The Beginner’s Guide to Conversion Rate Optimization

Chapter 9: Tools to Test and Optimize Conversion

The most basic and essential tool you can have, analytics software tracks and reports on what’s happening on your site day in and day out. You want an analytics package such as Google Analytics, KISSMetrics, Mixpanel or similar that allows you to get at the basics (like unique visitors and bounce rate), but also has advanced analysis tools like audience segmentation, cohort analysis and conversion tracking.

Segmentation can produce data for different sets of people,and you can isolate hiccups or trouble spots in your conversion funnel. Cohort analysis allows you to group a set of users together by a common attribute to see how group behavior differs as the common attribute changes.

A simple cohort analysis example is examining how many users downloaded your app in a given month. You can compare users who downloaded in August to those who downloaded in September to see which group is the most engaged or most profitable. You can use that information to learn how product changes impact behavior.

Google Analytics (free)
KISSMetrics
Mixpanel
Segment.io
Chartbeat
Clicky
RJ Metrics
Woopra
Chart.io
Custora
Sumall
GoodData
Omniture

There are more, and depending on your business size, type and traffic you’ll need to determine which is best for you. For most companies Google Analytics is plenty. If you want to do cohort analysis, using a combination of Google Analytics and KissMetrics will do the trick.

Think of this chapter as both a manual and a catalog. If you’re looking for a basic explanation of which tool does what and how, you’ll find that here. If you want recommendations, you’ll find those as well.

Analytics

To say there is a tool for every job is an understatement. You will run into more tools than you could possibly use. After reading this chapter, however, you should feel equipped to wade through all the options and select the right tools for your job.

If you’ve skipped right to this chapter without reading the preceding eight you may be in grave danger of making the classic CRO testing mistakes of implementing tools and tactics without a plan. If that’s you, take your hand out of the toolbox cookie jar and return to chapter one. You’ll be back here before you know it and better prepared to leverage the resources in this chapter.

Think back to our basic toolbox from Chapters 4 and 7. In it, tools are divided into three main types—analytics, user surveys and user testing. Think of the those categories as drawers in your toolbox. These drawers might contain any number of tools, depending on your needs and budget.

If you’re a hobbyist building birdhouses, you might only have a tool or two in each drawer. If you’re a professional carpenter with a twenty-person staff, you’ll probably have several. Still, the likelihood of the hobbyist and the professional having some tools in common is pretty high.

Now about those drawers…

But analytics can only communicate so much about your users’ needs; you need something that gives you the ability to gain insights directly from users in the moment, to hear their concerns in their own words; there is no such thing as too much user feedback. This is where user surveys come in.

For example…

You’ve used analytics to figure out that a large percentage of your visitors are dropping out at a certain point in your conversion funnel, but you’ve gone through the funnel yourself several times and can’t seem to figure out what the problem is. Implement an inline user survey at this exact sticking point.

In the most basic application of user surveying, all you really need is one good, open-ended question that allows users to express their concerns. If you’d like, acquire an email address for feedback purposes—but be sure to keep your word and actually follow up.

In more advanced iterations, user surveys can facilitate an A/B split test—run page-level surveys on both the original and the test page, and see how visitors’ answers differ.

Qualaroo offers unobtrusive inline surveys that allow you to ask questions on specific pages or at specific points in your funnel.

User Surveys

CONTINUE TO CHAPTER 10

User Testing

But what if users can’t articulate the problem they’re having? Luckily, there are plenty of user testing software options. These tools allow you to directly observe how users are interacting with your site. You can use their feedback to get a better understanding of your site’s functionality, as well as test potential changes and document how they’ll play out in real life.

Brace yourself, however, because the User Testing drawer is the most cluttered of them all. In the interest of keeping things organized,we’ve added some additional dividers to this this one.

To figure out what’s grabbing your users’ attention—where they’re looking, clicking,and lingering—use one of the following tools, which perform Heat Mapping and track Click Density. Much like user survey, these tools help you to make sense of how users engage with your site so you can form hypotheses about how to improve this experience.

CrazyEgg is an eye tracking technology that allows you to visualize where users are scrolling on your site through the creation of heatmaps, scrollmaps, overlay tools and confetti tools.
Click Tale is a tool that creates videos of user interactions with your site so that you can see how they use the user interface elements of your site. The tool also analyzes heatmaps, scrollmaps, among others, while also creating videos of user interaction.

Next, we have Concept Testing Tools. Once you’ve formed your hypotheses, these tools allow you to quickly and painlessly draft a test version of your site (also called a wireframe) and get feedback regarding potential trouble spots.

Balsamiq is a wireframing and mockup tool that allows you to create low-fidelity sketch wireframes and prepare hypothesis for testing.
Cacoo is a user-friendly drawing tool that allows you to create a variety of diagrams such as site map, flowchart, mind map, wireframe, UML diagram and network diagram.
Browsershots is a free, open-source web application that makes screenshots of your web design in different operating systems and browsers. You submit your web address and a number of distributed computers open your website in their browsers. Then they will make screenshots and upload them for your review.

There are A/B Split Testing and Multivariate Testing Tools. These are are invaluable when when it comes to trying out and comparing various iterations of your webpage.

Optimizely is a website optimization platform focused on A/B and multivariate testing and making them easier to use and understand on your site.
Google Content Experiments is integrated with Google Analytics, and is Google’s free website testing and optimization tool.
Visual Web Optimiser also focuses on an easier approach to A/B and multivariate testing, but includes behavioral targeting, heatmaps, usability testing, as well.
Unbounce also offers A/B testing, while focusing predominantly on the efficacy of your landing page.

Finally, there are User Testing tools. These allow you to get user feedback from users like yours for your existing site, new product features, or new site designs. These are great for getting detailed qualitative feedback about how easy your site is to use for first time visitors.

UserTesting.com lets you find users like yours and give them a series of tasks to perform or questions to answer by using your site. Their session is recorded so you can watch their screen. The users also narrate their thinking, so you can get detailed feedback about their thought process while using your site.
VerifyApp.com makes it easy to get feedback on screen shots or designs in terms of which design an audience likes better, is more memorable, or more trustworthy.

Try not to feel overwhelmed by the sheer volume of tools available. In the real world, we acquire tools over time and as we need them. It might help to consider stocking your optimization toolbox this way as well.

For example, if you’re still in your data-gathering phase, focus initially on implementing a great analytics package. Once you’re comfortable with that, examine your needs and select a user survey service that meets them. After you’ve implemented user surveys, begin researching testing options.