Checklists, A/B testing, user testing, analytics, heat maps, feedback, guerilla usability testing, eye tracking, graphic design testing, user testing according to Steve Krug. Does this sound interesting to you? If so, I have a great news for you! You will learn about all these tools and methods and much more in this fourth and final part of wed design process.
Phase No.4: Evaluation
Evaluation phase is inseparable part of professional web design process. It doesn’t matter how great results of your work. What matters is whether the result achieves business goals you and your clients set for the project or not. This is what evaluation phase is about. Its overall goal is to provide you with feedback and show you potential weaknesses of your solution – the website you deigned. Then, you use this feedback to fix these weaknesses and any gaps and make your solution flawless. Feedback provided by evaluation phase will also help you train your intuition.
You can do evaluation in different forms and on different levels. In all depends on how far are you in web deign process. Another important condition is what do you want to test for. You should never try test more than one factor or metric at the time. For example, let’s say you want to test usability of the website and also its persuasiveness. The worst thing to do is running both test at once. There is Russian proverb: “Person who chases two rabbits catches neither.” Another way to put it is that if you focus on more than one thing at the time, you focus on nothing.
Have you ever tried to focus the sun rays with one magnifying glass on two spots? Probably not. Don’t try to do the same thing here. Choose one test, run it, evaluate it. Only then move to the test number two and repeat the process. Another mistake is thinking that you can do testing with prototype. Sure, there will be times and types of test when prototype will be sufficient. However, this will be mostly about testing your hypothesis you do in earlier phases of web design process.
Depending on the number of people you have available, you can either do quantitative or qualitative testing. Quantitative testing will work best if you have a large sample of people. On the other hand, qualitative testing works with small number of people. Whatever the case is, make sure to choose the right people – people from the target segment. Otherwise, no matter how much feedback will you get, it will be useless.
There are two more things I should mention. First is that evaluation includes methods that can be used online and also methods that can be done offline. Keep this in mind while choosing the method you will use. Second, don’t forget to think about any limitations or restrictions your project might have. It is possible that these restrictions will influence what you can test and what you can’t. Keep these things on mind in order to avoid any potential problems.
The process of user testing is quite simple. First, you give people specific tasks to do and on the website. Second, these people will get these tasks done while commenting them out loud. Your job is to watch and listen. One of the relative downsides of user testing is that you already have to have something at hand. You can’t user test vacuum or ether. Well, you can try to test wireframes or sketches. However, only (at least partially working) prototype filled with content will provide you with any meaningful feedback.
Another relative downside is that user testing will help you find only certain amount of bugs and problems. Even this small number of problems is still much better than doing no testing at all.
The benefit of user testing is that it can support your findings with rock-solid evidence and convince your client about these findings. It is one thing to propose changes or adjustment if you have nothing to justify them. The situation will get whole different if you can show your client some proof provided directly by real users. User testing can also improve how programmers see users of the website. The last thing is that user testing can help you prove your client that your web design process is actually working and his money are well spent.
Guerilla Usability Testing
Let me tell you one relatively cheap and agile method of usability testing. This method is called guerilla usability testing. In order to get it done, you need to write down basic scenarios. Next, you will go to some coffee shop, book store, library or just any place with people and something to sit on. Finally, ask random girl or guy to spend next 15 minutes with you because you need some help with your website. You can also invite the person on coffee or tea. So, at worst, it will not be a complete waste of time for her or him. Don’t forget to take a lot of notes.
The benefit of guerilla usability testing is that it is blazing fast and also cheap (depending on the place). This makes guerilla usability testing very good tool in case your web design process is restricted in time and / or money. The downside is that it requires certain amount of assertiveness and courage. This method works best in projects where you need to test the general usability rather than specific expertise or knowledge. On the other hand, if you are working on a project for specific group of people that depend on some expertise, choose different method.
Let me give you an example. Right now, I’m working on UI for accounting web app (do you know intuit?) that will be used by professional accountants. In this case, it would be waste of time for me and my client to sit in coffee shop and let random people play with the website. The majority of the people around will probably not have adequate knowledge to use the website. These people are also not the target segment of user – accounts. In this case, guerilla usability testing is useless.
Steve Krug’s User Testing
User testing developed by Steven Krug is another fast and easy method to test your website. It will also deliver tangible results without spending a fortune on equipment. The whole process consists of couple steps. First, you create a variety of scenarios. Second, you prepare either complete website or working prototype. Third, you will find three to five people. Isn’t five people too low? No, not at all. According to study done by Jakob Nielsen, having group of just five people is enough find almost as many usability problems as you’d find using much larger group.
There are also exception to this rule. The same study states that for quantitative studies (statistics) you will need at least 20 people. If you are using card sorting, we discussed this method in Web Design Process Pt2–User research, the number to aim for is 15. And, for eye tracking, it is approximately 39 people.
The fourth step is to give the people concrete task to do. Every testing requires another two people from your team. The first person will assign the tasks and control the testing. The second person will take notes. I suggest that the creator of the test is also the person taking notes. Otherwise, he or she might have inclination to help or guide participants. There is also possibility that creator of test in position of note taker will be able to focus on and gain more insights. In examples of web design process, user testing is recorded for potential future use.
I have to admit that I’ve never been present during recorded testing session. I have no experiences with it. One of the arguments for recording the testing is to present it to your client. True to be told, clients will only very rarely watch these videos. That doesn’t mean these records are completely useless. You can use these records to support yourself if someone from the team will take a stand against your changes and adjustments. How often should you test? Do tests during your web design process often and start soon. This will help you find problems and fix them fast.
Over the course of usability testing, you find whether the website is usable or not. You will see if the UI is easy to understand, if people can orient in the UI and use it. However, there are other questions you need to answer. For example, does the website help visitors achieve their goals? Does it help in their work? Are people more productive? In other words, is the website useful? This type of testing is done in a similar fashion such as usability testing. The only difference is that now you will also use in-depth interviews and look for information about how he or she sees the UI in relation to his or her work.
It is quite possible that you and your client will have different opinions on the design. That’s why there is this method called desirability testing. The goal of this method is to test what effect the design has on visitors. This method can be quite useful when you have more than one version of website design. There are two approaches to this method. First is using card sorting and ask people to assign cards with specific terms according to what they feel. Second option is to use some kind of scale for rating their experience, instead of cards.
The last method for user testing, and evaluation phase of web design process, is eye tracking. This method is based on testing with use of eye camera. The downside of this method is that you need to have laboratory setting, professional equipment and large sample of people. It is quite possible that you will never be able to use eye tracking in user testing in you work. Still, it is better to know that method such as this one exists. Reading couple eye tracking studies will be beneficial for your web design process as well.
When we talk about checklists, we often think about to-do lists. In web design, checklists are tool that can provide web designer with feedback and unbiased view on the website. The thing is that you can’t argue with checklists. It is a list of points you can directly test. This makes checklists objective tool that ignores your feelings and assumptions. Your website either pass the point (condition) or not. One of my mentors I had, and still have, on my journey in web design recommended that I use three specific checklists.
The first checklist is for my web design process. This checklist keeps me on track and helps me remember everything I have to do. Otherwise, I may skip or forget some step and risk failure of the project. I suggest that you do the same thing. Write down all steps of your web design process or the one you are currently reading about so you will have all information on one place. With time and practice, this checklist will become more or less automated part of your work. Even then, you can still use this to double-check yourself.
The second checklist is focused on methods. You can create a checklist for the methods you use regularly in your web design process. However, this checklist will be much useful if you focus on methods you don’t use as often. It is these methods where your memory will not be at the best condition all the time. The less you use some method, the more probable it is that you will forget something related to it. This can’t be said about methods you are using almost daily. So, start with methods you don’t usually use and only then move to more familiar ones.
The third checklist is for procedure you will go through before and right after launching the website. This checklist can include points such as checking search friendly URL, working forms, favicons, configuring web analytics and so on. To give you a better idea of how this checklist may look like, take a look at these 45 checklist examples collected by Smashing Magazine. You can also use online Web Designers Checklist. Remember that checklists are tool for perfectionists. Meaning, they will push you to always deliver excellent results. Avoid them at your own risk.
Good feedback is the engine of good web design. Now I don’t mean only feedback provided by visitors of the website. I’m also referring to the feedback provided by other professional web designers and technical supervision. The key is to always look for someone who is either on the same level of expertise like you or better. Never go in opposite direction. If you are can’t ask for feedback someone in a team, you can search for other web designers and web design communities and discussion forums on the Internet.
One of the benefits of expert feedback is that it keeps your ego in check. Otherwise, there is a certain chance you could see your design as flawless. Trust me, I’ve been there as well. You can also ask for feedback your clients, but the only opinion you will often get will be either “I like it” or “I don’t like it.” This is useless. You need to know the reasons behind the opinion. Expert feedback will also help you polish your web design process and web design skills in general.
Another method or tool to get feedback on the result of your work is A/B testing. This method is about testing two or more alternative variants of the website to see which variant results in higher conversions. These alternatives are often created as a result of user research. These variants are then shown to sufficient group of visitors. The variant will bring the highest number of conversions wins. Contrary to what you may read about, it is not so beneficial to do A/B testing just to test random colours for buttons. It is not worth the time and expenses.
Heat maps are great tool that can help you understand what visitors are doing on the website and evaluate results of your web design process. Tools such as Google Analytics will show you how are people moving through the website. Heat maps will show you exactly where people are clicking and what are they doing. You can then use these information to improve your websites to show your visitors what they really want to see and increase conversions. Couple tools you can try are crazyegg, hotjar and SumoMe.
Closing Thoughts on Having a Web Design Process Pt4
Evaluation has a specific place in web design process. It helps you systematically test the results of your work. Evaluation will also help you train and improve your intuition. However, that doesn’t mean you should drop this phase after certain time. You should always use at least one method that will get you in touch with real users. From this point of view, checklists and expert feedback are only supplemental methods. Remember that evaluation is always cheaper than implementing something that doesn’t work.
Thank you for your time.
All parts of the Web Design Process series:
Web Design Process Pt4-Evaluation