What Is Conversion Rate Optimization Process: Five-Step Guide
We often go about running conversion rate optimization campaigns on the basis of the tips and hacks available widely on the internet. But these tips fail to deliver in most cases. The reason behind this is that these blogs on conversion growth hacks don’t talk about the actual conversion rate optimization process that we should follow.
Randomly tweaking your websites and landing pages without following the conversion rate optimization process is like walking on a dark road without a roadmap. So, in this blog, we will talk about the conversion rate optimization roadmap that you should follow before implementing the CRO tips and tricks.
We have divided the entire conversion rate optimization process into five steps:
Step 1: Research & Data gathering
Marketers often tend to copy the most popular CRO strategies that have proven to deliver for other businesses. But, they fail. Why? Because what has worked for them might not work for you!
And this is precisely why the research and data gathering phase is of utmost importance. This helps you understand how visitors are reacting to your site and why they are doing so. Rather than optimizing purely out of gut feelings or copying CRO strategies, it’s better to take data-backed decisions. So, you should track and analyze your micro conversions before anything else. Wondering how? We’ve done a detailed blog on micro and macro conversions and how you can achieve them. So, do check it out!
So, how do you do that? Let’s show you two major ways!
Quantitative data analysis
Quantitative data analysis helps you understand how your visitors are reacting to your site or landing page. Tools like Google Analytics, Adobe Analytics, etc. can help you gather quantitative data like:
1. Traffic source
3. Click-through rate (CTR)
4. Time on site
5. Bounce rate
6. User’s device and browser information
In order to improve the CTR, time on site, and other KPIs, we often recommend using interactive landing pages or any other form of interactive content.
But first, what are interactive landing pages?
An interactive landing page is nothing but a normal landing page with interactive experiences like quizzes, calculators, assessments, etc. embedded in it. Unlike static landing pages, users on an interactive landing page directly interact with the page and are guided through an experience that generally generates qualified leads at the end.
And because interactive landing pages require the user to keep interacting, it keeps them hooked in for a longer period of time. Thus, it helps to improve important KPIs like time on site, bounce rate, etc., and improve conversions.
So, how can you extract data from these interactive landing pages? Well, if you use a tool like Outgrow to create your interactive landing pages then extracting these data will be a cakewalk. Because Outgrow comes with a robust analytics platform of its own that lets you extract and analyze the data you need.
You can also look into tools such as Lucky Orange, Hotjar, Crazy Egg, etc. to gather more insightful data through:
Heatmaps show a graphical representation of cursor movements on a particular web page at a given time frame. The hot areas (where cursor movements are concentrated) and ‘cold areas’ (where cursor movements are less) are represented using a color scale from red to blue. If you are conducting a CRO campaign, heat maps can help you determine the ‘cold areas’ i.e. areas of your page that users are not going. So, you can avoid placing your CTAs in such areas and move them to the ‘hot areas’ instead.
• Scroll maps
Scroll maps are also a type of heat map that shows how far users are scrolling a particular page. You would want to keep your CTAs and USPs limited to the areas your users scroll.
• Click maps
Click maps are another type of heat map that shows where users are clicking on a webpage. You can use it to track user engagements on your pages like button clicks, link clicks, video clicks, etc.
• Session recordings
Session recordings capture user movement on a page in a video format. It shows the actual footage of a user navigating a page along with button clicks, taps, scrolls, etc. It helps you understand how users interact with your site better than anything. You can use this data to improve the user experience on your site and ultimately increase your conversions.
Qualitative data analysis
Quantitative data analysis is great to start with but it’s not enough to paint a clear picture. It fails to answer the ‘why’ questions. And this where Qualitative analysis comes in.
Qualitative analysis is a subjective approach that answers why your users are behaving in a certain way.
So, how do you get this data? Conduct tests like:
2. On-site surveys
3. User interviews
4. Focus groups
5. Satisfaction surveys
You can conduct these surveys using a tool like Outgrow. It’s a no-code platform to create interactive polls, surveys, etc. without any coding knowledge. And once you have created the polls, you can embed them somewhere close to a newly updated element on your site. Ask your users if they like the new feature or not. Also, you can use these surveys to ask users where they are facing difficulties on your site or what additional features they want on your site.
Here’s an example of a Product Testing survey template to consider.
Conducting these tests and keeping a keen eye on user interactions in real-time can help you understand:
• Why are visitors abandoning your cart? Is it a bug, is the process is too lengthy, etc.
• Why does traffic from a certain device or browser have the maximum bounce rate?
• Why are your visitors not filling up lead forms?
Must read: How to create a poll on Outgrow?
Step 2: Hypothesis
Based on the quantitative and qualitative data you have collected, it’s time to form your hypothesis. So what exactly is a hypothesis from the CRO perspective? Here’s an example:
“If we add urgency messages like ‘Only a few left in stock!’ and ‘Sold out!’ just above the ‘Add To Cart’ buttons on our product page, we can have a 3% growth in cart additions.”
Now that you have your hypothesis (based on statistical data) ready, it’s time to implement the changes. But, how can you be sure if your hypothesis is going to work? Well, this is where A/B tests and multivariate testing comes into play.
But before you jump off to the testing phase, prioritize your work first. After steps 1 and 2, you may have discovered ample optimization opportunities and hypotheses. It can be a lot to digest all at once. So, you need to schedule everything as per the priority.
Step 3: Prioritization
Prioritizing your ideas and hypotheses will help you focus on the most important and serious problems on your site first. There are a lot of prioritization frameworks, but to cut things short, let’s take a quick look at the best few:
The P.I.E. framework is formulated by Chris Goward of WiderFunnel. The framework considers three factors: Potential, Importance, and Ease.
Using your analytics data, customer feedback, and thorough analysis of user scenarios, you need to figure out your worst-performing pages and rate them out of 10.
Now that you have figured out your worst performers, it’s time to figure out which of these pages are of utmost importance. So, how to define importance? Well, your pages with the highest volumes and costly traffic are the answer to that. You can segment these pages easily with your web analytics data and start ranking them out of 10.
The final step of consideration should be the ease of carrying out your tests. A page with massive potential for improvement and good traffic might as well face huge technical barriers to run your tests. So, you must also rank your pages considering the ease of optimizing along with ‘Potential’ and ‘Importance’.
Here’s an example of how you can carry out this framework:
So, the PIE framework essentially prioritizes your worst-performing pages, with important traffic and ease of modification.
The PXL framework is devised by Peep Laja of ConversionXL. Now, unlike the P.I.E framework, the PXL framework has a more objective approach. It can be difficult to rate your pages out of 10 (in the P.I.E framework). So, the PXL framework asks yes/no questions and assesses value and ease objectively. Suppose the answer to a certain question (mentioned below) is Yes then the value will be 2 else 0. So, it is much easier to scale.
This framework asks a set of questions:
• Is the change above the fold? → Changes that are made above the fold (upper half of a webpage) increases the chance of users noticing the change. Thus, it increases the likelihood of the test having an impact.
• Is the change noticeable in less than 5 seconds? → If you are thinking of implementing new controls or variations on your site, then check if users can notice the change in under 5 secs. If not, it’s likely to have less impact.
• Does it add or remove anything? → If you change or add certain elements like removing distractions or adding key information, it should have more impact.
• Does the test run on high traffic pages? → Check if the page you are thinking to optimize has high traffic. Making improvements on a high traffic page results in more revenue.
The TIR Framework (Time, Impact, Resources) is given by Bryan Eisenberg. The rating scale here runs from 1 to 5.
This step ranks the pages out of 5 based on the number of days, man-hours, development hours, etc. required to achieve the optimum impact of the test you want to conduct.
Ranks your page based on the amount of revenue that can be generated if the test is successful.
This step takes into consideration the total cost (people, tools, space, etc.) that is to be incurred for running your test on a particular page.
Step 4: Implementation and Testing
This is one of the most important steps of the conversion rate optimization process. Once you are ready with your data, hypothesis, and prioritization, it is time to test and implement. Testing of your hypotheses can be done primarily in three ways:
A/B testing is a research methodology used to test user experience between two variants A and B. It is used to test simple optimizations like minor design and layout tweaks.
Grene, an eCommerce company, revamped its mini cart page and ran an A/B test that resulted in a 60% increase in revenues.
This is how their actual page looked like before:
They made the following changes to their mini cart page:
1. Added a CTA button at the top of their mini cart to facilitate a smooth and quick transition to the actual cart page.
2. Added a ‘Remove’ button alongside each item and the total value of products.
3. Increased the size of their ‘Add to Cart’ button to improve prominence.
They ran the campaign for 36 days and got the following results:
1. More cart page visits (from the CTA at the top)
2. Improved conversion rate from 1.83% to 1.96%
3. 2x increase in the purchase
Split testing, also known as split URL testing, is used to test more complex changes. In other words, split testing is done to compare two different variations of a page with separate URLs. Sometimes when your design optimization requirements are too heavy and complex, you have no choice but to create a different page with a different URL.
Multivariate testing is used when you have to make multiple changes on a page and at the same time test each combination separately.
Let’s look at this product page of Apple to understand what you can test using MVT testing:
But to run these tests, you need to have the best tools at your disposal. Tools like Google Optimize, Optimizely, etc. can help you with A/B tests, Split tests, and Multivariate tests.
Step 5: Learning and Reviewing
Now that you have your test results in front of you, either you will find that your hypothesis is correct or it has not turned out as expected. Generally speaking, if the test results come out positive then marketers tend to implement the changes straight away. But, if the results come negative, they go back to form new hypotheses.
But, you must dig deeper even if your hypothesis is correct. Always check if the cost of implementing the changes is doing justice to the revenue it is supposed to generate.
Also, if your hypothesis is wrong, review it for loopholes before discarding it completely. Recheck your test results and research data before you move on to form a different hypothesis.
Again, the best way to get primary and accurate data or feedback on any change you make is to directly ask your customers how they feel about it. So what’s the best way to do it? Well, traditional ways and long boring forms just don’t work anymore. Here is an interactive “Vote for your favorite #GoogleMarketingLive Feature” poll. It takes just a single click from the user and the brands collect real-time feedback on their feature/product updates.
So, these were the precise 5 steps to master the conversion rate optimization process. Well, so far you must have understood the steps you should follow while creating your Conversion Rate Optimization strategy. Try implementing these steps and results should follow soon! But what exactly can you do to optimize your CRO? We’ve got all the strategies, best practices, tips, and tools to take your CRO efforts to the next level.
Before we bid adieu, here’s one last gift for you! Remember we were talking about the effectiveness of using different types of interactive content in the data-gathering section? Well, if you are interested in creating interactive content for your brand, then follow this link and create them for free with Outgrow!
Roy is a full-time content marketer at Outgrow. He often travels to the Himalayas in search of his muse to breathe. Being a professional marketer & a passionate traveler, he struggles to keep the balance.