Raise your hand if you’ve ever taken a test without studying beforehand. 🙋
How did it go?
Probably not very well, unless you’re really good at taking tests.
Similarly, too many digital marketers skip the preparation stage and jump straight into A/B testing before taking a look at their competitors, industries, and customers.
Don’t take the test before studying. More specifically, don’t jump into testing before doing your conversion research.
Conversion optimization research helps you create a plan to keep your testing tempo, improve your conversion rate optimization (CRO), and make more money. 🤑
In this article, we’ll teach you our go-to CRO research methodology so that you can run better tests, for better optimizations. Get ready to take some notes.
Get brand new conversion strategies straight to your inbox every week. 23,739 people already are!
What’s conversion research
Conversion research is a strategic approach in CRO.
It’s used to interpret relevant data so that you can identify and fix possible friction points in your sales funnel and increase your overall conversion rate. The most valuable data is that data that helps you understand your audience better. Those who understand their audience better are the ones who will win over the customers.
As Brian Clark puts it:
“The battle is won or lost, right here. Put me up against the greatest copywriter in the world, and if I understand the audience better, I will kick his or her ass whenever it comes to connection, engagement, and conversion.”
When you know how to track and interpret the correct data, you can address real problems and not arbitrarily guess what’s going on with your audience. This puts you in a better position to increase your conversions.
Walmart case study
After implementing conversion research, many companies and agencies have experienced significant improvements in their conversions.
One of those companies? Walmart.
After doing some research, Walmart realized that a considerable percentage of its visitors came from mobile devices. However, the look and feel of the site on mobile were awful and it took forever to load, so many website visitors would leave without converting. 👎
Once Walmart redesigned its site to be more mobile-friendly, conversions increased by 20%. 👍
However, this conversion boost wouldn’t have been achieved beforehand without conversion research).
The bottom line?
All significant improvements in a sales funnel are preceded by rigorous research. If you really want to succeed, you must take conversion research seriously.
Check your buyer persona against your audience
Who is your ideal customer? Do they match up with the customers you’re actually attracting?
20 years ago, marketers struggled to understand their audience and create actionable insights from them. Traditional marketing channels were hard to track, and the results weren’t accurate. In other words, nothing was easily correlated.
But nowadays, it’s much easier because digital channels are entirely measurable. You can see how many users visit your site, the percentage who buy, and tons of other relevant information that helps you make better decisions, but here are the top 6 metrics you definitely should be tracking.
For instance, with analytics tools like Google Analytics, you can find your visitors' age, language, interests, and many relevance patterns like bounce rate and time on site. And with this information, you should be able to see if your customers match your buyer personas.
New to buyer personas? Learn all about them in this article.
Now we move on to this magical data. Here’s what to look for.
In this section, you’ll find information that will help you determine who your audience is.
First, head over to your Google Analytics account and click on “Audience.”
Then, you’ll see the above list of subsections you can click on. Some of the primary audience subsections are:
- Demographics: age and gender
- Interests: affinity categories and market segments of your audience
- Geo: language and location
- Behavior: the time users spend on your website, what pages they visit the most, returning visits, and new visits
- Technology: what browsers visitors use the most
- Mobile: the type of devices your audience uses
- Users flow: the path your visitors follow from the moment they arrive at your site to when they leave it
Examples of how to use this research
Each one of these metrics could be a gold nugget of info. If your demographics point to an audience that’s 75% male, but your hero shot is of a group of females, test what happens when you change it to an image that also features males.
Let’s say your geographic data points to a huge chunk of users coming from France, but all your content is in English. Test what happens if translate a high-value page into French. More conversions? Maybe.
Another hypothetical: Maybe 25% of your traffic is coming from Android Webview. When’s the last time you looked up your website or landing page from an Android? Check it out. If it doesn’t look good, fix it.
This section will help you learn how your audience interacts with your website.
Again, head over to Google Analytics and click on “Behavior.”
Here, you’ll see the above sections. Some of the most pertinent metrics are listed below:
- Site content: how many visitors you’re getting on each one of your pages, landing pages, and exit pages
- Site speed: the average loading time of your website.
- Events: you can use this section to track data about interactions with your content. For example, “Play” button clicks
- Site search: if you’ve set up an on-site search, this section will show you the keywords and search terms people have used to find content within your site
Examples of how to use this research
So. Much. To. Learn. When’s the last time you looked at a report of your top exit pages? You just might find that there’s one page that makes up a bulk of exits. Uh oh. If that’s the case, there’s a problem. Visit that page on incognito or private mode from different types of browsers and devices to try and find the problem. Betchya it’s a pesky pop-up without a close button or something from a million years ago that you thought you had fixed or turned off.
I know you might not be that excited about site speed, but it can make or break your conversions. Check for slow-loading pages and then make sure you’ve optimized your website to load as fast as possible because time is money.
If you’re not totally sure about the events part of this, you have a TON of opportunities ahead of you. It’s the best way to measure the actions you want your audience to take. Why’s this amazing? If you can see that your audience is adding products to their cart but failing to complete the purchase, you then have the power to fix it.
Further reading: 12 Google Reports That Show What To Optimize
Ready to graduate to the next step? Let’s go.
Discover what’s working and what’s not
Using web analytics can learn a lot about what works and doesn’t work for you.
And in the end, you’ll be able to use the data we talk about below for your user testing.
Types of data
There are two primary data sources: qualitative and quantitative. Both of these data types have two subcategories of their own.
They provide different pieces of the same puzzle, so it’s essential to look into both.
From looking at this data, you can look for patterns in your audience to align your marketing efforts with those patterns.
Let’s get into qualitative data first.
Qualitative research is all about categorical data:
- Why do users think and behave in a certain way?
- What's their opinion about your products or services?
- Why do they choose to convert or not convert?
This type of data can also help you find industry trends, your audience’s biggest challenges, and even the exact words they use to describe those challenges.
Furthermore, the two subtypes of qualitative data are
- Nominal data: data used for labeling variables without a discernible order (e.g. gender, hair color, marital status, ethnicity, etc.)
- Ordinal data: data used for labeling variables with a discernible order (e.g. place in a competition, letter grade, etc.)
For instance, you could integrate a customer survey on your website or landing page. This way, you can understand why people leave and what you can do to improve their experience.
Suppose they volunteer first-party data about why your site doesn’t apply to their marital status. In that case, that’s considered nominal data. If they’re ranking their experience on your page, that’s deemed ordinal data.
By performing this type of research, you’ll be able to understand your audience more deeply.
As a matter of fact, StubHub increased conversions by 2.6% by performing qualitative research.
They found that their “See Details” link was confusing to visitors, so they removed it. That minor change allowed StubHub to make millions of dollars in extra revenue.
Thankfully, there are many methods you can use to find qualitative data, so you can choose what fits best for your situation. Here are the most common ones:
- Face-to-face interviews
- Focus groups
- Direct observation
- Open surveys
- Frequently asked questions
Quantitative research is all about the numbers to answer “how many, “how much,” and “how often” (e.g. 2,000 user sessions, 300 session recordings, etc.). For example, data like this can be found in heatmaps (e.g. scroll maps in Hotjar).
Here are the two types of quantitative data:
- Discrete data: data that is meaningful only by integers (e.g. 10 dogs)
- Continuous data: data that is meaningful at any capacity (e.g. 3.67 inches)
To further understand the difference, let’s look at the example below.
Interviews conclude that 61% of companies carry out less than five tests every month.
In this scenario, both 61% and 5 tests are discrete data. It isn’t helpful to know that 61.5% of companies carry out 3.7 tests because they are measured as a whole entity that cannot be broken down into parts in this context.
On the other hand, if you realize that a .83 second page load delay decreases your conversions, that’s continuous data — a second can be meaningfully broken down into smaller units, as seen here.
But, there’s no one-size-fits-all method to find quantitative data. Few quantitative methods often allow people to express their opinions, only predetermined choices.
For instance, in some multiple-choice surveys, participants can’t respond with their own words unless they can comment when they select a specific response (e.g. “Other”).
When done correctly, quantitative research can help you answer crucial questions:
- Is there a market for your products and services?
- What awareness is there of your product or service?
- How many people are interested in buying your product or service?
- What type of people are your best customers?
- What are their buying habits?
- How are the needs of your target market changing?
As you can see, data can open your eyes The correct information can help you perform conversion research the right way.
Now go ace your exams
See, studying isn’t so bad.
Conversion research is much simpler than it seems, and it pays off.
You’ll be prepared to tackle technical, usability, behavior-related, and website content issues by evaluating and improving facets like user experience (e.g. functionality) with the proper solutions (e.g. usability testing, heuristic analysis).
So, go ace those exams—well, increase conversion rates.
If you want to be a real all-star student, put your smarts to the test with a 12-step CRO Audit (and get some really easy conversion wins in the process)