Whether you are working on a landing page or the homepage of your website, you may be asking yourself, "Why aren't people converting? What elements are helping or hurting my user experience?"
Those are good questions.
When it comes to website or landing page design, there are dozens — if not hundreds — of potential elements to test. And that's before you start testing how different combinations of elements affect performance.
Launching a test
The good news is, after running thousands of tests for websites in almost every industry you can imagine, we've created a simple way to quickly identify the most important areas of opportunity on your site or landing page.
We call this approach the "launch analysis".
Why? Well, getting someone to convert is a lot like trying to launch a rocket into outer space. To succeed in either situation, you need to generate enough momentum to overcome any resistance.
To get a rocket into orbit, the propulsion and guidance systems have to overcome gravity and air friction. To get a potential customer to convert, your CTA, content and value proposition have to overcome any diversions, anxiety or responsiveness issues on your site.
So, if you really want your conversion rate to "take off" (see what I did there?), you need to take a hard look at each of these six factors.
Prepping for launch
Before we dive into the launch analysis and start testing, it's important to take a moment to review 3 important testing factors. After all, no matter how good your analysis is, if your test is fundamentally broken, you'll never make any progress.
With that in mind, here are three questions to ask yourself before you dive into the launch analysis:
What is my business question?
Every good website or landing page test should answer some sort of important business question. These are usually open-ended questions like "how much content should be on the page to maximize conversions?" or "what does the best-converting above-the-fold experience look like?"
If your test is designed to answer a fundamental business question, every test is a success. Even if your new design doesn't outperform the original, your test still helps get you get some data around what really matters to your audience.
What is my hypothesis?
Where your business question may be relatively broad, your testing hypothesis should be very specific. A good hypothesis should be an if/then statement that answers the business question (if we do X, Y will happen).
So, if your business question is "how much content should be on the page?", your hypothesis might be: "if we reduce the amount of content on our page, mobile conversions will increase." (If you're interested, this is actually something we studied at Disruptive Advertising.)
What am I measuring?
We hinted at this in the last section, but every good test needs a defined, measurable success metric. For example, "if we reduce the amount of content on our page, people will like our content more" is a perfectly valid hypothesis, but it would be incredibly difficult to define or measure, which would make our test useless.
When it comes to online advertising, there are tons of well-defined, actually measurable metrics you can use (link clicks, time on page, bounce rate, conversion rate, cart abandonment rate, etc.) to determine success or failure. Pick one that makes sense and use it to measure the results of your test.
The launch analysis and countdown
Now that we have the testing basics out of the way, we can dive into the launch analysis. When performing a launch analysis on a page of your site, it is critical that you try to look at your page objectively, and identify potential opportunities instead of immediately jumping into things you need to change. Testing is about discovering what your audience wants, not about making assumptions.
With that being said, let's countdown to launch!
6. Value proposition
To put it simply, your value proposition is what motivates potential customers to buy.
Have you ever wanted something really badly? Badly enough that you spent days, weeks, or even months figuring out how to get it for an affordable price? If you want something badly enough (or, in other words, if the value proposition is good enough), you'll conquer any obstacle to get it.
This same principle applies to your website. If you can really sell people on your value proposition, they'll be motivated enough to overcome a lot of potential obstacles (giving their personal information, dealing with poor navigation, etc.).
For example, a while back, we were helping a college optimize the following page on their site:
It wasn't a bad page to begin with, but we believed there was opportunity to test some stronger value propositions. "Get Started on the Right Path: Prepare yourself for a better future by earning your degree from Pioneer Pacific College" doesn't sound all that exciting, does it?
There's a reason for that.
In business terms, your value proposition can be described as "motivation = perceived benefits - perceived costs." Pioneer Pacific's value proposition made it sound like going to all the work to get a degree from their college was just the beginning of a long, hard process. Not only that, but it wasn't really hitting on any of the potential pain points an aspiring student might have.
In this particular case, the value proposition minimized the perceived benefits while maximizing the perceived costs. That's not a great way to get someone to sign up.
With that in mind, we decided to try something different. We hypothesized that focusing on the monetary benefits of earning a degree (increased income) would increase the perceived benefits and talking about paying for a degree as an investment would decrease the perceived cost.
So, we rewrote the copy in the box to reflect our revised value proposition and tested it:
As you can see above, simply tweaking the value proposition increased form fills by 49.5%! The form didn't change, but because our users were more motivated by the value proposition, they were more willing to give out their information.
Unfortunately, many businesses struggle with this essential step.
Some websites lack a clear value proposition. Others have a value proposition, but it makes potential customers think more about the costs than the benefits. Some have a good cost-benefit ratio, but the proposition is poorly communicated, and users struggle to connect with it.
So, if you're running the launch analysis on your own site or landing page, start by taking a look at your value proposition. Is it easy to find and understand? Does it address the benefits and costs that your audience actually cares about? Could you potentially focus on different aspects of your value propositions to discover what your audience really cares about?
If you think there's room for improvement, you've just identified a great testing opportunity!
5. Call to action
If you've been in marketing for a while, you've probably heard all about the importance of a good call to action (CTA), so it should come as no surprise that the CTA is a key part of the launch analysis.
In terms of our rocket analogy, your CTA is a lot like a navigation system for your potential customers. All the rocket fuel in the world won't get you to your destination if you don't know where you're going.
In that regard, it's important to remember that your CTA typically needs to be very explicit (tell them what to do and/or what to expect). After all, your potential customers are depending on your CTA to navigate them to their destination.
For example, another one of our clients was trying to increase eBook downloads. Their original CTA read "Download Now", but we hypothesized that changing the CTA to emphasize speed might improve their conversion rate.
So, we rephrased the CTA to read "Instant Download" instead. As it turned out, this simple change to the CTA increased downloads by 12.6%!
The download was just as instantaneous in both cases; but, simply by making it clear that users would get immediate access to this content, we were able to drive a lot more conversions.
Of course, there is such a thing as being too explicit. While people want to know what to do next, they also like to feel like they are in the driver's seat, so sometimes soft CTAs like "Get More Information" can deliver better results than a more direct CTA like "Request a Free Demo Today!"
As you start to play around with CTA testing ideas, it's important to remember the 2-second rule: If a user can't figure out what they are supposed to do within two seconds, something needs to change.
To see if your CTA follows this rule, ask a friend or a coworker who has never seen your page or site before to look at it for two seconds and then ask them what they think they are supposed to do next. If they don't have a ready answer, you just discovered another testing opportunity.
Case in point: On the page below, a client of ours was trying to drive phone calls with the CTA on the right. From a design perspective, the CTA fit the color scheme of the page nicely, but it didn't really draw much attention.
Since driving calls was a big deal for the client, we decided to revamp the CTA. We made the CTA a contrasting red color and expanded on the value proposition.
The result? Our new, eye-catching CTA increased calls by a whopping 83.8%.
So, if your CTA is hard to find, consider changing the size, location and/or color. If your CTA is vague, try being more explicit (or vice versa). If your CTA doesn't have a clear value proposition, find a way to make the benefits of converting more obvious. The possibilities are endless.
4. Content
Like your value proposition, your content is a big motivating factor for your users. In fact, great content is how you sell people on your value proposition, so content can make or break your site.
The only problem is, as marketers and business owners, we have a tendency towards egocentrism. There are so many things that we love about our business and that make it special that we often overwhelm users with content that they frankly don't care about.
Or, alternatively, we fail to include content that will help potential customers along in the conversion process because it isn't a high priority to us.
To really get the most out of your content, you have to lay your ego and personal preference aside and ask yourself questions like:
- How much content do my users want?
- What format do they want the content in?
- Do mobile and desktop users want different amounts of content?
As a quick example of this, we were working with a healthcare client (an industry that is notoriously long-winded) to maximize eBook downloads on the following page:
As you can see above, the original page included a table of contents-style description of what readers would get when they downloaded the guide.
We hypothesized that this sort of approach, with its wordy chapter titles and and formal feel, did not make the eBook seem like a user-friendly guide. There was so much content that it was hard to get a quick feel for what the eBook was actually about.
To address this, we tried boiling the copy down to a quick, easy-to-read summary of the eBook content:
Incredibly, paring the content down to a very simplified summary increased eBook downloads by 57.82%!
However, when it comes to content, less is not always more.
While working on a pop-up for Social Media Examiner, we tested a couple different variants of the following copy in an effort to increase eBook downloads and subscriptions:
Just like the preceding example, this copy was a bit wordy and hard to read. So, we tried turning the copy into bullet points...
...and even tried boiling it down to the bare essentials:
However, when the test results came in, both of these variants had a lower conversion rate than the original, word-dense content!
These results fly in the face of the whole "less is more" dogma marketers love to preach, which just goes to show how important it is to test your content.
So, when it comes to content, don't be afraid to try cutting things down. But, you might also try bulking things up in some places — provided that your content is focused on what your potential customers want and need, not just your favorite talking points. Our suggestion: challenge whatever you have on your site. Try less, more, and different variations of the same. It should ultimately be up to your audience!
3. Diversions
Unfortunately, having a great value proposition, CTA and content doesn't guarantee you a great conversion rate. To get a rocket to its destination, the launch team has to overcome a variety of obstacles.
Same goes for the launch analysis.
Now that we've talked about how to maximize motivation, it's time to talk about ways to reduce obstacles and friction points on your site or page that may be keeping people from converting, starting with diversions.
When it comes to site testing, diversions could be anything that has the potential to distract your user from reaching their destination. Contrasting buttons, images, other offers, menus, links, content, pop ups...like cloud cover on launch day, if it leads people off course, it's a diversion.
For example, take a look at the page below. There are 5 major elements on the page competing for your attention - none of which are a CTA to view the product - and that's just above the fold!
What did this client really want people to do? Watch a video? Read a review? Look at the picture? Read the Q&A? Visit their cart?
As it turns out, the answer is "none of the above".
What the client really wanted was for people to come to their site, look at their products and make a purchase. But, with all the diversions on their site, people were getting lost before they even had a chance to see the client's products.
To put the focus where it belonged—on the products—we tried eliminating all of the diversions by redesigning the site experience to focus on product call to actions. That way, when people came to the page, they immediately saw Cobra's products and a simple CTA that said "Shop Our Products".
The new page design increased revenue (not just conversions) by 69.2%!
We've seen similar results with many of our eCommerce clients. For example, we often test to see how removing different elements and offers from a client's homepage affects their conversion rates (this is called "existence testing").
Existence testing is one of the easiest, fastest ways to discover what is distracting from conversions and what is helping conversions. If you remove something from your page and conversion rates go down, that item is helpful to the conversion process. If you remove something and conversion rates go up - Bingo! You found a distraction.
The GIF below shows you how this works. Essentially, you just remove a page element and then see which version of the page performs better. Easy enough, right?
For this particular client, we tested to see how removing 8 different elements from their home page would affect their revenue. As it turned out, 6 of the 8 elements were actually decreasing their revenue!
By eliminating those elements during our test, their revenue-per-visit (RPV) increased by 59%.
Why? Well, once again, we discovered things that were diversions to the user experience (as it turns out, the diversions were other products!).
If you're curious to see how different page or site elements affect your conversion rate, existence testing can be a great way to go. Simply create a page variant without the element in question and see what happens!
2. Anxiety
Ever have that moment when you're driving a car and you suddenly get hit by a huge gust of wind? What happens to your heart rate?
Now imagine you're piloting a multi-billion dollar rocket…
Whether you're in the driver's seat or an office chair, anxiety is never a good thing. Unfortunately, when it comes to your site, people are already in a state of high alert. Anything that adds to their stress level (clicking on something that isn't clickable, feeling confused or swindled) may lead to you losing a customer.
Of course, anxiety-inducing elements on a website are typically more subtle than hurricane-force winds on launch day. It might be as simple as an unintuitive user interface, an overly long form or a page element that doesn't do what the user expects.
As a quick example, one of our eCommerce clients had a mobile page that forced users to scroll all the the way back up to the top of the page to make a purchase.
So, we decided to try a floating "Buy Now" button that people could use to quickly buy the item once they'd read all about it:
Yes, scrolling to the top of the page seems like a relatively small inconvenience, but eliminating this source of anxiety improved the conversion rate by 6.7%.
Even more importantly, it increased the RPV by $1.54.
Given the client's traffic volume, this was a huge win!
As you can probably imagine, the less confusion, alarm, frustration and work your site creates for users, the more likely they are to convert.
When you get right down to it, conversion should be a seamless, almost brainless process. If a potential customer ever stops to think, "Wait, what?" on their journey to conversion, you've got a real problem.
To identify potential anxiety-inducing elements on your site or page, try going through the whole conversion process on your site (better yet, have someone else do it and describe their experience to you). Watch for situations or content that force you to think. Odds are, you've just discovered a testing opportunity.
1. Responsiveness
Finally, the last element of the launch analysis is responsiveness—specifically mobile responsiveness.
To be honest, mobile responsiveness is not the same thing as having a mobile responsive site, just like launching a rocket on a rainy day is not the same thing as launching a rocket on a clear day.
The days of making your site "mobile responsive" and calling it good are over. With well over half of internet searches taking place on mobile devices, the question you need to ask yourself isn't "Is my site mobile responsive?" What you should be asking yourself is, "Is my site customized for mobile?"
For example, here is what one of our clients' "mobile responsive" pages looks like:
While this page passed Google's "mobile friendly" test, it wasn't exactly a "user friendly" experience.
To fix that problem, we decided to test a couple of custom mobile pages:
The results were truly impressive. Both variants clearly outperformed the original "mobile responsive" design and the winning variant increased calls by 84% and booked appointments by 41%!
So, if you haven't taken the time yet to create a custom mobile experience, you're probably missing out on a huge opportunity. It might take a few tests to nail down the right design for your mobile users, but most sites can expect big results from a little mobile experience testing.
As you brainstorm ways to test your mobile experience, remember, your mobile users aren't usually looking for the same things as your desktop users. Most mobile users have very specific goals in mind and they want it to be as easy as possible to achieve those goals.
Launch!
Well, that's it! You're ready for launch!
Go through your site or page and take a look at how what you can do to strengthen your value proposition, CTA and content. Then, identify things that may potentially be diversions, anxiety-inducing elements or responsiveness issues that are preventing people from converting.
By the time you finish your launch analysis, you should have tons of testing ideas to try. Put together a plan that focuses on your biggest opportunities or problems first and then refine from there. Happy testing!
This article has been perfect for me, since I was just getting into the A / B test to improve conversions. Thank you.
Awesome! Glad it was helpful.
Great article, exactly what I needed. I've been looking for a resource to reference to start getting my CRO program off the ground. I will definitely be referring back to this post.
When first getting into testing do you recommend starting with the value prop, then CTA, and then continuing in the order you laid out in the post?
Hey John, I'm happy to hear you are getting started with CRO!
Initially I would recommend you work your way through the list. Over time, you will start to learn which ones really affect conversion rates (you won't always see results on every single area), and you can focus more heavily on the areas that impact your users.
A few tips I have as you get started:
Thanks really a lot for your great article.
Its easy to understand, to the point and contains really actionable stuff you can consider for your case and then check it out.
--> Whats the best way to find out what mobile users want in comparison to desktop users?
Great question Cesare!
Understanding what your mobile users want in comparison to your desktop users is a critical part of any site testing strategy. What I would suggest is that you take the first 5 areas of the launch analysis (value proposition, call to action, content, distractions, anxiety) and test each of these specifically for your mobile users.
For many of our clients, we will have a separate test strategy for mobile and desktop, since the two experiences should be different and customized.
The biggest things that you'll usually want to focus on discovering (testing) for your mobile traffic: What call to action do they want? (click to call can work VERY well for mobile users), how much content do they want? (a lot of times it's less than your desktop users want), and how should I lay out my mobile pages? (take into consideration the limited screen space).
Hi Chris,
Thanks a lot. You are right. The CTA is most likely one of the things that may differ at least you have to think about it properly.
By the way: articles about CRO (Conversion Rate Optimization) are the ones that interest me the most. So I am looking forward to reading more from you...
Thanks Cesare! Hopefully we will get another one up on the Moz blog soon :)
Chris that is definitely a must read on conversion optimization and a great reference piece. I am jealous you get to see all that data. lol
If I may, I would also add page load time and server response to the controls. Just to make everything even, especially on a shared server. Because as you stated time is of the essence for a visitor to decide to stay or bounce. And well sometimes the internet is just.. um wonky slow.
Haha thanks William, it is definitely fun to see test results like this on a daily basis.
You bring up a great point about load times and server response - slow load times DEFINITELY affects conversion rates, and is something to keep a close eye on as you go through the testing process. Excellent feedback.
Totally agree, load time is a very important point.
Thanks Chris!! Testing conversions is a really great topic that offers continuous improvements to our e-commerces. But to do it properly we should have a representative amount of traffic, if we want to take the right decisions. Any suggestion about how many visitors makes that tests trustful?
Great question Toni. There are a lot of articles written about testing sample sizes and how long you should run a test. Here is a great one from Optimizely that also includes a neat tool they built to calculate how big of a sample size you should have based on the conversion rate of your page: https://help.optimizely.com/Analyze_Results/How_lo...
A few rules that I have for my team:
Thanks for the answer Chris!!
The resource is so good, specially the Sample Size Calculator, and also your team rules are great ;-)
It's very impressive how little changes make so much benefit!
I think I have to work on "Diversions". Now I have many CTA to different pages, so I will test putting only one clear focus CTA.
Thank you Chris!
You bet. It's easy to try to cram a bunch of CTAs on a page to make sure you give users all the options possible. I would love to hear how your testing goes!
Great Post! I love A|B testing my website. I find it hard though in niche with low volume traffic to apply theses principles. Any ideas on how to test results on only a few users in order to understand their own preferences? Example, a manufacturer that wants to optimize its website for his 50-60 clients.
Good question. It's definitely a challenge when you have very low numbers of visitors like what you mentioned.
In those cases, it doesn't really make sense to test since any result you get is not going to be statistically significant (unless you run the test for years!). In a scenario like you mentioned, I would suggest doing your best to take these principles into consideration, and then using heat mapping (hotjar.com is a great tool), surveys, and other data to gauge how people are responding to the site, and make changes when appropriate.
Woahh, I didn't know about Hotjar! I juste installed it on one of my websites. I can't wait to see the results! Thanks so much Chris.
Hi Chris
Excellent tips on how to test a web page. I love the idea of not distracting the customer's attention with aspects of little importance. Sometimes, for wanting to make it flashy, what we do is create a design that does not favor conversions
Exactly! Design should never trump functionality. Thanks for the comment.
Great post. Do you know of any tools that can be utilized during the lading page development phase to help create a well optimized page?
For example, Yoast has a great SEO plugin for Wordpress that scans a page or post for key SEO elements in the editor, as it's being created. It generates a list of SEO improvements and gives a Green light when it is will optimized.
It would be really interesting if there was a similar tool for landing pages.
Thanks Luke, great question. I've actually been working on a conversion rate audit process that we hope to turn into a tool at some point in the near future.
Currently, there are a few tools that look for various things on your landing page, but unfortunately nothing that will just scan a landing page and give you great feedback about conversion rates. The best tool that I know of is from VWO (an a/b testing tool), that asks you a series of questions about your landing page to help you grade it: https://landingpageanalyzer.io/
Thanks for sharing. I'll check that out.
I'm building my new blog at the minute so this is great! I am going to follow it like a checklist and test it after launch. Thank you!
You bet! Good luck with the new blog.
Great article. Usually webmasters are trying to add diversions to their websites to show up how "complex" a company is and how many things they can offer. This often results in abandoned carts, confusion etc. Like in a painting, when the artist must know or feel when to stop to not mess it up, the same is with a website page. A webmaster needs to know that many information posted there does not necessary mean a lot of fans or conversions.
Great post I found it really useful. I totally agree that content should be clear, sometime customers get carried away and forget who they're actually writing for.
Muy interesante Cris!
Aunque lo leí así que parece fácil de llevar a cabo, no creo que sea tan fácil.
But I have taken many ideas that I will take into account and that will surely help me improve my conversion rate.
Very helpful article Chris, thank you! Just recently started with A/B testing and it is just stunning how the smallest changes in a sentence make the whole difference!
Hi , I have been reading on Moz for a year or so tried the trial version and countless other Seo tools and they all give my site a fairly good overall score and my competition a much lower score. But they still out rank me by a lot, from using the tools to study them, the only difference is they have more back links and that's it. So is that what is has come down to ? If you do not have the links you will not get the rank no matter if your site is of more use to the visitor? Thanks for any input and all the free help from Moz.
Useful article. Landing page testing is critical to improve our AdWords campaigns.
Great article and all things are explained in a very specific manner.
One more thing that I want to add here:
Motivation + Emotion = Perceived Benefit
Thank you so much..