Patrick Murphy
Why should you test your website?
Every colleague, employee, and customer has an opinion about your website. Most of their beliefs are out of date, biased, based on experience in some irrelevant realm or, most often, just plain wrong. Rather than going with your opinion, their opinion, the opinion of your CEO or the latest trend from the marketing department, pursue the most functional option: testing. Most likely, “your website is underperforming.” The gap between its performance and its potential importance and return on investment is the “opportunity gap.” You can fill that gap, but proceed with care: Whenever someone suggests something that affects your site, test it.
Your goal in testing is “conversion optimization”: getting “more revenue-generating actions from the same number of visitors.” Whatever your conversion rate – the ratio of website visitors “who complete your desired action” compared to those who don’t – you can improve it. Most people who visit your site leave disappointed. You may not even know about it – much less, why that happened.
Look for ways to improve; a site can always be faster, easier to use, and more functional. Most people can’t define what website success looks like. Their designers don’t provide objective criteria for success. They use their competitors’ websites as examples and base their sense of success on their personal impressions. They follow industry “best practices,” which might or might not offer the most appropriate solution.
In our opinion utilize “the scientific method…to find optimized solutions” before redesigning your site or selecting another design using the same flawed techniques. If you make changes that lack a scientific basis, you might lose customers, fix some design flaws while introducing others or alter something that isn’t a problem. With “controlled testing,” you can identify what works and what doesn’t, and change only areas that need fixing. This way, your site will evolve in positive steps.
Understanding conversion optimization
“Usability testing” – or “user testing” – helps fix basic problems and articulates your hypothesis for actual conversion optimization. Surveys are useful for gathering “qualitative data” from your site’s users, but those opinions do not constitute conversion optimization.
“Heatmap tracking” – using tech tools to portray visually where people click on your website – is useful, but isn’t conversion optimization either. The practice most commonly mistaken for conversion optimization is “Before & After” or “ “Pre- & Post-” testing.
Before and after testing is dangerous because it only appears scientific: You formulate a question, make a change, do a test and gather data. Yet, before and after testing does not provide a controlled test. Instead, it lets “external factors” distort your data. Imagine that you make a change on your site at the same time a competitor also introduces a major change.
Are the differences in behavior you observe among your visitors due to your change or to your competitor’s? You can’t tell. True conversion optimization requires three elements:
1. “Persuasion marketing” – Craft messages that call your users to action.
2. “Experience design” – Create a website that’s easy to use.
3. “The scientific method” – Shape a hypothesis and test it in two different areas. Identify the “target audience” for your test and set goals for it.
Your goals should cascade from the most important to the least in a goals waterfall. Identify priority goals and assign them a relative value. To make your ranking more utilitarian, determine what financial return each achieved goal would produce. Ideally, conversion optimization operates in a “continuous-improvement cycle.” Design a hypothesis, test it, and make changes. When circumstances alter, test again.
Our marketing strategy
Follow the “Act / Plan / Do / Check” loop before starting the continuous improvement cycle again.
Digital & Social Articles on Business 2 Community
(276)
