December 2, 2016
Conversion rate optimization best practices
Conversion rate optimization is a process. It’s not one-off tests based on hunches, guesses, random tactics, or any consultant’s voodoo foreknowledge. The process starts with looking for opportunities. You sniff around for things that are glaringly wrong with your user experience and that palpably feel out of place. Optimizers call this heuristic analysis or heuristic evaluation. In this post, I’ll give you a rundown of three common heuristic analysis frameworks.
- Nielsen Norman Group’s Heuristic Evaluation
- Marketing Experiments’ Model
- LIFT model
Executing quality heuristic analysis will give you an endless roadmap of conversion optimization opportunities, and put you miles ahead of other companies who are A/B testing things at random.
3 common heuristic analysis frameworks explained
Nielsen Norman Group’s Heuristic Evaluation
NN/g is a powerhouse in usability and user experience information. They’ve outlined 10 steps to consider in a heuristic analysis: 1. Visibility of system status Users don’t like to feel lost or confused. Try to keep them aware of where they are and appropriate next steps through clear navigation, breadcrumbs, and user onboarding messaging. To use NN/g’s words, “The system should always keep users informed about what is going on, through appropriate feedback within reasonable time.” 2. Match between system and the real world Your UI should match the user’s language and expectations. Don’t bog users down with jargon and hard-to-understand language. Always try to make information appear in a natural way. Voice of Customer research complements this and helps you hone in on that messaging. Chubbies knows how its customers talk: 3. User control and freedom NN/g puts it well: “Users often choose system functions by mistake and will need a clearly marked “emergency exit” to leave the unwanted state without having to go through an extended dialogue.” So give users an adequate amount of control over navigation and mistake correction. An example of violating this heuristic…using auto-rotating sliders with no way to control the speed or direction: 4. Consistency and standards Do you use different words that users are used to? How about different words across your own interface? Always opt for consistency to limit your language’s cognitive load. 5. Error prevention There’s a whole art and science to error messaging, because no matter how great the design, users will make mistakes. Optimize to minimize them, of course, but make sure you’re also following error message best practices.
6. Recognition rather than recall According to NN/g, “The user should not have to remember information from one part of the dialogue to another.” The instructions, instead, should be visible or easily retrievable. In practice, that could mean the use of accordions, sticky menus, and changing the color of links that have been clicked. 7. Flexibility and efficiency of use As NN/g says, “allow users to tailor frequent actions.” This means offering a limited (but relevant) user flow, personalizing CTAs for returning visitors, and tailoring the experience to make it easier to complete frequent actions. 8. Aesthetic and minimalist design While minimalism is a design trend, it’s also heavily supported by research as a good usability/conversion heuristic. According to NN/g, “every extra unit of information in a dialogue competes with the relevant units of information and diminishes their relative visibility.” Translation: if it’s not helping the user experience, it’s probably hurting it. 9. Help users recognize, diagnose, and recover from errors A previous step outlined the need to minimize errors. This step is the follow up. NN/g claims, “error messages should be expressed in plain language (no codes), precisely indicate the problem, and constructively suggest a solution.” I wrote a whole post on common error message mistakes if you want to read that, but here’s the tl;dr:
- Don’t blame the user.
- Write like a human, not a robot.
- Make sure errors are clear, and the messages are positioned in a place that is intuitive.
- Make sure users know how to fix said errors.
- Don’t list all errors at the top of the page. Inline validation is a good solution.
Of course, enact form analytics, conduct user testing, and watch session replay videos to get the data about your own site, but the above best practices will get you further than most CROs.
10. Help and documentation NN/g: “Even though it is better if the system can be used without documentation, it may be necessary to provide help and documentation.” This is where FAQ pages, live chat options, pricing table, and microcopy come in. “Help and documentation” empowers users to accomplish their goals on your website and in your product.
Marketing Experiments’ Model
This next model is more conversion optimization specific. It was developed by Marketing Experiments, and goes like this: Looks like some complicated stuff, but it’s really a weighted system of heuristics to analyze on a given site. Here’s what the variables mean: C = Probability of conversion m = Motivation of user (when) v = Clarity of the value proposition (why) i = Incentive to take action f = Friction elements of process a = Anxiety about entering information In essence, the formula says that the probability of a conversion is a function of the match between the offer and visitor motivation + clarity of value proposition + (incentives to take action now – friction) – anxiety. Then there are multipliers to weigh importance (motivation, as you can see, is the highest). Friction is defined as a psychological resistance to a given element in the sales or signup process. Anxiety is a psychological concern stimulated by a given element in the sales or signup process. Reduce these as much as possible and do what you can to increase the users’ motivation and incentive, and to clarify the value proposition.
The LIFT model, by WiderFunnel, is similar but comes with a nifty graphic: Your value proposition is the basis of this framework. Relevance and clarity boost conversions, while anxiety and distraction kill it. Urgency is what propels people to take action right away. The nice thing about this framework is it’s easy to remember and the visual makes it clear what you’re looking for. Does this social media icon distract from the product page? Does this form need some microcopy to add clarity to the process? You can build up a solid list of opportunity areas by checking in on these heuristics.
What heuristic analysis looks like in practice
There are more models. These are just the tip of the iceberg, but I think you get the point: they’re standardized, repeatable lenses by which you and your team can analyze a site to get some preliminary ideas on optimization. Jeff Sauro has written some great things on heuristic analysis if you’re still curious about the practice. At ConversionXL, where I work, we don’t have a branded process for a heuristic analysis. Rather, it’s a hybrid of all the above and changes based on the type of website and its goals. What does the structured website review look like? We assess each page for a certain set of criteria:
- Relevancy: Does the page meet user expectation—both in terms of content and design? How can it match what they want even more?
- Clarity: Is the content/offer on this page as clear as possible? How can we make it clearer, simpler?
- Value: Is it communicating value to the user? Can we do better? Can we increase user motivation?
- Friction: What on this page is causing doubts, hesitations and uncertainties? What makes the process difficult? How can we simplify? We can’t reduce friction entirely, we can only minimize it.
- Distraction: What’s on the page that is not helping the user take action? Is anything unnecessarily drawing attention? If it’s not motivation, it’s friction—and thus it might be a good idea to get rid of it.
What heuristic analysis does not do
Heuristic analysis doesn’t tell you the solution to a problem. It also doesn’t tell you how important a problem is. Customer research helps you find out the former, and digital analytics helps you find out the latter. Heuristic analysis stands alone as an _opportunity finder_—something that lets you understand the landscape of a website better and holistically understand what you’re working with. So after you run a heuristic analysis, you should have a list of opportunity areas. Then you’ll want to conduct:
- Technical analysis
- Digital analytics analysis
- Mouse tracking/eye-tracking analysis
- User testing
- Qualitative surveys
Basically, the works. Combine these sources of data, and you’ll start to answer some formative questions:
- Who is my customer?
- What are their desires, fears, doubts, hesitations?
- Why would they buy from me instead of the competition?
- Where are they dropping off on my site?
- What are the fundamental usability problems I can fix right away?
And on and on and on. In other words, it gives you the ammunition to build and prioritize a highly effective A/B testing roadmap. You’ll never run out of testing ideas, and you’ll be miles ahead of other optimizers testing things at random.
A heuristic analysis is a rigorous process when done right. You set the variables by which you’re judging a site, get multiple parties to commit to the walkthrough, and judge based on the pre-established criteria. Most importantly, you’re not making any decisions with this data alone. You’ll then commit to digging through digital analytics, setting up customer surveys, conducting user testing, and running through proper conversion research. Then maybe you’ll up your A/B test win-rate and minimize regret. Which of these heuristic analysis frameworks matches your in-house process? Any insights you’d add to finding CRO opportunities? Let us know in the comments.