Landing page testing: Increase conversion rates and adoption with data
Discover how landing page testing can help to increase conversions and ensure your message resonates with your ICP.
Your landing page is a sales pitch aiming to convert visitors into leads. However, most landing pages only convert 4.6%[1] across all industries.
Landing page testing refines your sales pitch by checking how your company’s marketing messages land with your ideal customer and evaluating how likely they are to convert after encountering your page.
Testing might seem like a “nice-to-have,” but the results speak for themselves: clients using Wynter’s message testing tool have seen customers willing to pay 10x more and a 15% increase in conversions.
Unfortunately, not all landing page testing is created equal.
In this article, we’ll explore what landing page testing is and how to test your landing page copy the right way. We’ll also reveal some common mistakes companies make on their landing pages so you can start testing ahead of the game.
Landing page testing evaluates web page performance for user engagement, conversion rates, and overall success metrics to improve the visitor experience and generate higher conversion rates.
Marketers design landing pages with one objective: they want prospective customers to click on the call to action (CTA). This singular goal is why landing pages are critical to increasing your marketing campaigns’ landing page conversion rates and lowering the CPA.
Testing ensures you display the highest converting version of your page for visitors. When a visitor clicks on your CTA, that action is a result of everything put in place before it: the website’s loading time, the landing page’s design, its content, the CTA placement, etc.
Testing is particularly crucial before running PPC ads. You can test your messaging efficacy before you invest in your campaigns to optimize ROI.
Say you want to run a PPC campaign directing clicks to a landing page targeting a specific segment (say coaches and consultants), but your conversion rate on this page is low.
The only way to know why your page isn’t converting is by asking your target market. Then you can make changes to increase resonance before paying to direct more traffic to the page.
There are three traditional approaches to landing page testing:
Each test is used in specific cases, but they’re also limited in what insights they can provide.
A/B testing is the most common landing page testing method. Landing page A/B tests, also known as split testing, compare two or more versions of a web page and show them to different visitor segments to determine which version drives maximum results.
For example, you might test whether or not changing the CTA button color from blue to green results in greater CTRs.
The problem with A/B testing is that most B2B websites lack the traffic necessary for A/B testing since they target specialized market categories with fewer total customers. This is especially true when targeting multiple ICPs.
Conversion rate optimization (CRO) expert Michal Parizek says that even a 10% improvement on a 2% base conversion rate would require a sample size of 39,288 per variant.
A/B Tasty developed an A/B testing sample size calculator so you can see how many visitors you’ll need for an effective A/B test, depending on your conversion rate goals. Typically, tests require tens of thousands of visitors over at least a few weeks.
For these reasons, A/B testing isn’t a realistic landing page testing tactic for B2B companies.
Like A/B testing, multivariate testing assesses multiple variables simultaneously to determine which variant combination yields the most successful results.
For example, testing the color of the CTA plus the placement of headers and subheaders to see which variants increased conversions the most.
Testing for multiple variations requires significantly more traffic to reach statistical significance, so multivariate tests are not recommended for B2B companies with lower traffic volume.
Split URL testing compares multiple variations of your website hosted on two different URLs. Website traffic is split randomly between the variations, and you track conversions to discover which variation performs the best.
For example, if you want to redesign a website page, you can use a tool like Google Optimize:
Then, Google Optimize will send half of your traffic to URL A and the other to URL B.
Although split URL testing is similar to A/B testing, split testing requires that you host the variations on two different URLs.
Also, split URL testing is best for completely redesigning a web page, whereas A/B testing is preferred when testing small changes (e.g., button color, pop-up elements, form fields, etc.) to your web page.
Message testing measures the effectiveness of marketing messages by gauging how well it connects to and resonates with your intended audience. It does this by putting the landing page copy in front of users matching your ICP and collecting qualitative insights to improve it.
Qualitative insights reveal customer motivations through the lens of their behaviors, opinions, desires, and expectations.
During message testing, test givers ask respondents open-ended questions about the current messaging to reveal opinions and feelings in their own words which provides deeper insights and removes potential bias.
Message testing doesn’t have a traffic requirement, unlike the other landing page testing. You aren’t pitting different versions of your landing page against each other. Instead, you use direct, qualitative feedback from your ICP to determine the five components of effective messaging:
The other testing methods measure a landing page’s effectiveness by quantifying the data: How many customers signed up after going to Variant A on the website versus Variant B?
The problem is that there’s no way to know why it worked, or didn’t.
Was there language on Variant A that resonated better with the ICP? Did Variant B do a better job at differentiating our company?
Only message testing removes the guesswork. It reveals how your ICP perceives your landing page’s messaging: what’s confusing or unclear, what resonates, what doesn’t, and why.
The typical landing page testing process looks like this:
However, without asking your customers directly, or monitoring how they interact with your site through tools like heat maps, you’re still guessing at your messaging’s effectiveness.
Instead, use the following methods to compile a data-driven hypothesis before investing resources in page updates.
The easiest method to test your landing pages is through message testing. Wynter is a message testing tool that gives you feedback on your messaging efforts by allowing your ICP to examine and evaluate your messaging.
Wynter allows you to test using various methods:
When you sign up for Wynter, you can set up the test within minutes. The process looks like this:
After completing these steps, you’ll submit the test and will receive results within 12–48 hours.
When you receive your results from Wynter, you can immediately reflect the feedback in your copy for better conversions.
For example, when we message tested Cognism’s landing page, its users said many of the same things:
Cognism learned they had a differentiation issue and took steps to make their unique offer clearer—increasing conversion rates by 43%.
Web surveys collect data and information from a larger number of people, called samples, to reveal your target audience and uncover their motivations and desires.
Follow the steps below to conduct a web survey:
Every web survey must have an objective or you won’t know what to track. For landing page testing, refer to the five messaging components to guide your goals.
By the end of the test, you should know if your messaging:
Your goals at the beginning of your research act as a blueprint for the rest of the survey.
Full-scale market research methodologies use a blend of qualitative and quantitative methodologies.
Quantitative data is information you can measure or count given a numerical value. Qualitative data provides non-numerical information like qualities, feelings, or characteristics.
For web surveys, quantitative data is usually in Likert Scale form (i.e. 1 to 5) asked using closed-ended questions with predefined answers for respondents to select.
Qualitative data uses open-ended questions to provoke respondents to describe their answer using their own words, with no predefined prompts.
Use both quantitative and qualitative research methods so you can quantify how well something is working (or isn’t) and why it’s working (or isn’t).
Using both quantitative and qualitative methods requires a mix of closed- and open-ended questions.
Closed-ended questions:
Open-ended questions:
Once you write your questions, consider how many you will use in the survey. Stick to three to five closed- and open-ended questions total so you don’t overwhelm your respondents.
Your web survey must have a well-defined target audience, otherwise you won’t know whether or not the right people see your messaging in the way you want them to.
For example, if you want to find out how well your ideal customer receives your messages, consider emailing your past customers to find out how they feel about your landing page.
After you run your survey through SMS, email, or in-app or on-page surveys, synthesize the data using a spreadsheet, much like Wynter does for its message tests.
You can also use tools like Zapier to automate this process by extracting the results and loading it into a spreadsheet.
Uploading the results allows you to see any patterns that show up. Are people saying the same thing? Are they having similar hang ups on your landing page?
If a user says, “it took me a bit of time to fully read the headline. I believe it’s a bit too long,” now you know to shorten the headline so it says the same thing in fewer words.
Web surveys supplement the richer insights from message testing tools like Wynter.
Heatmaps are graphical data representations that use colors, shades, or patterns to show where customers interact on your page.
They help you understand where your customers click, what they click on, and what they ignore, helping you to identify trends and optimize your page to increase engagement and conversions.
By experimenting with how users interact with your website, you can evaluate your landing page’s overall effectiveness and tweak it for improvement.
There are many types of heat maps, including:
To set up a heat map on your landing page, you’ll need software like Hotjar or Glassbox. These services also offer tools to analyze the data.
When analyzing your heatmaps, ask the following questions:
Heatmaps are a powerful way to understand how users interact with your landing page—where they click, what they look at, what they ignore, and how far they scroll. You can use these insights to make changes to your landing page and improve the overall user journey.
Session recordings are recorded renderings of visitors' actions when browsing your website. They capture mouse movement, clicks, mobile taps, and on-page scrolling across devices.
Session recordings also reveal user hangups: where they’re confused and frustrated, which can also improve your website’s usability.
To set up a session recording, use a tool like Hotjar. Choose what pages to record, the test’s duration, and how many sessions to record. You can also set trigger session captures based on certain actions users take on your site.
With session recordings, you can:
Here’s an example of a screen record from Aillum Ltd:
It will take a few recording sessions to spot trends in visitor behavior and reveal how they interact with your website. You especially want to look for:
Session recordings give a holistic overview of your landing page’s user experience.
Now that you know how to conduct landing page testing, let’s review some common mistakes found in landing pages using some examples of B2B companies that Wynter message tested.
Rutter is a universal API that integrates with every commerce, accounting, and payments platform.
Back in September 2022, Wynter tested their website copy against its ICP of Product Managers (PMs). One PM commented, "The specific use cases that this enables feels a bit fuzzy. What exactly should I be able to implement now that I have these unified APIs with commerce data?"
The headline, “One API for commerce and business financial data,” wasn’t clear enough. PMs couldn’t tell what problems it solves and it lacks excitement.
The right-aligned design doesn’t show users what the app looks like, so users don’t know what the product looks like, further confusing them.
The CTA “Get API Keys,” left PMs wondering whom they’re selling to. The language sounds more geared towards developers, not the FinTech PMs they’re targeting.
When analyzing your landing page copy, ask yourself, what problem does this solve for my target audience? Did I spell the problem out for them or am I leaving them to guess what value I provide them?
Zuko Analytics is a form analytics platform that tells you when, where, and why potential customers drop out of your web forms and checkouts.
In June 2022, Wynter tested its messaging against a targeted panel of CRO and digital analytics professionals and found its copy to be extremely wordy, a common problem among companies who don’t want to limit themselves to select audiences.
The problem with this is it muddles the messaging, confusing the target audience and burying the company’s value beneath layers of words.
Even though Zuko says a lot in its subheading, they never get to the heart of what they or how they offer it to customers.
One respondent said,
“Show more rather than tell. The product can do a lot of the heavy lifting. Lead with the solution then introduce the product. I'm also very curious about some of the problems of the target audience. How will it save my marketing budget? Who exactly is this right for?”
When customers read your landing page, they want to skim. Less is always more when skimming and making decisions.
Metadata.io allows companies to automate their paid campaigns to drive more revenue. In June 2022, we tested its messaging against a panel of demand generation and Growth Marketing Directors.
One common problem in the panelists’ feedback was Metadata’s confusing positioning. Even though Metadata.io is an ABM (account-based marketing) tool, they never mention it on their website, except when they compare themselves to other tools using the G2 chart.
Although they compare to other ABM tools, this is too subtle for their ideal personas who, when skimming their web page, might not even look at this comparison chart.
Many panelists found Metadata’s positioning confusing or didn’t understand it at all.
They said:
Your customers shouldn’t have to guess at your use cases or your product positioning. Spell it out for them in your messaging. Are you an ABM tool? Say it. Don’t leave anything for customers to guess.
Landing page testing is an effective way to refine your messaging and increase conversions.
Wynter’s message testing results are fast, with results in 12–48 hours so you can quickly apply these changes to your landing page and roll out your new copy.
With these tools and testing ideas, you can create a high-converting landing page that resonates with your target audience and reduces friction in the buying process.
Out now: Watch our free B2B messaging course and learn all the techniques (from basic to advanced) to create messaging that resonates with your target customers.
