CRO: Conversion Research Methods

A good Conversion Rate Optimization Process starts with a good amount of research. Before diving into research techniques let’s see a situation I encountered recently.

  • The sales team from an SEO tool company finds me and asks if I would like to schedule a call to know their tool.
  • Takes details of the tools I currently use for my work.
  • Calls me on the scheduled day where I was waiting to know more about the tool.
  • Wants me to straightaway go for a free trial and was ready to answer questions after my trial.

Of course, I have a lot of questions to be answered such as,

  • How is the tool different from the tool I use?
  • What new problems that the tool solves?
  • What is in it for me?

So did you notice what the salesman missed here that could have earned him a potential customer to bring down the funnel? Yes, he did miss

  • Telling me how they are different from the tool I currently use.
  • Highlighting the features of the tool.
  • Giving a demo, so it is not overwhelming for me to try.

The same applies to the web pages since the websites are the salesmen in the digital world. It is important for the web pages to motivate the user, solve anxiety, answer questions, and reduce friction for successfully bringing the customer down the funnel. How can you do this methodically? The first step of a good conversion optimization process is good conversion research.

Conversion research has three major components.

  1. Experience Research
  2. Qualitative Research
  3. Quantitative Research
Types of Conversion Research

Site Walkthroughs

Walking through the key journeys of users is important to identify and fix the major gaps that users go through in the experiences with various devices and browsers. Identify the parts of the site where key metrics are affected to have quick wins.

Heuristic Analysis

Analyzing the problems without data helps to understand your site better and gives the necessary hypothesis which you can support with data. This analysis can be executed effectively only with a structured approach. You can use the popular heuristic analysis formula by Meclabs,

C = 4m + 3v + 2(i-f) — 2a


  • C = Probability of conversion
  • m = Motivation of user (when)
  • v = Clarity of the value proposition (why)
  • i = Incentive to take action
  • f = Friction elements of the process, a psychological resistance in the process
  • a = Anxiety about entering information

Usability Analysis

This approach is to analyze how easy the website is for the users to learn the design, quickly perform the tasks, remember the design when they return after a period, avoid errors, easily recover from the errors that occur, and feel satisfied with the overall experience.

To perform this test, you can have guidelines or checklists downloaded from the internet and start analyzing your website for each of these factors.

Survey Design Theory

It is important to know the theory behind a good survey design that focuses on getting the best information out of the audience without annoying them. You can follow the below suggestions to perform an efficient survey.

Be careful with your choice of questions, keep the number of questions limited, and make them short and precise. Make sure that the client does not have to spend more time than needed.

Always keep questions of similar nature together, avoid questions that don’t relate to your target audience, and keep the learning process that users have in the survey neutral.

Learn the voice of the customers, the words they use to connect closely to them. Try rotating the order of the questions as the questions at the top tend to get more positive answers.

Follow up with customers who have already interacted with your brand at least once. There is a lot of difference between the answers you will receive compared to a person who is not your customer yet. Make the survey short and while designing the questions, don’t assume their knowledge about your product.

Online surveys with recent customers

Survey the customers who bought your product or service most recently. The best sample size is somewhere between 100 to 200 and 7 to 10 questions will be an ideal number. Prioritize free-form answers over a selection from multiple choices. Give incentives like free downloads or a coupon to be used in the next purchase in return which can be used to pressure them to complete the survey in a fixed period of time. You can do it with Google Forms, a great free tool that is simple to use.

On-site polls

These are the web and exit surveys that appear on the website asking about the user’s experience. They are made to pop up based on certain trigger conditions like the time spent in a session or the number of pages in the session. This helps you understand if anything is holding the user from performing certain actions. Most often you find these polls on the bottom right side of the website.

Phone interviews

Some very successful tools are doing one-on-one conversations with the customers to know about their experiences and ideas to improve the website. Interviewing is the toughest type of surveying but this process gives the most unique information about each customer. You can begin a customer interview with permission to record the call and follow up with questions focusing on the below topics.

  • What customers like about your business
  • What you can improve
  • What customer’s pain points are

Live chat transcripts

Live chat is a great way to understand top pre-sales questions, the problems people have, their doubts, and what answers of yours persuade them to make decisions. So make sure all your chat transcripts are being recorded. It is best to start with people who already have given you the phone number.

Customer support insights

Finding out the problems that customers encounter from the customer support team gives great information about the frictions that customers are having.

User Testing

This deals with selecting people to try your website where you can gain insights from them. Testers who are not your target audience are also helpful in this process. A sample size of 10 to15 people is good enough to perform user testing. This can be done by assigning a set of tasks to the users.

Ex: Find pink shoes of size 6.5 under $75

Web Analytics

It is important to make sure that Google Analytics is perfectly installed and you are collecting what you need. What data should be included and what should be avoided. As an example, you must make sure that the necessary goals and funnel configurations are in place and check if the traffic from the office IP is filtered.

For an E-Commerce funnel, you can make sure you have the following data in place

  • Category pages
  • Product Pages
  • Add to Cart
  • View Cart
  • Checkout
  • Purchase Complete

Funnels and Goal Flows

Through the goal funnel, you can analyze, where users enter and exit the conversion/sales funnel. You can then eliminate bottlenecks in your conversion/sales process in order to improve your website conversions and sales.

Goal Flows help you analyze if users are having any problems in the funnels like higher drop-offs from a particular page.

Audience Insights

Audience insights are very important in understanding the following things

  • Geographical locations as a country, region, and city
  • New and returning visitors: A returning user to the website has much higher chances of making conversions than a new visitor
  • Demographics as age, gender
  • Interests such as affinity categories
  • Browsers, devices, and Screen Resolutions

Site Search

Site search data helps analyze the searches made by the user inside the site. You will find the search data in the behavior — site search area of Google Analytics. You can view the search terms as well as the pages on which the searches were made. Seeing these pages helps us understand which part of the website people use the internal search more and helps them have a better experience by giving more visible search boxes.

Although Google Analytics is a great tool and free to use, you can also try other quantified data tools like Adobe Analytics, KISSMetrics, MixPanel, and Heap Analytics.

Mouse tracking analysis

Mouse tracking analysis is performed to identify two things

Clicks: Click maps communicate about where users click and don’t click. If a large number of people are clicking on a particular spot where there is no button or link, then it is time to analyze if a link is needed in the spot.

Scroll: Scroll maps give an idea of how far users scroll. If a large number of people are not scrolling after a certain length of your long-form content, then it is time to analyze what makes them lose interest.

You can check HotJar, which is a widely used tool for this analysis. However, you need to keep in mind that these are algorithmic tools predicting stuff and not the actual users.

Having explored all the applicable means of collecting both qualitative and quantitative data, you are now ready to use the information to improve your user’s experience.

This is Kiruba Sekaran, a marketer, and a search engine optimizer sharing my experience with the CXL’s Conversion Optimization Mini Degree. I am happy and excited to be learning more from CXL in the following weeks.

SEO and Web Analytics Enthusiast

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store