Conversion rate optimization (CRO) is about understanding your users to bring them the experience they need or want to encourage a conversion. It is partially possible to gain such understanding through quantitative research like data measurement, analysis, and comparison. However, quantitative data alone tells you little about thought processes, motivations, and frustrations. For that, you need qualitative research.
Qualitative research for CRO involves observing and making inquiries. It gives you insight into the why, whereas CRO quantitative data tells you the what. In other words, qualitative data gives a voice of the customer.
When planning data collection, there are several qualitative research methods you should consider.
Surveys and Polls
To gain input from users, you can conduct full surveys or you can run something simple like a poll. The latter is a good option if you have a single question you want to be answered. It is also more scalable and gives you data from a larger number of users at a faster rate. CRO surveys, on the other hand, provide you with a wider range of information about who your customers are and how they feel about your product, service, or website. Both methods can have a place in your CRO strategy.
To run an effective survey, you need to know why you are collecting the information. The survey should have a specific purpose to yield actionable data for CRO. For instance, you may want a clearer picture as to who makes up your audience, what problems they are facing, and why they are interested in your product or service. Plus, it may be useful to know if they ran into any problems using your website or making a purchase, what their research process was like when picking your brand, and how long it took to make a purchase decision.
In addition to what questions you ask, you should think about when you provide the user with the survey. For web traffic surveys, you want questions to appear right as the user is about to leave your site, after the visitor has had the chance to interact with features and content. However, for CRO customer surveys, a better time is right after customers make a purchase and their experience with the buying process is fresh in their minds.
Users are more likely to answer a poll than a survey, as it requires almost no effort. Usually, a poll will take the form of a single question (although it can have two). It will appear at the bottom of certain web pages, perhaps at a particular moment in the user’s visit. Users who decide not to answer can simply close the window or ignore it.
You can encourage more responses by providing multiple-choice answers. Better still, add the option for users to supply more information if they wish. Even then, it will take just a few seconds for users to respond, meaning there’s still no need for incentives.
Another great thing about polls is that they allow you to reach users who are not yet leads (as you don’t need an email address). This means you may be able to gather valuable information about why some visitors never become leads. Furthermore, you can target just the visitors you want, displaying the poll to users based on criteria like whether they are new or returning visitors and traffic source.
Interviews take surveys a step further, allowing participants to answer questions freely. They also give you the chance to answer followup questions according to responses.
The problem with interviews is that they are even more to quantify (and, therefore, analyze) than surveys. However, there are ways around this. For instance, you can search for keywords or examine what words and phrases are occurring most often in responses. This will highlight similarities among your audience as well as common problems you may need to deal with.
User testing allows you watch how people complete tasks on your website. It gives you a better idea of your website usability. This is directly related to conversion rate — after all, if it is difficult for users to find the product they want to buy, to compare items, or complete the checkout process, they are unlikely to convert. The actions they need to take may seem obvious to you, but when you perform user testing, you may find that the process is actually counterintuitive.
To conduct a CRO user test, you need to enlist users who match the demographics of your target audience in terms of criteria like age, gender, location, and income. In addition, you need to ensure that you test with users on a variety of device types — to represent how your audience accesses your website. Aim to recruit between five and 15 participants. More than that is usually unnecessary.
Ask participants to carry out specific tasks, defining the actions you want them to take. You have two options to watch them in real time: in person or remotely using audio and screen captures. Either way, you should record your tests rather than relying on your memory alone. As you review, take notes.
In the case that you choose to watch tests in person, users should comment aloud about what they are doing, including if they run into any difficulties, and you can ask them questions. It is even possible to set up live views of users’ faces to read their expressions. Bear in mind, there are some disadvantages to this method. For instance, users may react differently due to your presence. Plus, it is complicated and expensive to organize. Lastly, you’ll only be able to test users in your area, which is no problem if you are a local brand but will be unrepresentative of your audience if you have a presence in a larger area.
Remote testing can lead to more accurate results than in-person testing. It also has the advantages of convenience and the capability for you to run a larger number of tests. Although you will lack commentary, you can still ask followup questions — besides, it is more important what users do than what they say. The main problems with remote testing are that it may require participants to sign up to a testing platform and that you may need to limit the length of testing sessions.
Live Chat Logs
The main purpose of live chat on your website is likely to answer FAQs, to provide customers with fast responses (including outside of business hours), and to take the strain off your customer service team. However, live chat can also join the suite of your CRO testing tools.
When you read live chat transcripts, you’ll likely find that many users are seeking support for the same onsite problems. They may be asking about missing information (or information that is difficult to find), details about certain products, or other issues that impact conversion rate. What is great about live chat is that it, instead of you needing to identify potential problems, users bring the problems to your attention.
To manage the data from your live chat, download the logs in a PDF or CSV format. Then, read through the transcripts to see what kinds of things users are talking about to detect patterns. Write down all the common issues in a spreadsheet and note how many times they occur. Issues that are mentioned the most often should form a source of further research.
Support Call Logs
Whether or not you have chat enabled on your website, you need to check out your support call logs. These may well reveal different queries to those from your live chat, especially if your onsite chat uses a chatbot.
Use the same procedure as above, identifying the most common queries. Also look at how your customer service team currently responds to questions — the solution to the issue may lie in your website or it may be in how you handle the query. Just remember that this stage in the conversion research process is about considering solutions, not about implementing anything.
You should already be monitoring user reviews of your products and services, as plenty of positive reviews are key to maintaining good SEO. Furthermore, you need to respond to reviews to show that you are listening. In addition, using reviews for qualitative research allows you to put a strategy in place to act upon common concerns.
Search for reviews in all the top places, including review sites, industry-specific websites, and Google My Business (if you are a local business) as well as your own website. If you currently lack reviews, consider beginning a campaign to garner more. For instance, you could ask for a review right after a purchase or offer a coupon to people who leave you a review.
Other Forms of Direct Feedback
Finally, you have direct feedback from places like social media, forums, and blogs — all of which can function as conversion rate optimization tools.
There are a variety of ways to gather qualitative data from social media. The first place to look is your own social media profiles. See what people are saying in the comments on your posts and in the reviews they leave. Another way is to monitor mentions of your brand — you have a number of tools to choose from to receive alerts in real time.
Watching for complaints about your brand, products, and services is a good reason to be active in industry forms, communities, and groups. You can also listen for mentions just like with social media.
Lastly, you need to monitor your own blog and the blogs of other people in your industry. On your blog, stay up to date with the comments you receive — you may receive unique insights into issues you have overlooked. As for the blogs of others, once again you can listen for mentions.
Once you have all the research you need, organize it to make it manageable and start looking for patterns. Highlight any important findings. Rather than jumping in to make changes, use these insights as inspiration for quantitative data gathering. Sample sizes from qualitative research tend to be too small to provide statistical significance, but they are excellent starting points for running website conversion testing. Then, you can use the results of these tests to implement website conversion optimization.
Originally published at insightwhale.com on November 8, 2018.