While a website redesign can be an effective way to improve the user experience, making smaller, more focused adjustments as part of ongoing optimization can also deliver meaningful results. Through experimentation and testing, businesses can identify the strongest opportunities for improvements, explore different approaches, and make evidence-based decisions to enhance website performance. This ongoing process helps businesses understand what is working well and make targeted refinements based on the evolving needs of their audiences.
Site analytics can help teams surface which parts of the website would be good candidates for A/B testing and redesigns. Certain metrics like conversion rate, bounce rate, rage and dead clicks, click-through rate, and screen recordings give insight into where users are encountering issues. This research helps inform design decisions, and offers the opportunity to develop multiple variants and test how each of them performs. Because these changes are not permanent, A/B testing can be a lower-risk way to explore new ideas, providing flexibility to experiment and evaluate their effectiveness before committing to a change. This data and user behavior can help signal to businesses what works without relying on assumptions or opinions.
Testing even small changes can lead to measurable improvements in a website’s performance. For one travel client, data revealed that one of their booking widgets was causing user frustration and confusion. We tested two alternate solutions where we incorporated filter pills into the design to give users more visibility into their selections, and new buttons to help navigate users through the multiple tabs. Following the experiment, we learned that both variants consistently outperformed the control, with the winning version achieving a 3% increase in bookings, and a 7% increase in overall revenue.
Conversion and revenue are just some of the ways we can track these improvements over time. Other metrics that are often considered are rage clicks or dead clicks, which can be seen in screen recordings from users who are unclear about where to find what they are looking for. For an e-commerce client, we worked on a redesign of the homepage, with the goal of improving hierarchy, scanability, and product education through clearer content organization. After implementing these changes, the website homepage showed a significant decrease in dead clicks (89%) and rage clicks (88%), indicating that users found the interactive elements less confusing and eliminated friction points in navigating the page.
These collective changes made during experimentation can add up to substantial improvements over time. By running multiple experiments on various parts of the site, businesses can evaluate the different approaches and apply those insights to similar components across the site. With an ongoing approach to optimization and testing, websites can continue to evolve based on measurable results. Every test is an opportunity to learn something new, and each small shift is a step toward a better user experience.
| Metric | Definition |
| Conversion rate | % of visitors or sessions that complete a specific, desired action |
| Bounce rate | % of visitors who land on a page and leave without clicking to another page or taking further action |
| Rage click | Rapid clicks on the same element or area 3 or more times within a short window of time |
| Dead click | Single click or tap on an element that produces no response, page change, or visual feedback within a few seconds |
| Click-through rate (CTR) | % of visitors who click on a link, ad, or webpage out of the total number of people who see it (impressions) |
| Screen recordings | Video reconstructions of real visitors navigating a website, capturing mouse movements, clicks, scrolls, taps, and keystrokes. |
Small changes can drive big results, but only if you know which ones matter. Chartis’s UX and Experimentation teams help you identify, test, and prove the improvements that move your metrics. Ready to see what a few smart changes can do? Let’s talk.