Testing is an important ingredient in the recipe of a successful web site. If a site becomes stale or needs a burst of flavor to reach a wider audience with a more discerning palate, testing allows for making changes on the fly that can lead to a much larger impact.
However, with so many different aspects of your site you can test, a well-designed testing plan is needed to ensure you truly understand the ‘what’ and ‘how’ of the test results. To help you cook up a good plan, we will go through some of the most common and impactful varieties of testing to add the right amount of flavor to your site.
Low Hanging Fruit
These types of tests are where you can make small changes that have potential for BIG impact - changes that are simple and cost you little-to-nothing to implement.
Simple changes, like changing the text on a ‘Call to Action (CTA)’, can bring greater interaction by grabbing the visitor's attention and making them feel like they are missing out if they don’t take action. "Free Trial" sounds like it would work, everyone likes free, right? But, doesn't "Start Your Free Trial" sound like you are actually doing something, like you are making something happen?
Another small change/ big impact test could be simply changing the background color on a page, which could change the readability of a page as well as highlight buttons and other CTA objects.
Visitors of today have a lot less patience when it comes to consuming digital information. They consume so much of it every day; they need to quickly evaluate what is valuable and what can be skipped. The website real estate above the fold grabs 80% of the attention of a page. So by removing large images or unnecessary filler content, you can fit more content "above the fold" in order to quickly drive messaging without requiring your visitor to scroll.
Meat and Potatoes
These tests are the middle ground of testing and are likely where most of your tests will, and should, fall. Tests of this variety may require a little more time and energy to create, but allow you to test more complex ideas. They incorporate some of the ‘low hanging fruit’ concepts with more detailed\complex concepts.
CTA Button Design & Position:
You’ve already changed the CTA text, but why not also include updates to the aesthetic look of your CTAs (and all other CTAs)? Small changes like increasing button size, allowing more white space around the button or changing color to make the button more visible are easy changes to make. By taking it a step further, you can run tests on the styles and position of buttons to help increase conversions.
Note: just be careful to keep standard CTA buttons uniform, it’s important to make sure they are consistent so visitors can easily find them!
Similarly, you have tested moving content above the fold but you could test the layout of other site components, such as reordering your menu. Reordering your menu can help visitors access pages more easily instead of having to dig through your site.
Or perhaps your page layout needs a refresh, try moving CTA buttons to be more visible or move images below/above content paragraphs to see if it allows for better page structure. By taking a bit more time re-structuring, your visitors experience will flow easier and may increase the overall rate of goal completion.
One of the best tests for product and services pages is to test page design by adding new elements that show greater detail for a product or service – images are processed faster than text and create an emotional connection to what is being viewed. In fact, content featuring images has 94% more total views
than content without images so it’s worth including in your testing efforts.
These types of tests involve BIG changes, requiring large amounts of time to setup and can require additional resources (such as designers, developers, etc.). These types of updates are typically done when adding new functionality or when performing major site overhauls but can result in MAJOR improvements.
Maybe it is time for a complete overhaul of your site - this could include changing the entire stylized theme of the site; background images\colors, menus\navigation, page structure, features, etc.
This could be any set of changes that involves lots of development time (both in the planning\design phase and in the coding\implementation phase), or requires bigger changes to more than one element or group of elements.
Keep in mind; the more you change at one time, the harder it is to attribute success (or failure) to any one change that was made. Perhaps a portion of your changes are successful, but others caused the overall goal rate to drop because they made it more difficult to understand or interact with the page.
And also, remember that the increase in development may not yield as big of an impact because you may have changed so much that you are now comparing apples to oranges, instead of Golden Delicious to Granny Smiths. You’ll have to wait to perform tests on your new site once you have a good baseline of data to work off of.
Testing has a very important purpose, it helps to determine whether your current model, or a new one, provides a better user experience or return on investment (ROI) per visitor. A/B or multivariate testing can give you a much-needed look into the preferences of your visitors or the performance of the elements of your site.
The first step is to decide how what your site or page's actions are that lead to the overall conversion\goal - is it button clicks, filling out a form, ordering a product, or something else? Once you know HOW a user must interact with your site to reach the defined goals, then you can create events to track those actions. Keeping in mind what was mentioned previously, the more changes you make, the more actions you need to track to determine which changes, if any, are responsible for the lift\failure of a test.
You can, and should, create segments of your population (each site may be different) in order to be able to track usage of the page\site during the test. Those events are then tracked as a whole, and can be compared across segments and variations in order to determine if there is an overall winner, or a different winner in different segments.
Truly optimizing your site is a 4-course meal, it’s not a microwave dinner. You should optimize one course at a time, don’t overdo it by trying to change everything at once - changes are best made when you can properly track their effectiveness in a controlled environment. As you continue to optimize your site, don’t forget to test and test, then test again. Once you start testing, you will have good baseline metrics to follow and can start building out your site persona based on visitor behavior. Bon a ’petite!
What are your best testing & optimization tips? Leave a comment below!
For help with your testing or optimization strategy, feel free to reach out to us at firstname.lastname@example.org
Sources: National Retail Federation, Moz.com