This is the second part of the Conversion Research episode. After talking about heuristic evaluation and the Lift model, we took a deep dive into quantitative and qualitative research to optimize your site for conversion.
If there’s one takeaway from this topic, don’t simply rely on best practices. Understand your audience and always do your own testing.
Links in this episode:
In part one, we talked about heuristic evaluation and different frameworks you can use for that. In this episode, we will talk about quantitative and qualitative analysis.
Quantitative research is data that you can represent using numbers and normally the tool for this would be Google Analytics. You’d go to your Google Analytics account and you would look for things like bounce rate, percentage of exits, conversion rates, stuff like that.
It’s important, especially for small businesses to make sure that first, you have tracking tools installed. And the good news is you don’t need really complex tools are very expensive tools. You can get started with Google Analytics, which is totally free. The goal of this quantitative step is really to understand or try to understand what kind of data you can get and understand how your users are currently behaving on your site.
Installing Google Analytics, like the basic installation, that tracks page views, sessions, users, bounce rate. You need to add a snippet of your coach to your website, or ideally to Google Tag Manager, if you have it.
Using Google Analytics for Conversion Research
So install Google analytics and then where do you start? Well, the first thing is never to look at total numbers. I never looked at users like they’re one group, like one uniform group, because they absolutely are not. And I’ll go back to mobile. It’s best to start by looking at mobile versus desktop. You’ll still look for things like where they land or your channels, consider the device if that matters to you.
If you want to start simple, just look at the page that you’re trying to fix and see how many people are leaving, through that page or ending their sessions there. Split mobile versus desktop. If you see that 20% of people are ending the session they’re on desktop, but 80% on mobile. It’s not the content most likely. It’s how it’s presented on a mobile phone and that’s where you need to work.
Using Heat Maps And Scroll Maps
The other method is more visual. Heat maps and scroll maps allow you to see how people are interacting with your page or how far people are scrolling down the page. There are many tools that you can use for this, like HotJar, which we talked about many times, Microsoft Clarity, Crazy Egg and more.
To get started, install a heatmap tool in Google Tag Manager, you can add the snippet of tag there. You give it a few days to gather enough data on your most important pages or landing pages. Let’s say, you have a product page, you set up a heat map, you set up a scroll map. You can see where people are clicking and where people are not clicking even more importantly.
If you find that people are clicking the wrong element or something that’s not even clickable because it looks like a button, that’s a very obvious fix. And, and that’s really important because, you know, maybe your design or your interface is not clear enough and data. They don’t really get what they should be doing, or your copy is not guiding them that way.
Next thing is scroll maps. It’s really similar. You can analyze scroll. And this helps you, especially on mobile again, because if add to cart button is not above the fold and 40% of people don’t reach that button and scrolled that far, you obviously have a problem. A tool like Hotjar will help you confirm these issues.
So with Google Analytics, you will find which page the issues are on. With heat maps and scroll maps, you will find what exactly is wrong with that page. So those two work really well together.
This is something that can not be directly represented with a number. So with the quantitative, you’re trying to gauge what their behavior is. With qualitative research, you figure out the why behind their behavior. This is where you try to see their drivers and motivations and barriers, and it will be of course, qualitative, but this will make your data richer when you try to interpret it and use it to help you prioritize which pages to fix first.
So this is the most consuming one. There’s no way around it. You need to analyze the data. But this is a gold mine. This is really people telling you, hey, I’m having a problem with and if you fix that thing, if a lot of people are having problems, your website will perform better.
That’s the gold standard. Ideally what you want is to have someone who falls under your ICP or your ideal profile and watch them interact with your page. Usability testing is the best by far. It is a sort of a lab controlled environment where the user sets you sit next to you, look at them and they have to talk to you while they’re interacting with the website.
So you give them a task. I want you to find seven different green jackets, and I want you to find a t-shirt and put that in cart and then remove five jackets, something like that. Something that your users may want to do. And you’ll sit there, you look at them, you look at their face because you know, when you’re frustrated by websites, you can’t really hide that and you have them walk you through what they’re trying to do.
This is very expensive because you get into find a users it’s best done in person, because you want to really make them feel you’re nice and easy before they started their session. Now it has to be done remotely, I guess, because of the situation we’re in, but this is something that you don’t need a hundred people testing. Three to five people you’ll find the most obvious issues or most of the obvious issues. We just have many people and it can be done with honest friends or honest people or who are not going to try.
So surveys can be done in many ways. Some of the things that we can try doing is an onsite survey. So maybe a survey that will pop up on the chatbox, or maybe when you’re, about to check out or it’s like an exit-intent survey, or sometimes you can do it right after the process. You can follow up with a post-purchase email and ask about their experience. And the other thing is the NPS survey to check, like how satisfied they are or how likely would you recommend them to others.
But don’t show this immediately on the product page, you know, maybe a good way. You mentioned a post-conversion survey. That’s one thing where they recap their experience. And if you have one that’s before conversion, everything like that is going to risk hurting that conversion.
One other way to do it is to contact your existing customers. You’ll be surprised that maybe a coupon code for their next order or some that are really happy with our product or your service will be very happy to participate in a survey.
Customer Service Data
This is your bug tracker in a way, customer service data. If there’s something that a lot of users complain about frequently, even if you think it’s right, even if you think it’s a perfect UI, you have to fix the issue for them, or explain on the website.
So definitely check out your customer support, talk to your customer support staff. They’ll be able to tell you off the bat, this is the most common complaint that we get. Sometimes it will be an easy fix. It might be just a link that’s not working, or it can be something that’s a bigger issue.
You can watch session recordings in tools like HotJar. Sessions that had an event, like added to cart or sessions that start with a certain page and would just end with a certain page. This is time-consuming but at least do it like once a week or once a month with this tool, you’ll be able to see, even if you don’t have the budget to have a full usability test, you can see how they actually interact with your website.
And that’s how you can find the most painful issues that users have with your website. So sessions. Yes, they are very time-consuming dare, dare faster ways to get customer data, but sessions can be extremely valuable because if quantitative research and web analytics show you the page where the problem is, and then heat maps and scroll maps are, show me exactly where, on the page, with sessions, this is the element that’s causing an issue like this is exactly the one.
Final Thoughts and Recap
Combining qualitative feedback with your data will help give you a deeper understanding of your user behavior. When you have a better understanding, it will also help you point out the main issues. Because of course there are lots to fix for sure. You’ll have your own things that you want to fix. And then those that will be surfaced from this customer data and your qualitative data.
It all goes back to understanding your user to guide you on where to prioritize your CRO or your A/B tests. Understanding your users and these three let’s call them pillars of conversion research in heuristic evaluation, quantitative research, and qualitative research. What you’re trying to understand is what your user thinks or struggles with when using the website.
And always, always, always remember when you’re optimizing a website, unless it’s speed optimization, you’re optimizing it for a specific audience. Never try to apply anything general. There there’s nothing that’s going to work for everyone.
Don’t forget to subscribe to the No Hacks Marketing Podcast. And if you learned something new today, we would appreciate it if you can leave us a review on your favorite podcast platform.