Figure 9-8. Original shopping cart design, without assurance center
Can we say with absolute certainty that adding an assurance center for this client will always produce negative results? No. Several elements could have influenced this particular design and caused the drop in conversion rate. For instance, the assurance center design, copy, or location could have caused the drop. Analyzing how a hypothesis is validated through testing data and creating a follow-up hypothesis is at the heart of conversion optimization. In this case, we needed to test many different elements around the assurance center before we could determine its impact on conversions.
Tests that produce increases in conversion rates are excellent in terms of validating initial assumptions and hypotheses. We do not mind tests that result in reduced conversion rates because we can learn something about our hypotheses from these tests. It is tests that do not produce any increases or decreases in conversion rates that we worry about.
Two factors can impact your conversion rate. These are:
Internal factors that you can control, such as your design, messaging, copy, and so on.
External factors over which you have little control. For example, when a competitor runs a large sale, your conversion rates can suffer. If the economy gets worse, you can expect your conversion rates to suffer as well.
Successful testing requires that you pay close attention to external factors. And although you have no real way to control these factors, you should attempt to minimize the time it takes to run your experiments. The less time it takes to run tests, the fewer chances external factors have to influence your test results.
DON’T FORGET!
When conducting online tests, your goal is to arrive at statistically valid results in the shortest possible time. If you are not careful when designing your test experiments, you will have to run test scenarios for years to come. Pay close attention to create tests that conclude within six weeks.
As mentioned, you can test a number of elements on your website. In the subsections that follow we discuss the 49 things you should test across your website. We have discussed some of these elements in different sections in the book, but we are including them all here for easy reference.
Too many companies slap too much information into the header. Is your header crowded with information? How much information are you presenting to the user? We consistently find that clean and well-organized headers work best. Figure 9-10 shows the Frys.com header. The header is crowded with information that can overwhelm visitors. Compare this to Figure 9-11, the header for NewEgg.com, which carefully uses white space.
A good tagline that communicates your value proposition will have an impact on your conversion rate. Test different taglines to see which resonates best with your visitors. A word about taglines: think of something unique about your business. It is rare for excellent customer service or lower price to make a good tagline.
With most ecommerce stores, you can expect 15% to 25% of visitors to rely on the search box for navigation. Some online stores, such as auto parts websites, report higher percentages of visitors using the search functionality (close to 40%). Placing the search box in a prominent location will encourage visitors to use the search tool. We recommend testing different placements and sizes of search box. Figure 9-12 shows search boxes on BestBuy.com, CompUSA.com, and Buy.com.
Many ecommerce stores fail to pay attention to navigation. Even worse, many stores let their technical teams and engineers determine the site navigation. Site navigation is designed to help users find what they are looking for quickly, so a good scheme should be focused on visitors and the tasks they complete on your website. We recommend investing the time to test different categorization schemes and their impact on conversion rates.
Start by testing multiple navigations for your website with the product page as the conversion page (micro tests). This will allow you to determine which of these navigations helps visitors find their way to the product page more effectively.
Visitors will most often look at the image first when scanning a landing page. This image can make the difference between whether visitors stay on your site or leave. The most important criterion for choosing an image is its relevance to the product or service being promoted. If you are selling widgets, a great shot of the widget will show your customers precisely what they will get. If a product shot is not appropriate, your image should tie in with any creative or graphic design elements that were used in the advertising that got the visitors to your page.
Images must connect with a value proposition. They also need to be considered from the perspective of personas. A client that sold communication devices displayed an image of nature scenes on their landing page to communicate that their product can be used anywhere and at any time. This did not work for visitors, since they all happened to be busy businesspeople who needed a better reason to purchase the product. We tested an image of a person in an airport using the device against the original image of nature scenes. The new image improved the conversion rate by 34%.
