Matt James

Engagement

The Client

Aviva is a financial services company that offers a range of insurance products. From car, home and health insurance to pensions, investments and asset management, they offer a wide range of insurance and saving products to help protect what's important to their customers. Aviva are a 'composite insurer' meaning Aviva is one business made up of smaller units that specialise in different types of insurance.

The Challenge

I was part of Aviva’s pensions team, focusing on bulk annuities, which involved acquiring pension schemes from businesses. Aviva would purchase these pensions in bulk, and our goal was to create a self-service portal for end users to check their eligibility. We divided users into two groups: those nearing pension age who wanted to assess their retirement timeline, and those who were already receiving their pension. My role was to design a pension dashboard that users could access on desktop, mobile web, and app platforms. The dashboard would not only serve as a self-service tool but also as a sales tool to attract more pension products and repackage them for Aviva’s offerings.

Research and Planning

I joined the project midway through, after the discovery phase had already been completed. The stakeholders were eager to move forward with ideation due to tight deadlines. As a result, I inherited high-level user journeys and personas. There were two distinct user groups: one group consisted of individuals considering the age at which they wanted to begin drawing their pension, while the other included users who had already retired.

We knew that the primary concern for these users was their income—specifically, how much they would receive from their pension. To address this, we brainstormed features such as income calculators and information about various pension options. Compliance and regulatory considerations were also important aspects of the project that had to be factored into the development.

Each of the segmented groups had different needs, which required tailored user journeys. However, a common theme emerged across both groups: users were primarily concerned with how much money they would be receiving. We focused on providing information relevant to their specific stage, whether they were pre-pensioners or retirees, and also included budgeting tools to help them plan their finances.

The artifacts above were provided to me when I joined the project. They were created by an agency that had worked on developing personas and gaining insights into the users, as well as understanding the stages they go through—from initially being curious about retirement to eventually making the decision to retire.

“I began mapping the user journeys, and below is an example of the iOS journey. Aviva has an established design system, so certain parts of the journey followed a predefined pattern, starting from the home screen as shown above.”

Prototyping and User testing

As a team we had assumptions of what the different user groups would want to see and where the level of priority would be giver to the content on the dashboard. I came up witha design but we wanted to test it with users. we used usertestinglcom and Optimal Workshop with mediated workshops. we had 2 research methods with the following objectives.

1. Card sort (Using Optimal Workshop)

• Uncovered the most and least popular grouping of cards and their category labels.
• Understood what content or tools participants allocated to different groups and why.
• Understood which cards go together according to participants.
• Ascertain whether any content or tools should appear in more than one group (repeated), whether groups could be merged, or if cards didn’t belong.

2. Usability Testing

Retirement dashboard: projected income before tax banner:

• Ascertained if participants grasped that the project before income tax banner tells the pre-retirees that will be their income when they retire and the retirees that is their current income.

Your pension income tab:

• Uncovered if participants understood that the guaranteed income for life projection would show them an income based on the age, they choose to take money from their pension and whether they take a tax-free lump sum.
• Assessed if participants understood that the guaranteed income will be lower if they take a tax-free lump sum.
• Whether the case study clearly signifies it’s an illustration and not based on a real-life customer experience.
• Determined if participants grasped the Terminology explained section contains articles to help them understand pension and retirement jargon.
• Understood whether participants thought that the :Transfer your defined benefit pension” section didn’t recommend that they transfer their pension to Aviva.

Tooling and budgeting tab:

• Uncovered whether participants understood the differences between pre-tax (gross) and post (net) tax and that they could pay tax on their pension income.
• Ascertained whether the budgeting tool communicated that it was there to help them plan their finances during retirement.
• Uncovered if the cost summary and the monthly spending breakdown outline the income and outgoings from the information entered within the budgeting tool.

---

---

The card sorting exercise was valuable for gaining insight into the users' perspective on information architecture. It was especially helpful for organizing the article and informative text, as it provided clarity on how content should be grouped and what should appear on each page.

The two pages below focus on the key areas of interest for users when making retirement decisions, specifically around age and expected income. We also proposed adding a budgeting tool to further support users in their decision-making process. All controls were based on Aviva's established design system, so the foundational elements were already in place. My role was to ensure the correct design patterns were applied based on the users' needs.

The primary goal was to create an intuitive dashboard that would help users make informed decisions without feeling overwhelmed. By following this approach, we aimed to provide the best possible user experience.

Below are the pages we tested across both desktop and mobile devices. The test also included questions about device preferences. Desktop and iPad were perceived as the best devices for using calculators, while mobile was favored for reading articles and gathering information.

“Mobile and iPad designs were also tested with users, who were given tasks to complete while being observed and listened to in order to gather feedback. It's important not to influence the users, but rather to observe their actions and listen to their responses when they are left to complete the tasks on their own.”

Card Sort results

The findings were valuable in validating the hypothesis we had as a team regarding users' needs and the tools they felt were helpful, or unnecessary. The results gave us clarity on how users interacted with the tools and whether they found them effective or not.

From a UX perspective, the card sorting exercise was particularly insightful, as it helped pinpoint where users were getting confused. The results were quite conclusive in identifying these areas. Specifically, the dark blue sections below were successful, as they indicated that 15 of our participants correctly placed the cards in the expected categories.

On the other hand, the lighter blue sections highlighted areas that needed further refinement. There were 15 participants in total, and the card sorting matrix below shows where each participant placed the cards within the top-line categories.

User testing results

A total of 12 participants were recorded at this stage, as some users dropped out before we reached this point in the process. The participants are labeled P1 to P12, which corresponds to each individual user. In the left-hand column, you’ll find the tasks that users were asked to complete during the study. These tasks were designed to test specific aspects of the user experience.

The color indicators next to each task represent the users' performance levels. These colors provide a visual representation of how well each participant was able to complete the tasks, giving us clear insights into areas of success and where further improvements are needed.

The calculator was surprisingly well received by the users. I initially thought that many would not see the point and would prefer to use more advanced tools like Excel if they were familiar with them. However, the feedback we received indicated that the pages were too long. In response, I proposed testing a new approach using tabs to address the issue of lengthy pages. The next set of designs included a summary section as an additional tab.

This design was intended as an MVP (Minimum Viable Product), and when budget and needs allow, more functionality can be added. These user tests were focused on assessing feasibility and understanding the users' aspirations.

Conclusion

Testing is crucial, and I don't believe you can test enough. However, in the real world, there are always budgets and deadlines to meet. Ideally, I would like to channel the testing results into more designs and continually optimize them. This project will go through another round of unmediated testing, which fits within the budget allocated for this MVP project. While there’s a lot to cover, I have compiled a full report. For now, I focused on the Calculator pages, as they were highly functional and task-oriented. The feedback has been positive, and I'm looking forward to moving forward with the next steps.

The card sorting exercise was especially insightful, and I plan to collaborate with a copywriter to address some of the misunderstandings that were revealed. Some sections will be moved to different areas of the dashboard to better meet user needs. Terminology was another key area that users found tricky. We tested placing it in a contextual location, but upon reflection, we found that having a single, dedicated space for this content worked better, as highlighted by the testing. For more information, feel free to get in touch—I’d be happy to discuss further.