Property

Turning scattered data into clear, actionable insights, empowering claims teams and analysts to respond faster and smarter.

Role

Sole UX and UI Designer

INDUSTRY

Insurance

Company

MS Amlin

The challenge

When a major catastrophe occurs, such as an earthquake or hurricane, insurance claims teams and portfolio analysts struggle to quantify the financial impact. The challenge is even greater in specialist insurance, where a single policy can cover multiple properties rather than individual assets.

Currently, insured properties are stored in tables in spreadsheets and policy administration systems, making it difficult to analyse the exposure of insured assets in recently impacted areas. While live geographic hazard data is made available by local governments, such as wildfire zones in California or flood and earthquake risk maps from Japan, this data was not being leveraged effectively within the business.

Background

The development and data science team had already began an effort to create a tool that could capture catastrophe-affected policies. While the tool met the user's basic needs, there had not been a design approach to creating the tool, making it hard to navigate intuitively, inconsistent in it's design and lacking a purposeful user journey to help users achieve their goals.

The pre-existing tool

1

Unintuitive search that doesn't meet the users' requirements. Users requested a filtered searched with predefined parameters to avoid errors and minimise the need for typing.
2

Table repeated.
3

The information presented in this table is the most important in the tool. Users wish to cross reference the table with the map, and must scroll between and retain the information. This is inefficient and could cause avoidable mistakes to be made.
4

The total sum insured is only given per policy, which undermines the goal of the tool to easily perform analysis on policies as a group. This solution requires users to manually calculate the aggregated financial data from the group, adding an unnecessary manual step.

1

No navigation used. Users may find themselves on the wrong path and not have the opportunity to correct their course.
2

Key real estate is taken up by low-priority information.
3

Jakob's law is not observed; For example this download button does not follow UI patterns expected from users, both in it's placement and it's design.
4

A single row table is not the best presentation for users to ingest information.
4

A single row table is not the best presentation for users to ingest information.
5

Table heading titles are inherited from the database, featuring underscores and technical terminology. This decreases legibility and affects the ability for the text to be translated, which is a key requirement in this project.
6

The data provided has not been refined to consider the users' needs and give the most valuable analysis.

Research

I began by speaking to the stakeholders to find out exactly what they needed their teams to achieve by using the tool. I was able to gauge the following:

There was no easy way for insurers to assess the financial impact of large-scale disasters on insured properties. Analysts relied on manual processes and disjointed data sources, making the rapid decision-making required in these scenarios near impossible.

Affected users

Claims handlers needed a way to quickly estimate the financial impact of an event.
Portfolio analysts required a clearer picture of exposure across all insured assets.
Underwriters needed insights to refine future risk assessments.

Business goal

Improve efficiency in claims handling and exposure analysis.

User need

Access to an interactive way to visualise insured properties and risk areas.

Having thoroughly defined the goals of the business, I sent out a spreadsheet for existing users to document their requests for a design refresh and record recurring issues. Designing for Japanese users requires a different approach to designing for Western users and illuminated some extra considerations.

Japanese users are found to be more risk averse, especially when making financial decisions. There is therefore a general preference for all data to be made available.

Japanese users have a higher context tolerance, with a strong preference for dense information displays rather than minimalist interfaces.

Standard search functions wouldn't work due to Japanese language complexities, requiring a structured filtering system instead.

1

Navigation added to aide users in finding their way around the site and restarting their journey.
2

Search has been designed to include user's requirements to refine search results through predefined prefectures and cities. This helps to address users' reports that many towns in Japan share a name, making this level of granularity important to achieve the correct result. Users also expressed a preference to avoid typing where possible, due to the extensive Japanese alphabet. In order to prevent unnecessary typing, dropdowns are provided.
3

Aggregated statistics are calculated automatically, prioritised and highlighted at the top, eliminating the need for users to perform extra tasks outside of the tool.

1

The page has been given a title to signpost to users and minimise the possibility of errors being made.
2

Information previously stored in a single row table is redistributed into an easily digestible card of information.
3

At users' request, the risk score has been split into different catastrophes to allow for better risk understanding. I also added a comparison to the portfolio average, as well as semantic colours, to give users an immediate understanding of the statistic.
4

The map with the plotted location and information regarding the property has been included on this screen. This gives users a better understanding of the property's context and ties it back to the first map screen.

Outcome

The redesigned tool has been deployed to the property team at MSIG Japan and is used to support weekly reports undertaken by portfolio analysts. Users have reported high satisfaction levels in being able to perform complex analysis and having a stronger awareness on the impact of localised catastrophes on claims.