Leading ux testing to increase website conversions

Discovery

Uelz is a young startup focused on creating a Billing API to automate all types of recurring charges. Over time, it has evolved its commercial strategy and shifted its focus toward a higher-end target audience. It is currently aimed at large companies that process a high volume of recurring transactions and need to modernize and delegate their entire billing infrastructure.

Uelz offers two usage modes: via API or through its proprietary software with its own interface and development. It allows clients to adapt to any type of payment model with an international reach.

Although they are confident in their product’s potential, they acknowledge that it is not easy for their target audience to grasp or visualize. As a result, the marketing and support teams have been gathering common doubts and misconceptions for quite some time. So far, they have tested various copies on their website as their product has evolved. They now believe they have found the right message that clearly defines who they are and what advantages their product offers. However, they are still struggling with user acquisition.

Additionally, many users who access their demo are not their actual target audience, and in some cases, they are confused about the platform’s functionalities.

To achieve this, it will be necessary to conduct a study where objectives and hypotheses are defined, along with determining the best methodology for validation.

Define the problem

First and foremost, it will be necessary to speak with the department that brought this issue to the design team and gather as much information as possible about the doubts users have. This includes verifying whether that information is present on the website and how it is communicated. Once all relevant data has been collected, the next step is to formulate the main hypothesis that will guide the entire research process.

Similarly, I break down the general hypothesis into smaller ones that help identify more specific friction points.

As a result, I obtain different hypotheses to iterate on in order to validate the general hypothesis. These specific hypotheses will help determine the research method that best fits the inquiry.

Methodology & structure of the process

Qualitative studies generate data on behaviors or attitudes based on direct observation and listening, whereas in quantitative studies, data on the behavior or attitudes in question are collected indirectly through measurement or tools such as surveys or analytics platforms. Qualitative methods are much better suited for answering questions about why something happens or how to fix a problem, whereas quantitative methods are more effective for answering how many or how much types of questions.

Since the key question to answer is why users do not understand what Uelz is through the website, I choose a qualitative methodology.

Below, I have attached a table from the Nielsen Norman Group website that illustrates 20 popular research methods based on their study dimensions. This chart serves as a decision-making guide to help determine which method best fits our research objective.

Among all the methodologies, I choose Contextual Inquiry, where users can interact with the product naturally while being observed and asked questions to better understand how and why they behave the way they do.

Study Structure: Contextual Research

Based on real product usage—the actual website—I structure an online interview for ideal users. The type of information to be collected relies on observation, impressions, and reasoning regarding their understanding of the website (Narrative Research). To achieve this, and considering the specific hypotheses, I develop a script aimed at validating them.

Additionally, before the interview, I decide to include a short test called the 10-Second Test. This will allow me to quickly analyze what users think, their first impression, and what they understand at first glance. Many studies suggest that the first few seconds of navigating a digital product are crucial for making a positive impact. Users should understand what the site is about as quickly as possible.

User Recruitment

Recruiting the right users for the study is almost as important as selecting the appropriate UX research method. In this case, the ideal participants are those with a professional background in technology-related businesses, who are familiar with or have worked with subscription-based business models, or those involved in developing and implementing code infrastructures for their companies—specifically, backend profiles.

I define two ideal interviewee profiles, categorized by their professional background, and send them a screener with the following criteria to determine their eligibility for the research.

Analysis & insights

To analyze the collected information, I transcribe the recordings and create an Affinity Diagram, visually organizing the different responses using post-its. These responses are grouped based on the research objectives, allowing for a clearer identification of patterns and insights.

• Website Navigation

• Product Understanding

• Value and Functionalities

• Documentation

Once all the ideas are placed into their respective categories, the concepts are grouped to extract the final conclusions.

Presentation of conclusions

After grouping the different categories and clustering post-its that reflect the same concept or idea, the final conclusions are drafted.

I write them in a way that reflects the users' own statements, either confirming or refuting the hypotheses.

Conclusion

Following collaboration with the marketing department, I redesigned aspects of the website and created new design structures to clarify information and provide users with useful content.

The redesign successfully increased user engagement, reduced repetitive questions, and improved the overall site experience