range of colors from the most popular (red) to the least popular (blue) elements of a web page.
Image credit: Hotjar
Using the heatmap, you can easily pinpoint areas with high concentration (red) along with statistical analysis of what percentage of web visitors clicked what —and what percentage didn’t.
The same also goes for areas of low concentration (blue).
3. Analytics
This is one of the most used methods of gathering data for CRO. Analytics takes into account the overall performance of your website, especially in terms of user experience and engagement.
Google Analytics is one of the many tools that does this best.
With Google Analytics, you get insight data that comprises different metrics, such as your overall website performance, bounce rate, click-through rate, sessions, traffic sources, engagement, time-on-page, sales, etc.
4. Customer Survey
Surveying customers lets you understand the experience of customers before and after interacting with a website.
It dives deep into market research and customer experience as a driving factor in collecting quantifiable data.
A tool that does this well is Survey Monkey as it takes out the guesswork for you.
The survey goes in-depth to analyze different reference points across your customers. This helps you prioritize the type of customer data and the best approach.
Once you have these tools, you already have a rough sketch of your customer’s psychology when they interact with you. With that, you can begin segmenting and profiling your data.
Then, you can easily analyze your customers better and understand at what level their preferences influence your conversion process.
This idea you have then allows you to move to the next crucial stage — Hypothesis.
5. Analyze Your Hypothesis
Now that you’ve gathered enough data, interpreting it into a piece of usable information is next on the list.
This is a crucial stage in the conversion optimization process because you’re taking ordinary data and transferring it into information that impacts your decisions.
Your first thought might be to test different methods since you have enough data to back you up.
But all you’ve worked for in gathering that data will be lost if you test the wrong method on your audience.
Take a moment, pause, and ask yourself these three questions:
- Which of our customers are we optimizing for?
- What are we optimizing for?
- Where do we perform this optimization?
Which of our customers are we optimizing for?
Those in the awareness stage? Consideration stage? Or exactly who in your sales funnel do you want to run the test on?
The sales funnel in your business is a drawing board and pillar to all your business end goals. It bridges the gap between your product or services and which customers need them.
This then allows you to map out a customer journey that flows across your conversion process. But a common mistake is not considering the stage your customers are in your sales funnel.
The way you’d convert someone who hasn’t visited your website is way different from the one who already added a product to the cart.
For someone who has never visited your website, your focus is likely moving them down the funnel. Hence, the process will include things like targeted ads and blog posts.
However, for someone who already trusts you and knows they need your product —and already has it in their cart, the process is different. Things like abandoned cart emails and reminders will do the trick.
Consider to whom all your micro conversions will be tailored.
What are we optimizing for?
In the first question, you figured out which of your customers you’re optimizing for.
The next one is “what are you looking to optimize for?” Sales? Downloads? App installs? Email Open Rate?
If in the first question, you figured out you’re optimizing for customers who abandoned their carts —then you’d likely hypothesize optimizing for better open rates for your abandoned-cart emails.
The purpose of this is based on prioritizing what’s most important to your end goal —or a key contributor to your end goal.
Where do we perform this optimization?
You’ve answered the first two questions. The last is “where do you want to optimize in your whole conversion process?”
What landing page do you want to optimize? How many landing pages? If you’re selling products, how many products do you want to optimize for at once on a single landing page?
Or perhaps if it’s a blog post or a Facebook ad — the question is still “where?”
There’s no one-size-fits-all for when you’re hypothesizing this.
However, we recommend you use the data you’ve gathered from the first stage to break down your process.
Let’s say you have a conversion process that goes like this:
Google ad — Landing Page — Demo Booking —Newsletter subscription — Blog Post — Email Drip Sequence — Final Conversion.
At every point of this process shows “where” you can hypothesize for more conversions. But you cannot optimize for two things simultaneously.
You can’t optimize your landing page while also changing the email drip sequence attached to it. The answers you’ll get will conflict and won’t help you.
A better way is to optimize one at a time and use the results to understand how customers go through your whole process.
Create Test Wireframes
Regardless of how much data you’ve collected, you shouldn’t immediately implement your hypothesis.
Wireframes in this context designs what will be incorporated on your website based on the changes you’ve decided to make.
It’s easy to imagine how it’ll play out after implementing one or two changes. But on paper versus on your computer screen, it’s a different ball game.
Your test wireframe gives you an idea of how certain changes will look in the face of your proposed customer.
If you feel it doesn’t look good, or perhaps isn’t user-friendly, perhaps now is a good time to change it.
The essence of wireframes is to mimic the actual test —and even though most people ignore it, it’s a crucial part of understanding how what you put out reflects on your brand as a whole.
Your wireframe should do three things:
- Be a product of your data
- Test out your hypothesis
- Align with the end goal of your sales funnel
Your wireframe aims to ensure you’re creating changes not based on guesses, but actual data.
The sole focus of any conversion is the “end-user” — if it fails in that, it’ll fail in others.
Implement Your Test
After creating a test wireframe and you’re satisfied with it, the next step is to implement everything you’ve learned.
Well, not so fast.
CRO is more of a process, it redefines what it means to run a test. Before you implement your test, there are two things to get right.
- What’s the duration of your test?
- What test method should you use?
What’s the duration of your test?
Depending on how you want to run your test, you need to establish how long you want to run it.
At the first implementation of your test, both new and old visitors will react differently to it—therefore, giving you different conversion rates.
At some point, these numbers will be at their highest, lowest, or perhaps stagnant. However, it shouldn’t make you create an assumption based on your sole expectation.
Instead, let the number play out based on the duration of the test.
Let’s say you run a test on two variants (A & B) for 5 weeks.
In the 1st week,
Variant A resulted in a 17% conversion—while Variant B resulted in a 14% conversion.
In the 3rd week,
Variant A resulted in a 28% conversion —while Variant B resulted in a 23% conversion.
End of the 5th week,
Variant A resulted in a 12% conversion — while Variant B resulted in a 48% conversion.
Looking at this, most people will be convinced that Variant A is the best choice by the 3rd-week conversion rate —and even though that might be true in some cases, it’s better to let it play out.
What test method should you use?
There are two methods to test your hypothesis, namely:
- A/B Testing
- Multivariate testing
A/B Testing
This is by far the simplest form of testing and is mostly used when changes between designs and texts aren’t complex.
This is because it compares only two variants.
Multivariate Testing
For changes that are complex and require multiple variables to function, multivariate testing is best.
This method goes in-depth in testing variables of multiple functions such as text, color grading, fonts, font size, CTA position, and more.
Refer to your hypothesis to pinpoint the best method depending on your desired outcome.
For instance, you’re focused on optimizing the readability of your website’s text, and you’ve recognized the font as the problem.
Simple A/B testing will compare the two fonts together to see which one converts better.
With this test, you see that using the Roboto font will give you better conversion, which makes it a clear winner.
On the other hand, if you put this in a multivariate test, the results are likely to be different. Let’s say you’re running four variables to get the best font for your website.
By using multivariate testing, you can see that using the Calibri font gives a way better conversion compared to Roboto.
If we hadn’t used this method, we wouldn’t have known of other possibilities beyond those two variables.
Meanwhile, though multivariate testing can produce better insight into how web visitors react to changes, testing for too many variables simultaneously can affect the result.
This is because the essence of testing your hypothesis is to discover a pattern based on how users engage with the website.
In a case where 10 variables are tested at once, there’ll be little to no usable pattern available to establish quantitative results.
Examine Your Results
This is the point where you draw a temporary conclusion on your conversion rate optimization. It’s temporary because CRO is not all about a test, it’s an ever-changing process.
You can determine whether your test was a success or a failure from its results. Regardless of the outcome, you must dig deeper in knowing how it came to be and what can be done better next time.
Don’t create another test based on your previous wins or losses without examining every step of the process.
In two possible scenarios, here’s how you examine it.
- If your test was a success:
Analyze the whole conversion optimization process and answer these three questions.
- At what cost was the test a success?
- Could it have been done better and faster next time?
- Are there more opportunities to explore these results?
By breaking down your process this way, you’re able to guide yourself on your next conversion rate optimization test.
- If your test was a failure;
Don’t trash the whole thing —or you’ll make the same mistake. Instead, do this:
- Re-examine your initial data and look for possible corrections.
- Create a segment for similar data and another for different, but possible combinations.
- Go back to your hypothesis and see if it aligns with your initial sales funnel and customer journey.
- Look at the results from the user engagement, interaction, and drop-off point.
- Change your hypothesis based on different perspectives and new information.
- Run a split test on your wireframes to a small proportion of your visitors before pushing it live to a larger audience.
- Use the overall result to run the test again.
The best thing about the conversion rate optimization process is at every point your test fails or succeeds, you learn something valuable.
A test you ran 6 months back and resulted in a 68% conversion rate might not be recommended for you to run again due to how badly it’ll possibly turn out.
But take time to continuously iterate that test and learn more about the process every time you fail or succeed.
On a final note:
Here, I’ve discussed the steps to optimize your conversions. But I want to encourage you to continue to test at every point in your business. Conversion rate optimization is focused on people —and people change the way they interact.
Looking at how few years have impacted our system and adjusting from there proves that current “the right way” might not always be the right way.
So, it’s best to prepare your business for quick diversity and adaptation to give you a leg up.
We’ve come to the end of this chapter in our CRO Series. I hope you learned something today you can use in your next CRO campaign.
Before I go, I’d like to ask you a question:
Did you have a hard time with any of the steps listed here?
If customer experience, good conversions, low bounce rates, and overall — speed — matter to you, then you’ll definitely love the Enterprise Grade Web Acceleration Edgemesh offers.
With our intelligent, automated, and next-generation client-side caching, your website is ready to move at a full speed — with just a single line of code.
Plus, it takes less than 5 minutes to set up, so what do you say?
Start your 14-day trial here to get a feel for what speed means to your business.
