Conversion rate optimisation (CRO) is often misunderstood as simply testing image selection or button color, etc. Approaching CRO with such a narrow lens, though, greatly restricts conversions growth potential.
To maximize conversions, you should implement CRO holistically. By building a CRO framework instead of treating CRO randomly, you’ll have a greater probability of unlocking insights, achieving breakthroughs, and driving a higher number of conversions.
Start the CRO process with audience personas. Why? Because you likely will find that different adjustments and customizations to your marketing are required to maximize conversions for each respective persona. What resonates with one persona may fall completely flat or be irrelevant to another.
Configure each CRO process as centered on a specific persona to maximize results. The more details you can define and understand about each persona, the more opportunities for testing will likely come to mind, and the quality of such testing ideas will likely be of higher quality.
You’ll typically find that there’s a hierarchy of conversion events in the customer journey. Especially in B2B, there could be many micro-conversions that occur along the purchase path.
Identify your ultimate conversion event that directly leads to increased revenue for your business. Then, define a “ladder” of conversions. This may include whitepaper downloads, mailing list subscription, event registration, webinar registration, survey response, request for additional information, etc. An optimisation that’s overly weighted towards your ultimate conversion event will often underperform if you ignore optimization of the various steps leading up to such purchase.
By all means, you want to test images and button colors. But to maximize your marketing results, you want to apply experimentation and A/B testing to just about anything you touch. Whether your web page design, or the psychology of your site visitors, or the copywriting, test everything. Analytics and data analysis should be applied to all aspects of your marketing.
Let’s say you test a whitepaper landing page. You should test not only the layout of the page, but also the images, messaging, CTA button, number of form fields, copy, trust elements, and offer, along with any secondary micro-conversion event on the page (even if only a behavioral test).
Go Beyond Your Landing Pages
To that end, A/B testing should not be limited to landing pages. Instead, treat your entire website as a petri dish. Whether your homepage or your products page or even your blog, test various elements continuously. You can even test elements in often ignored areas of your site, such as your global footer.
Go beyond your website. Use A/B testing with your email marketing, direct mail, advertising, no matter what you are trying to do in your marketing.
Ideally, you’ll identify a series of potential conversion roadblocks and then translate these into a test plan. Many companies run tests as they think of them rather than establishing a logical flow, and this can artificially limit your results. Instead, be methodical and plan it out.
Define a hypothesis for each test, and test against the control with such hypothesis in mind. Ultimately, allow the data to dictate what works and what doesn’t.
What’s difficult to understand is bad news for your marketing. The easier a site visitor understands a website, the more that they feel it’s a professional, well-designed site. Interestingly, the faster the process occurs successfully in the brain, the more likely a person will experience good feelings towards the site, and therefore the brand. We are hard-wired to enjoy the information that we intake seamlessly.
No level of CTA button testing will prove definitive until you look more holistically at your marketing. Is it immediately clear what you are talking about? You may have endless internal meetings and know your product inside and out, but don’t assume that your site visitor is going to know what “Get Started” (or any other call to action) means without a description or explanation.
Layered into all of this, explore the feelings of your prospective customer. As the neuroscientist Antonio Damasio discovered, it’s extremely difficult for a human being to make purchase decisions without emotion. As such, testing different colors for your CTA button or text length on the page can get you only so far.
At a certain point, test very different messages to elicit a strong emotional response from your audience. You can conduct small tactical tests until you’re blue in the face, but until you evoke a strong emotional reaction out of your audience, you’ll be greatly limiting the impact of any A/B test that you carry out.
Use the same technology across your tests. This ensures that you’ll be comparing apples to apples. There are so many useful CRO tools on the market, it can be tempting to jump around among various options. Resist the temptation, standardize on specific tools, and then use them religiously through continuous testing.
Complementary Behavioral Data
With CRO, it’s critical to measure conversion events themselves, whether form submissions, downloads, sign-ups, etc. What’s also highly helpful is to measure complementary behavioral data, such as heat maps, click maps, cursor movement, and scroll maps. With the right CRO software, you can also view actual site visitor sessions, providing you with qualitative behavioral insights that cannot be found by looking purely at numbers.
Ensure that regardless of the test, you have a sufficient sample size to make intelligent, informed decisions. Making decisions with an insufficient data set can often lead to the wrong conclusion. When conducting analysis, collect as much data as possible. Don’t take shortcuts, and don’t panic. When the data size is sufficiently large, then make clear decisions and execute decisively.
If you go through all the effort to test designs, layouts, forms, images, copy, offers, etc. and you’re cutting your tests short before you have sufficient data, you’ll never truly understand what works, and the lack of clarity will only compound over time as one unclear test is layered on top of another.