What Is Conversion Rate Optimization?
Imagine you’re trying to sell your chief marketing officer on SEO, but he’s hesitant. “I understand SEO has the potential to drive a lot of traffic, but how do we know it’s the right kind of traffic?” he asks. “How do we know that traffic will turn into revenue? The CEO doesn’t give me high-fives for more traffic, she wants to see marketing directly tied to revenue. SEO seems like a lot of guesswork–we need science.”
If you’ve ever had an exchange like this with an executive, a client, or within your own head, then conversion rate optimization (CRO) is the answer. Not because CRO replaces SEO (it doesn’t), nor because it necessarily changes SEO (although it can help), but because while much of SEO focuses on generating traffic, CRO turns that traffic into leads and sales, or what we call conversions. And it doesn’t do it through guesswork, but rather through a methodical, scientific, straightforward process that virtually guarantees ever improving results.
“Conversion Rate Optimization (CRO) is the art and science of getting people to act once they arrive on your website,” says Tim Ash, CEO of SiteTuners, author of the book Landing Page Optimization: The Definitive Guide to Testing and Tuning for Conversions, and founder of the international Conversion Conferenceevent series. “It typically involves elements of visual design, copywriting, user experience, psychology, testing out different versions of your website content, and the neuromarketing to influence people to act.”
If you’ve ever heard people talk about A/B split testing, or changing the colors of buttons on a website to see which one converts better, then you already know something about CRO. But that’s kind of like saying if you’ve ever driven a car then you know something about the auto industry. There’s more behind the scenes. In this blog post Andy Crestodina, author of Content Chemistry: An Illustrated Handbook for Content Marketing and Founder of Orbit Media Studios, a digital marketing agency, shows how far you can take something as simple as button design when it comes to CRO. CRO is a term that includes everything you do with your marketing and website that influences conversions. It’s scientific and straightforward, but is also complex, because people are complex, and it’s people CRO seeks to influence.
1. Persona research. A “persona” is a fictitious profile of a potential visitor to your website, or potential customer of your services. For example, a simple persona for a jewelry website in New York might look like this:
Sam is 28. She lives in New York. She’s single. She has an undergraduate degree in journalism from Columbia, and an MBA from Harvard. She works as an associate for a private equity firm. She has a 40 minute commute each day on the subway. She enjoys triathlons and vegan cooking–she’s a health nut. Oh, and she’s been set up on a blind date, which is coming up in three days, and she wants some jewelry she can wear to it.
“Buyer personas help to break down the complexity into insightful understanding, allowing for improved decision-making,” says Tony Zambito, a leading authority in buyer insights and buyer personas. In a 2013 blog post, Zambito explains how various factors, including globalization, make the creation a buyer personas a more important practice than ever. As he says, “Companies are entering new markets with little knowledge of regional buying behaviors and little room for errors.”
Personas are important to CRO because they influence how we design the website, the content we create for the website, and what keywords we target for SEO purposes, amongst other factors.
2. User journey scenarios. Going back to Sam, our buyer persona from above, we can ask how will Sam and others find the aforementioned jewelry website? What will Sam do once she’s on the website, based on what we know from her persona? What information is she looking for? What is she likely to click on? What path will she take through the website before making a purchase? Asking and answering these questions are the basics of creating user journey scenarios. Ralph Wolbrink, co-founder at Pure Internet Marketing, explains in this blog postthat customer journey mapping is essential because it aids in “understanding the customer better and providing value where they require it.” Like buyer personas, mapping out user journey scenarios will help you figure out the more obvious changes to make to your website as part of your CRO efforts.
3. Focus groups. Although there are various opinions, when used correctly qualitative research like focus groups can add another point of reference and provoke valuable hypotheses for testing. “By all means, incorporate qualitative methods into your marketing system,” says Chris Goward, Founder of the optimization firm WiderFunnel and author of You Should Test That: Conversion Optimization for More Leads, Sales and Profit or The Art and Science of Optimized Marketing. “Then, use the input from them to generate better hypotheses for split testing. The scientific method of marketing starts with formulating questions to ask of your visitors. If you can use these qualitative studies to help you develop better hypotheses, that’s great!”
4. User surveying. When the social media management tool Buffer decided to redesign their homepage, part of their process involved user surveys. “We did a lot of user surveying and found that ‘saving time’ and ‘letting me focus more on work’ are actually two really key reasons that people love Buffer,” says Brian Lovin, who worked on the site redesign. Buffer didn’t base their entire redesign on the results of these surveys, but as with the other parts of CRO listed above, it provided one more viewpoint and allowed them to come up with experiments they could perform to get hard data. When the new site launched, Buffer saw a 16% increase in overall new user conversion.
5. Demographic and psychographic analysis & targeting. It may be obvious how demographics affect CRO–if you’re marketing to someone in Hong Kong you’re going to use different messaging on your website than what you would use to market to someone in the US. Psychographics goes further, in that (as defined by Wikipedia) it is “the study of personality, values, opinions, attitudes, interests, and lifestyles.” As Maggie Hibma of Hubspot says, “Demographics explain ‘who’ your buyer is, while psychographics explain ‘why’ they buy.” Hibma says you can collect psychographic data in a number of ways:
- Marketing data analysis
- Sales and support teams
- Social media
- Focus groups
The better you use these methods to understand why your customers buy, the better you’ll be able to engage in CRO to help customers get what they want.
6. Web psychology. Nathalie Nahai, author of Webs of Influence: The Secret Strategies That Make Us Click, came up with the term in 2011 to describe empirical studies of online behavior. “It’s a point of convergence for any type of research that looks at online behavior,” she says, “So things like human-computer interaction, cross-cultural psychology…social psychology…advances in neuroscience, how we can use neuroscientific studies to tell us about how we respond online at a brain-activity level…” Easy enough, right? Or as she then explains, “All of these different disciplines and many more give us different glimpses into how online behaviors are shaped and can be affected.”
As Nahai breaks it down in this online video at Moz, the three secrets of marketing online are:
- Know who you are targeting.
- Communicate persuasively.
- Sell with integrity.
Studying the psychology of your online audience will help you know your customers better and understand how to communicate effectively with them, adding value to your CRO program.
7. A/B split testing. Pick a webpage on your website. Make a copy of it. Leave Copy A as it is. On Copy B, change one thing, such as the wording of a headline, the color of a button, or the placement of an image. Direct 50% of the traffic that normally goes to that webpage to Copy A, and 50% to Copy B. Let the test run long enough to generate a sufficient sample size. Analyze results. Keep the copy that performs better, and drop the other one. Repeat.
This is A/B split testing in a nutshell. The point is that as you continue to run tests, your pages will convert more and more of your traffic. You may not be able to predict exactly how much better, but you know the results will improve, because while there is guesswork involved in creating the experiments, there’s no guesswork involved in choosing the page that performs better.
When people start learning about CRO, A/B split testing is generally what they latch onto. It’s easy to understand how it works, and the value in it. My only word of caution is to not mistake A/B testing as being all you need for a successful CRO program. A/B testing is sufficient for many websites, but it does have limitations. For high-traffic websites with more time and budget to put into CRO, multivariate testing may produce additional benefits.
8. Multivariate testing. In many scientific experiments it’s important to change one variable, and one variable only, in order to be able to effectively measure results. Otherwise you don’t know which change that you made in the experiment led to the change in results. Multivariate testing deliberately changes multiple variables at once to test various combinations. A simple example is changing the primary photo on a homepage as well as the large, bold text over that photo. Multivariate testing gives you more data than simple A/B testing, in which one variable is changed each time. If you were to run an A/B test on your homepage by changing the photo, and you saw an increase in conversions, and then you ran a separate A/B test to try out a new headline on the homepage, and you also saw an increase in conversions, you might assume that by putting the new photo and new headline together, you would see even more conversions. But what if you put them together and the result was lower conversions than you were getting before you started testing? Multivariate testing helps you avoid jumping to incorrect conclusions like this.
The catch with multivariate testing is the need for data–lots of it. “It’s not about visitors,” Tim Ash says, “The limiting factor is the number of actual conversion events. I would not recommend multivariate without 100 conversions per day minimum.”
9. User testing. Have you ever watched someone use your website? It can be eye-opening. What you find simple and intuitive may be confusing to your customers. You could set up your own user testing study to get more data, or use a service like User Testing to do it for you. If you’d like to get a laugh along with your user testing, you could hire web design expert and founder ofTheUserIsDrunk.com, Richard Littauer, who says “Your website should be so simple, a drunk person could use it.” For $500, Littauer takes that statement to its literal conclusion by getting drunk and testing your website. The results are not only entertaining, but useful as well. Here’s his review of HubSpot’s website:
10. Usability testing. User testing and usability testing might sound similar, but they’re different things, as pointed out in this post by Shari Thurow, author of perhaps the best foundational SEO book out there–When Search Meets Web Usability. User testing is when normal people try out your website and tell you what they liked or didn’t like about it. Usability testing is when usability experts run expectancy, closed card sort, brand perception, and other tests on your website. Both are useful, and there can be some overlap, but don’t confuse one for the other.
11. Eye-tracking and heatmap analysis. Tools that help you see what users are looking at and clicking on can give you valuable insights into how small changes on your website can improve conversions. In this post by Gregory Ciotti, who runs marketing for Help Scout, he shares seven case studies based on eye-tracking analysis. In one example, it’s shown how people focus on the face of a baby when its photo shows the baby staring back, but when the photo is change to a baby looking to the side, website visitors focus on what the baby appears to be looking at, which happens to be what the website owners want people to click on.
12. Analytics analysis. Right there in Google Analytics is a gold mine of data related to CRO. Not only can you track conversions within the software, but all the data leading up to a conversion. Perhaps more importantly, you can track the data that doesn’t lead to conversions, and figure out what you need to change.
These 12 areas contribute to CRO, but none of them by itself is CRO. Optimizing your website for conversions is an ongoing process that includes all these activities and more. Finding the right place to start will depend on what you can identify as the low hanging fruit, your current objectives, and what kind of budget you have both in terms of time and money. The key is to get started today. The sooner you start applying a structured, data-driven approach to your digital marketing, the sooner you’ll reap the rewards.