The Importance of Testing: Does Your Website Make the Grade?

Andrew Gross, Indaba Group

Andrew Gross, Indaba Group

Author Bio

Andrew Gross is a technical writer and test case developer with Indaba Group, a strategic ecommerce agency based in Denver, CO. For more information on Indaba Group, visit their website.

The big red button for your newsletter signup seems so obvious. Why isn’t anyone clicking it?

When attempting to optimize a website, we often focus on the big picture. Why aren’t shoppers clicking the button? Why are they abandoning their carts? Why are they leaving? While these questions are important, they don’t immediately address conversion rates.

A more pragmatic approach, and one that leads to more immediate results, is to focus on the specifics.  From what page are visitors dropping off? What content are they engaging with? What are they ignoring?

When you track down the answers to such questions, you can make changes until you find what works for your customers. This iterative testing process allows you to systematically adjust any element on a site, from a headline to a button color to a specific form field, to optimize a site for the majority of visitors. Such incremental changes add up, and on a high-traffic site, small changes can result in a major lift.

It All Starts With Data

To get started, you need access to current, reliable user data. Fortunately, there are a host of tools that allow you to pinpoint elements on your site that are performing poorly. These tools provide a baseline for testing.

Google Analytics:  Assuming GA is set up properly, identify your top exit, funnel drop off and high-traffic pages.

Heat Mapping: Heat mapping code, embedded on your site, shows which elements users are engaging with and what they are ignoring.

Form Analysis: Software can tell you how long users are spending on specific form fields. This problem can often be addressed with clearer instructions or by limiting options.

Video Journey: Video journey software allows you to see exactly what visitors are doing on your site. It’s the next best thing to a focus group. This type of information is very useful for identifying problems, but it’s not as scalable as the other options. The same goes for customer surveys, which collect more qualitative data.

Prioritize Your Plans

Once you have gathered the necessary data, it’s important to set some priorities and determine where to start. Focusing on high-value, low-cost fixes will give you immediate satisfaction and help you address the issues that impact the most visitors.

In Google Analytics (GA), find pages with the following:

  • Highest bounce rate
  • Highest exit rate
  • Lowest time spent on page
  • Most expensive adword traffic

You’ll also want to look closely at the pages associated with your carting and checkout processes, account creation and newsletter signup, and your top product pages.

For each, locate the corresponding heat mapping, form and video data, and use this information to create individual tests. Dissect each page, starting with elements above the fold. Look at navigation, calls to action and content engagement first.

Build Your Tests

When creating your tests, it’s best to start with items that are easy to implement on high-volume pages. This allows for multiple scenarios and quick data collection.

Let’s say you’ve determined that customers are not engaging with your newsletter signup button. In this case, there are a lot of things to test. It’s important to follow a systematic approach to allow for gradual and complete optimization.

  • Start with a single proposition. For instance, you might change the button color.
  • Create a test with multiple scenarios (red, blue, green, yellow). The current color will serve as the control.
  • Collect results.
  • Once you’ve determined the best performer, move on to the next proposition. Note: Often, the control will win, and this is fine. It tells you that color is not an issue.
  • Proceed to the next change. Perhaps this will involve the button wording.
  • Make the copy change, and repeat steps 2-4.

For each test, it can be helpful to track your results with a simple spreadsheet. For example:

Indaba Group Example

Starting simple and working toward more complex scenarios allows for ongoing optimization. In the next phase, you might consider:

  • Segmenting tests based on platform (mobile vs. desktop), customer demographics and customer behavior.
  • Scrambling scenarios to test different combinations. For example, change the wording first, then the color.
  • Multivariate testing (testing multiple elements at once).

Expand Your Efforts

Everything can benefit from a little testing and optimization. Once you have a solid web testing process in place, apply the same principles to your other business collateral and marketing efforts to help your team’s hard work pay off and get the most bang for your buck. Such testing isn’t just for websites and email marketing campaigns. It can also make a big difference in the effectiveness of your social media engagement and brand recognition. A little testing can definitely go a long way.

For more information on Indaba Group, visit our website or contact me at agross@indabagroup.com.