Are people actually using your website correctly?
I’m a builder by nature. I’ve always created things and inventions and not once have I ever made something without a purpose. I always created with a goal in mind for the end user. This is why we create anything really isn’t it? We want to build something that is useful and provides a positive outcome.
Otherwise, we’d just be wasting our own time.
A website is no different. We create them for a specific usable goal. Sure, I’d like to believe they are art and that I’m painting the Mona Lisa. But even great art has a goal and strives to develop a reaction from the user. Organizations, brands & businesses may have different reasons for building a website. A local retailer may decide to open a digital store to start selling to a global audience, a brand may want to connect and attract a larger market and a school may want an easier way to communicate with parents.
There are many reasons as to why we build a website.
But at no times is it to say I want to spend alot of money and build something pretty.
So we spend alot of money and time building something we believe will generate the results that we want. We look to our competitors and other similar websites in our niche. We critique what others have done and seek an agency whose portfolio best meet our needs. The next 3 months are spent gathering information and developing a large scale project that we hope will generate the results that we seek.
And then we launch the website and hope for the best.
But there’s that little tiny dark space in the back of our mind wondering is it actually working as it should?
Sure, we can point to the fact that people are going to the website and you may have lead requests or even online purchases or reservations.
But could it be better?
That really is the question we ask ourselves. Whether we generate 1 customer a month, 3,000 or have a conversion rate of 8% or 25%. Do we really know how people are navigating our websites?
We ask ourselves this question because we don’t know how to properly do online user testing. It may sound a bit creepy but we all have a tiny desire to sit behind all of our users as they navigate our website. We want to clap with joy when they look to the details of our wares and products. We want to scream at them to click the button they search for on 3 different pages.
But really we just want to test what our web users are doing so we can provide a better experience for them.
While we can’t legally nor logistically sit behind all of our customers as they browse through our website, we do have access to some really good tools for user website testing. These data-driven services allows us to actually see what the user experiences without the creep factor. We can use these tools to optimize and provide a better user experience so that we can achieve our digital goals.
1. Google Analytics: the baseline for user testing
This is the stalwart of the group. Google bought the analytics software company Urchin and launched this service in 2005. It is an out of the box solution that is good to go as soon as you register. You will need to install a line of code to your website but there’s a good chance your website cms or framework has a plugin for it. It is the single most popular analytics service on the market. It provides the most fundamental of metrics such as how many visitors came to your website and what pages are the most popular as well as it can scale to some truly in-depth information using custom evens and goal tracking.
Is it right for me?
Simple anser — Yes.. You need to stop reading this article and get Google Analytics installed on your website right now if you haven’t. It is completely free and sets the baseline for all of your testing efforts. This should be installed on day 1 of your launch even if you are expecting 0 visitors to your website because of historical records.
What does it show me?
Pretty much everything. It is a surface level view of your website. Google breaks it down into 3 main categories when you get it out of the box. There is Audience, Acquisition, and Behavior. These are great ways to organize your thoughts about the fundamentals of your website. Who is coming to your site, how they got there and what are they doing whe they get there.
Google Analytics also has some really great customization features for website user testing. You can set up custom event tracking and goal conversions. This is where Google Analytics can really be taken to the next level. You can set up conversions for when someone purchases a product, does an online registration or even submits a form. Once you have a goal created, you can see the actual funnel it took for someone to make that action whether it took a Facebook visitor 10 page views to arrive at your goal or a Google visitor 2 page views and a return visitor 1.
What are the drawbacks?
Yes, I think Google Analytics is amazing but alas its not perfect. The first is that it is a cookie based service that runs on the individual user’s browser. If you’ve ever had to delete your cookies or block a cookie then you can see how it may affect Google Analytics. Anyone can block or delete their Google Analytics cookie so this makes the data that Google Analytics collect open to not be completely reliable.
That being said, all of the analytics packages mentioned are based on cookies so this concern would apply to any below.
The bigger concern is the misuse and misreading of the results displayed. At its core, Google Analytics tells you fairly generic stuff. You can see how many people visited your website, how many pages they viewed and where they came from. This is interesting and useful information can be damaging in a vacuum. People can get attached to a particular metric whether it is the amount of visitors / sessons or the bounce rate and use it as the measure of success but this is bad. Real bad. Focusing on how many people come to your website is misguided as the only effects of that is that it eats your bandwidth and increases your cost. The bigger metric is what the visitors accomplish on the website.
2. CrazyEgg: User Testing for Landing Pages
We now get into a deeper look a some testing tools for website users with CrazyEgg. CrazyEgg is an older testing tool that began in the mid 2000s. It takes a different approach than Google Analytics to test what people see and do on single pages. It’ll take a screenshot of a page and then post a couple of relevant heatmaps on top of the screenshot so that you can see what people are clicking and viewing the most when looking at a page. It is a really neat tool and certainly revolutionary when it came out.
Is it right for me?
Single pages rather than whole website. The CrazyEgg Service is best for testing important pages like your landing pages rather than your entire website. You enter specific page urls that you would like to optimize and it will generate the screenshot for those pages. There isn’t much interaction between the given page and any other pages on the site. It is really great for focusing in with a laser on your most important content pages.
Tested page views should be greater than 1,000. CrazyEgg isn’t expensive but you may be wasting your time if the pages you want to test doesn’t have significant amount of traffic. I don’t hold fast to the idea that you must have at least 1,000 to get insights about the page but it shouldn’t be a ghost town either. I’d also steer clear from paying for extra traffic to test your results as well. In some cases this can be useful, but you want the online user testing results to be natural.
What are the drawbacks?
Its outdated. CrazyEgg is really useful in some cases to test website users because it is fairly inexpensive and easy to use. The on-ramp is straightforward and you can easily see and interpret the results. It’s just a bit dated and many other user experience testing services implement its features within its own suite of tools. If you want something in-depth then you may want to look elsewhere.
Avoid on dynamic or interactive pages. Do not use this tool to test your online users if the pages have any significant interactive elements. The service takes a screenshot of the initial page and then showcases the clicks on top of the screenshot. If you have interactive elements like contextual hover menus or dynamic content that appears then it would just provide confusing results because the mouse events will be pasted on top of the screenshot without the dynamic content.
3. Mouseflow: test the user journey
Mouseflow is a newer online user testing tool. It amplifies the traditional heat map ux tools by recording entire user sessions so you can see the entire journey of how people interact with your site. The benefit of this tool is that you can actually visually see what people are doing when they come to your website. You can see where people get confused. Why they don’t take certain actions and you can optimize for specific goals. I have used several times in the recent past and have always received positive results.
Is it right for me?
Tested website should have greater than 20,000 monthly visitors. This service isn’t ideal for brand new websites with few visitors. You should have a substantial amount of monthly visitors your website in order to take advantage of this service so we recommend an average around 20,000 visitors when using this service for testing website users.
Test website user funnels & business related goals. This service is geared best to test how the user interacts with the entire website so you can optimize your funnels and navigational structures. You will see why people get confused and lost on your website. I’ve experienced in many cases where people bounce back and forth from 2 different pages only to find what they need in an entirely different section of the website.
Ideal for larger content websites. This isn’t the tool to test your users if your website consists of 5 pages and a blog. This is best utilized for sites with significant amount of content with complex funnels and diverse user needs. If you have 3 or more different types of visitors that come to your site then you may find this user testing service useful whereas a website with 1 primary user target and a straightforward sales process may find other tools more useful
What are the drawbacks?
Significant amount of useless user testing results. The primary reason that you need a larger traffic numbers is you will receive considerable amounts of useless data and videos. There are times when you will need to delete dozens of videos just to arrive at 1 that you can use. It is a bit like finding a needle in the haystack but at least you can use this needle to sew together something useful (see what I did there?)
Time-consuming to see actionable patterns. We look for patterns in much of life and they just always take time to find useful information. It is the same way for this tool when testing your online users. You need to watch potentially hundreds of 30 second to several minute videos and there is no chance of cheating time here. You will need dedicate a good amount of time to make the most of this service.
User privacy. This service can start to edge towards creepy and an invasion of privacy. You are actually viewing how your users interact with the website a single visitor at a time and therefore will need to take steps to be responsible and mitigate any concern. The service does have a feature to hide any forms from being viewed which helps tremendously. You may also consider offering an opt-in version to test your users much like many software packages do these days.
4. SurveyMonkey: personalized user testing questions
SurveyMonkey is not an online user testing tool but a general survey service. I use the service to ask quick an easy questions to web visitors randomly. You can embed a survey into the website via a popup and allow the person to answer specific questions. I’ve discovered that using a survey is a good change of pace to the traditional data based metrics because it gives you real person feedback and can shed information that data just can’t show.
Is it right for me?
Need real person feedback. Surveys are the best way to test online users through feedback. You get real people’s opinions at scale and that is always invaluable. You can and should always ask your Facebook friends and relatives what they think of your website but this is fairly limiting. The results that you can get via a survey is eye-opening. Using a mix of open-ended questions with rating scale based questions will provide some of the most useful insights to your user testing efforts.
Test for online user opinions over time. The Net Promoter Score is one of the most used metrics in developing a brand. The goal is to ask whether you would recommend this product or business to a friend and then see the results over time. Asking the same basic questions over significant periods of time can produce important insights and provide a barometer for your brand’s customers.
What are the drawbacks?
Can be intrusive. Survey popups are annoying. There isn’t any getting around it. You can paint it pink and put a horse on it but nobody likes something irrelevant popping in your face asking for a favor. This is a little sad because people love basic surveys and giving their opinions. There is a right way and a wrong way to implement a survey on a website. If you need to ask yourself if this is annoying your users then it probably is.
Biased results. The good part about data based analytics services for web user testing is that unless your visitors have a cookie blocker on their machine then they will provide useful feedback. This isn’t so with a survey. We have yet to develop a robot that sits next to someone and gives them an evil stare until they fill out that survey. This means that people will choose to participate in the survey and participatory survey results generally don’t reflect a consensus. You will get more negative feedback than positive as well as feedback from the more opinionated users to your website.
Online user testing with the end in mind
Should you be testing and gathering insights about your web users? Yes. You also need to know what you want to achieve with the data that is collected so that you don’t fall into the analytics addiction trap. I’ve been there. It’s scary. Now I choose chocolate and coffee for my preferred addictions.
But know what matters in the end.
The amount of visitors coming to your website is a fairly useless metric as well as the bounce rate. They rarely tell any kind of useful picture but people become mosquitos to a lantern with those metrics because its flashy and they don’t know any better.
Better metrics are what moves you closer to your goals and are tied to more business related functions. For example, It is more useful to know how many times a visitor comes to your website before they give you their email address is far more useful than knowing that 30,000 people came to your website in the last 2 weeks.
Always start with your goals and work your way down to the metrics that move the needle to your larger vision.