What UX Designers Need to Know about Conversion Rate Optimization

Conversion is most often defined through sales, but it can also apply to clicks, sign-ups, repeat visitors, or any other metric that meets your organization’s goals.

The real problem many organizations face regarding conversion, is that content is often still considered “the stuff that goes into the design.” Putting content at center stage means changing some of the fundamental ways we think about content in the design process and how it helps conversion.

The principle of ‘form follows function’ states that the form of an object must be based on its intended purpose, and UX is no different. Since the primary function for any brand’s website is to engage, inform, and convert visitors into consumers, then the design of that website should be created primarily to support its content, or the “stuff” that has the greatest potential to engage, inform, and convert. The folks over at GatherContent, as well as several others, have begun calling this a “content first” approach.

From our viewpoint, content first doesn’t mean sacrificing design, or other elements of UX, but it does mean that content goals, which are often defined through business goals, should drive design. The elements of design—from the architecture and navigation to the look and feel to the code functionality and everything in between—are all components that come together to help the user reach their intended goals in the most efficient and enjoyable way possible.

Both CRO and UX should be about helping people reach their goals. That’s how you convert. You give people what they weren’t even sure they were looking for.

But how do you do that?

Strategy first

In our combined 15+ years experience working with clients in our respective industries (digital marketing and technical communication), we have been amazed with how many folks want to rush to design before they have content goals in place. Clients often begin conversations by talking about technologies they want us to use (blogging platforms, social media platforms, analytics platforms, mobile platforms, development languages, and the like), before we’ve even established what they’re trying to accomplish.

Here’s a representative comment from a recent client: “Well, I created a Google Analytics account several years ago and created an ad that pops up when you search for [specific keywords]. Well, now I get all kinds of people headed to my Facebook page from that ad. And those people are probably sorely disappointed when they find my Facebook page. Can you help me with that?”

There’s an old term in technical communication that refers to what happens when technological innovations force workers to re-skill: technoshock. Well, we call what clients do “technolock.” Clients lock into a particular platform that they’ve heard of or had positive experiences with and that platform becomes the strategy.

Developers can be guilty of this, too, however. We have had conversations with many developers who have a particular language or development platform they swear by, whether that’s Ruby, PHP, or a complete, open source CMS like WordPress or Drupal. Conversations with a developer who is experiencing technolock are very similar to those we experience with clients. The first step is to get them talking about strategy and not talking about technology.

Content goals are business goals

The best way we’ve found to get past technolock is to get everyone who has decision-making power over a design in a room together—developers, stakeholders, business managers, marketing specialists, UXers, content people, the works. Only then can you start to build a comprehensive strategy that is goal-directed, not technology-dictated.

UX strategy that keeps conversion as a key goal is about planning for the creation, publication, and governance of useful, usable content, content that converts. UX strategists must work to define not only what content will be featured but why it’s being featured it in the first place, before they decide where and how it should be featured.

This process typically starts with a detailed audit of the organization’s existing web presence, and even the smallest business or non-profit typically has at least a couple social media profiles.

This audit, however, is an audit for what is already working, or what converts. Such an audit often includes questions and methods like the following:

What are people responding positively to? If the organization is already interacting with users, the strategist should look at some of those interactions. What are people asking for? What is falling on deaf ears?

What data is readily available? Digging into qualitative or quantitative data that the organization has already been collecting, intentionally or less-than-intentionally, can render terrific insights into what they’ve been doing. Existing customers or clients of the organization are also great resources for discovering insights. Interviews with a representative sampling of them can uncover what excites them, what their goals are, and what they dislike about the organization.

What goals are apparent? While digging around in an organization’s web presence—whether that involves doing some simple organic searches to see what comes up, doing formal keyword research, interviewing key stakeholders who maintain particular platforms, or doing a formal content audit of an existing website—strategists should pay heed to what it seems like the organization is trying to accomplish. We have often found that when we ask organizational leaders what their goals are, they differ strongly from the apparent goals demonstrated by their organization’s existing web presence.

The bottom line for any digital effort undertaken by an organization is return on investment (ROI), and UX strategy is no exception. The expectation is that developing and adopting a comprehensive UX strategy that centers on CRO will create a benefit, either through increased revenue or operational savings.

Know your users so you can empathize with them

Another thing clients often want us to do is jump right to the conversion stage, which can also be a mistake. They want to know what will convert before we’ve interacted with a single user. Understanding users’ needs, wants, hopes, dreams, aspirations, attitudes, and goals relative to the brand is essential to converting these users to customers. Interviewing users directly is the best way of understanding these needs.

We like to combine interviews with usability testing and other methods. This approach goes something like this:

1. Complete audit of organization’s existing web presence.
2. Work with organization to define business goals and articulate how these goals connect to UX. The question we always try to answer at this point is: what kind of experience should users have, given the goals of the organization?
3. Create quick prototypes to test with or test existing prototypes. These tests should be conducted in a context that is as close to the users’ real world context as possible.
4. Ask qualitative questions of users at the beginning and end of these tests. Start by asking them about who they are, end by asking them about who they think the brand is.
5. Create customer personas based on user research data and content audit.
6. Redesign and retest as needed.

Observing how users perform tasks on a prototype is one of the most valuable forms of research we’ve conducted. Users are not always good at articulating what would convert them, so it’s better to pair interviews with tests to not only gather usability data, but also to elicit better feedback from users.

Finally, creating personas to represent each segment of a user population can help connect business goals to UX to CRO. They bring the whole strategy together. A persona is based on primary research that uncovers the real attitudes, goals, and behaviors of the users it represents. To turn research into a persona, give the persona a real sounding name and a face, and write a description of him or her that includes details uncovered during user research.

These personas can also be used to stand in for actual users later on in the process. We highly recommend having developers adopt one of the personas and do a walk-through of a prototype before committing the prototype to live code. This helps developers empathize with the users they’re designing for.

As needed equals it will be needed

Steps 1-5 can happen before a live site or redesign is fully deployed, but there are several activities post-launch that are just as important as these pre-launch activities, if not more important. One of the culprits for a lackluster user experience that doesn’t convert is what we like to call “self-centered design.” Organizations often lack a sharp, research-based understanding of their target customers. Without this information, decision makers advocate for things that they personally like.

Even UXers, developers, and marketers may have different opinions about what should be improved on a page. The only way to improve landing pages that have already been launched is to run statistically-valid tests of new user experiences, designs, and content variations. A/B split testing evaluates the differences in the visitor’s reactions between a limited set of page executions (frequently just two, A and B). By running controlled A/B split tests on the live site, organizations can get real data from live visitors and know with certainty which page variation would maximize conversions.

To optimize for multi-screen experiences, you not only have to control for screen size but also how users react to various screen sizes. Content can look, feel, and thus convert very differently on different screen sizes. Further, by multi-screen, we mean that UXers should take into consideration the ways users construct their own touch points with organizations across screens. Each screen size has its own unique affordances, and thus constant, iterative A/B testing must be performed post-launch so that CRO is continuously improved rather than plateauing.

Recommendations for getting started with conversion rate optimization

One of the first tools we recommend to clients is Google Analytics. As soon as a site is launched, or even if a site was launched but never tracked, GA is the first thing we do with clients. It’s just too usable, powerful, and simple to pass up. And it’s free, of course.

Next, we often help clients choose a testing platform. Unbounce is expensive, but effective. Optimizely is inexpensive and easy to use. Visual Website Optimizer is relatively inexpensive. Google Analytics has a testing component to it called Content Experiments. Some other tools we’ve used are Keyword Eye, UberSearch, and Crazy Egg. The point is, there are a number of relatively inexpensive tools out there that you can help clients with or use yourself to generate useful data.

But for many of the clients we work with—small business owners or non-profits—it really comes back to the why, the strategy. After a site is launched, after all, clients usually experience a lot of benefits, such as an increase in traffic, more sales, lots of positive vibes coming from their customers. Or the opposite may be the case. Sometimes we encounter clients whose honeymoon period with their site is long over, and now they’d rather just scrap the whole thing and start over.

Regardless of the directions strategists guide clients, you can’t do real CRO without analytics. At the same time, sometimes small business owners don’t have the traffic for robust, systematic optimization to be worth their time. For example, if a restaurant owner only uses their website for posting the menu and taking reservations, but not offering online ordering, it is not worth their time or money to optimize constantly. On the other hand, for a corporate restaurant site that has 5,000 franchises and offers online ordering, then it makes sense.

Does it never end?

Clients often ask us some version of: “Does the work of doing this ever end? Will I ever get to a point where I don’t have to keep improving my web presence?”

Our answer, of course, is no. Just because a solution worked today, or this quarter, or this year, doesn’t mean it’ll work tomorrow. Users’ behaviors and habits change. If Google changes their algorithm, suddenly the traffic from organic search is different.

We also encourage clients to experiment on their own, however. We put it this way with clients: if you were going to run a newspaper advertisement in every newspaper in America, it would be a big deal if you accidentally included a typo. But how big a deal would it be if you were just running it in your local paper?

No one action, outside of deleting a domain or having repeated tirades on your blog, can permanently damage CRO. If anything, more content is probably not going to drive someone away, even if the experiences associated with that content are less than optimal.

If your product does not change, your offer does not change, and for the most part people do not change, then once you have developed this kind of flow state that marketers call the “conversion funnel,” that’s a good time not to experiment. If it ain’t broke… you know the rest. Even in this instance, the client can try going further downstream (e.g., toward smaller scale interactions with existing customers on social media) or further upstream (e.g., toward a more expansive social media presence that tries to draw new customers in).

Clients will typically want to start somewhere in the middle with a shopping cart or comparable experience. Once the cart is converting, it may be time to go upstream and try to get more people into the shopping cart. And once that happens, they might try going farther upstream with paid search, direct mail, or banner ads. Once users have purchased, they might start testing within email with retargeting ads that convert customers to repeat customers.

We try to teach our clients that CRO is kind of like a challenging game of customer psychology. You’ll never fully understand what every user is thinking, and how a given user becomes a customer, but you can observe their actions and extrapolate a strategy from it.

This strategy needs to be a living thing that grows with your CRO efforts, not against them.

Further reading

Ash, T., Page, R. & Ginty, M. (2012). Landing page optimization: The definitive guide to testing and tuning for conversations. Indianapolis, IN: John Wiley & Sons, Inc.

Bailie, R. A. (2011). What’s the buzz about content strategy? Bulletin of the American Society for Information Science and Technology (Online), 37(2), 19-22.

Cummings, M. (2010, August 16). User Experience. UX Design.

Gofman, A., Moskowitz, H. R., & Mets, T. (2009). Integrating science into web design: Consumer-driven website optimization. The Journal of Consumer Marketing, 26(4), 286-298.

Goward, C. (2012, March 2). Landing page optimization for more profitable experiences. UX Magazine, 797.

Gualtieri, M. (2009, September 4). Best Practices in User Experience (UX) Design: Design compelling user experiences to wow your customers. [PDF] Forrester Research.

Halvorson, K. (2008, December 16). The discipline of content strategy. A List Apart.

The Institute of Direct and Digital Marketing (2013, December 8).

Kotler, P. & Armstrong, G. (2010). Principles of Marketing. Upper Saddle River, New Jersey: Prentice Hall.

Krug, S. (2006). Don’t make me think: A common sense approach to web usability. Berkeley, CA: New Riders.

Krug, S. (2010). Rocket surgery made easy: The do-it-yourself guide to finding and fixing usability problems. Berkeley, CA: New Riders.

Law, E. & van Schaik, P. (2010). Modelling user experience – An agenda for research and practice. Interacting with Computers, 22 (5), 313-322.

Nielsen, J. (2007). Long vs. Short Articles as Content Strategy. Nielsen Norman Group.

Strom, D. (2013, November 27). A/B tests: Cut the fluff and spend the pixels on design elements of interest to users. IT World.

Siroker, D. (2010, November 29). How Obama Raised $60 Million by Running a Simple Experiment [Weblog comment]. Optimizely.

Thirumalai, S. & Sinha, K. (2011). Customization of the online purchase process in electronic retailing and customer satisfaction: An online field study. Journal of Operations Management, 29(5), 477-487.