Skip to main content
Library homepage
Loading table of contents menu...
Business LibreTexts

20.3: What can you test?

  • Page ID
  • \( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)\(\newcommand{\AA}{\unicode[.8,0]{x212B}}\)

    The short answer: everything! However, there are some good tactics to start testing.

    Email marketing

    The most obvious place to start testing your email marketing is with subject lines. A simple split test will help you to determine which version of a subject line improves open rate. Within an email, you can test your call to action copy to see how you can improve clickthrough rates. Email is also a good medium to test different kinds of offers to see how they influence sales. These can be combined with testing different types of content: long copy versus short copy, or images versus video.

    You can also test delivery days and times for your email, either with a split test or by testing with your whole group at different times of the year.

    Display and search advertising

    There are many different conversion points you may want to test with advertising. You can test to see how different adverts may improve clickthrough rate, or you can test to see how different adverts affect the conversion rate of the traffic coming to your website.

    Different calls to action in the adverts can be tested, as well as different headlines. When it comes to display advertising, completely different versions of banner adverts can be tested.

    Most Ad Servers, including Sizmek ( and Google’s AdWords, have built-in testing. This means that no additional code is required in order to run tests. In most cases, the Ad Server will also serve adverts based on the results of the test, so that the better-performing advert is given preference over time.

    Figure \(\PageIndex{1}\): Two search ads with different headlines Adapted From Screenshot, Google, 2017

    Social media

    If you are broadcasting or distributing messages via social media, you should test your messages to see which perform better. Usually, you will be testing to see how you can increase engagement, whether that’s replying to messages, liking Facebook posts, retweeting posts shared on Twitter, or clicking through on links shared.

    Types of messages can be tested to see which increase interactions. Some options to try are direct questions, shared links, overt requests to take an action, or other messages to generate engagement. Types of media can be tested, such as images versus video. Time of day or day of the week can also be tested, to see how these affect interactions.

    Figure \(\PageIndex{2}\): Facebook Insights showing what kinds of social media posts make the biggest impact on your followers? Adapted From Buffer, 2017

    It can be trickier to test social media messages scientifically, as the environment is difficult to impossible to control.

    Landing pages

    A landing page is the first page a visitor sees on a website. In some web analytics packages, it is referred to as an entrance page. Any page of your website could be a landing page, especially if users are coming to your website via search engines.

    However, when you are running online campaigns that utilise tactics such as email marketing or online advertising, you often send visitors to a specific landing page. Because you choose the page that visitors see first, you have an opportunity to craft a page that converts. There are many things you can test on landing pages.

    • Heading: Different headings can make your visitors behave differently.
    • Copy: Style, tone, layout and length of copy can all be tested, as well as things like the font size.
    • Call to action: Different calls to action could increase actions.
    • Colour: Test the colours of buttons, green and red are two common choices to try.
    • Images: Ddifferent images can have an impact on conversions.
    • Offer: Don’t forget, you can also test different types of offers on a landing page or word the same offer differently.


    There are many things you can test on eCommerce websites, but the most obvious are to test your product pages and your checkout process. With eCommerce, you are usually testing to increase your conversion rate ( when more consumers buy from you), or to increase your basket size (when consumers buy more from you). As your testing becomes more sophisticated, you will hopefully be achieving both.

    On product pages, you should be testing to see how you might encourage users to start the purchasing process. Some things to test include:

    • Images: Images of different sizes, or entirely different image content, could make a big difference.
    • Call to action: Seeing the words ’Add to cart’ or ‘Buy now’ may affect shopper behaviour.
    • Shipping information: Test to see whether displaying shipping costs before actual checkout affects the volume of customers starting the checkout process, as well as its effect on completing the purchase.
    • Credibility information: Do ‘secure shopping’ badges affect conversion rate?

    Product pages are also a good place to test how you might increase basket size. You can test displaying complementary or similar products. If you spend time on Amazon, you’ll see that they display items that customers bought together, suggesting you should do the same.

    The checkout process is incredibly important to test. Often, an online purchase takes place over several screens, or steps. Instead of looking at the overall conversion rate, you should be looking at the conversion rate between steps. This is referred to as funnel analysis.

    If possible, you should test a single-page checkout versus the multi-page approach. Test credibility and reassurance statements in the checkout process. Font size can also have an effect on conversions.

    The above should give you an idea of where you can start with your conversion optimisation, but it is by no means an exhaustive list.

    This page titled 20.3: What can you test? is shared under a CC BY-NC-SA 3.0 license and was authored, remixed, and/or curated by Rob Stokes via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.