ING Belgium / NN Group

Car insurance

My Role

UX Design Lead
Expert reviews
Workshop organiser
Stakeholder management

Team

Insurance

Product Owners

Alexandru Boboc (1)
Cedric Hoste (2)
Aurélie Heyninck (3)

Lead/PAL

Chantal Hansoul

THE SCOPE

and a bit of context

ING Bank in collaboration with NN Insurances had already developed a user flow that allows users to subscribe to a car insurance online. However, the user flow was several years old and it was not performing according to expectations any more or make use of the latest UX patterns developed by ING.

I was contracted by the ING Insurance tribe to build, redesign and improve this user flow.
As a side note, at that point in time, INGs Insurance tribe was perceived as having a very low UX maturity level, so one of my tasks outlined by INGs BE/NL Head of UX was to evangelize UX methods and raise the overall UX maturity of the tribe.

Business requirements

  • Migrate from web 1.0 to 2.0
  • Use ING Flow (ie the design system of ING)
  • Mobile friendly and mobile first approach
  • WCAG AA compliance
  • Strengthen the strategic partnership between Ing and NN
  • Reduce the number of screens to maintain
  • Increase sales by 2000 contracts/year
  • Increase conversion rate from 18% to 30%
  • Decrease abandonement rate
  • Less CT calls (ie. Support phone calls)
  • Less visits in ING branches

User requirements

  • Guidance in the flow
  • Eliminate identified pain points
  • Decrease cognitive load
  • Offer a digital experience at least as good as the competition
  • Allow the user to quit and then resume the flow
  • WCAG AA accessibility compliance
Original-ING-Car-insurance-start-screen

THE PROCESS

and my role in it

As opposed to Project Atomium, we started with a functional user flow that had previously shown promising results. However, as we entered the dynamic landscape of 2023, we realized the need for a significant transformation. The user flow, though once successful, no longer met the elevated standards expected in the present era. Moreover, it had yet to embrace the latest version of the cutting-edge design system developed by ING.

Step 1: Evaluate the current flow

Besides the obvious, technical issues – like not utilising ING Flow – there were some glaring issues.

  • The flow is running in a modal window.
  • Accessibility issues
  • Illogical information architecture
    • ordering of the questions
    • overall screens order
  • Weird, apparently out of place questions
  • Business centric instead of user centric
  • Low conversion rate
  • High abandonment rate
  • No measuring plan is implemented (ie. proper analytics)
    • We know the abandonment rate between screens but we don’t know why
    • No event tracking tied to buttons
    • Improper Adobe Analytics dashboard
    • High probability of analytics errors (ie. Improper implementation)
  • Lack of consistency between ING Insurance flows

Step 2: But what is the competition doing?

A competitive benchmark was done to evaluate and compare the user experience of the car insurance products with our direct competitors. It involved conducting a systematic analysis and assessment of various UX aspects:

  • UX/UI
  • Behavioural design
  • The type and order of questions asked
  • The layout of the login screen
  • Promotions/X-sell

Therefore, the following companies and their services were analysed

  • Yuzzu
  • Corona Direct
  • Ethias
  • BNPPF
  • KBC
  • Belfius
  • Argenta
  • Yago
  • Progessive
  • ANWB
  • Inshared

Step 3: Who, what, why, how?

Stakeholder map

From a UX perspective, a stakeholder map is a powerful tool that helps identify and categorize the different groups of people who have a vested interest in the success of a particular product or project. This map is used to visualize the relationships between stakeholders, their level of influence, and their interests.

The stakeholder map for UX includes both internal and external stakeholders such as users, designers, developers, managers, customers, investors, and regulators. Each stakeholder is assigned a specific category, such as primary, secondary, or tertiary, based on their level of influence and importance to the project.

The primary goal of creating a stakeholder map in UX is to design and develop products that meet the needs and expectations of all stakeholders involved. By understanding the interests, motivations, and goals of each stakeholder, designers can create user experiences that align with the business goals and objectives of the project.

Moreover, a stakeholder map in UX can help identify potential conflicts or issues that may arise during the project’s lifecycle. This allows UX professionals to proactively address these challenges and ensure that the final product meets the needs of all stakeholders involved.

In conclusion, a stakeholder map is an essential tool for UX professionals, providing a clear understanding of the stakeholder landscape, helping to identify potential issues, and creating products that meet the needs of all stakeholders.

In the course of a few meetings with the tribes, the mapping was done.

The stakeholder map for Car insurance

Current flow mapping

To my surprise, there wasn’t any comprehensive user flow map. From both the user’s perspective or the intricate processes involved, it became evident that mapping the current flow was an essential task we couldn’t overlook. It was clear that by creating a detailed map, we could gain a deeper understanding of the user journey and uncover potential areas for enhancement.

With a collaborative spirit, ING and NN joined forces, actively involving all stakeholders. This shared effort allowed us to have diverse perspectives and tap into a wealth of collective expertise, ensuring a holistic and comprehensive approach to mapping the user flow.

The user flow of the original car insurace

Time to get everybody in one room

For those unfamiliar, a design sprint is a structured process that brings together cross-functional teams to solve complex problems and create innovative solutions in a short amount of time. It involves rapid prototyping, user testing, and iteration, all within a compressed timeframe.

By organizing a design sprint, we were able to bring together the right people and resources to tackle the challenge at hand, while also fostering a collaborative and creative environment. It was an exciting and exhilarating process, one that allowed us to explore new ideas and test them quickly.

Ultimately, the design sprint was the perfect solution for our needs, allowing us to create an innovative solution that was grounded in user needs and backed by rigorous testing.

The Scoping Canvas

Before diving into the design sprint, we needed to ensure that we were all on the same page and had a clear understanding of the scope of the project. That’s where the Scoping Canvas came in.

A Scoping Canvas is a tool that helps teams define the scope of a project and set clear goals and objectives. It’s a collaborative process that involves bringing together stakeholders and team members to brainstorm ideas, identify key challenges, and map out the project’s key components.

By using the Scoping Canvas, we were able to identify the key challenges and opportunities of the project, and set clear goals and objectives for the design sprint. It helped us align our thinking and focus our efforts on what was most important, setting us up for success in the design sprint.

Scoping canvas for the Car insurance project

Step 2: Plan and execute the design sprint

  • Conceptualise a new user flow
  • Scoping canvas
  • Wireframing
  • Detailed designs
  • Legal & Compliance validation (1)
  • Iteration of feedback
  • Legal & Compliance validation (2)
  • Prototyping
  • Testing preparation

The first priority was to look at the as-is user flow and create an new, optimised and user centric flow. Recognizing the indispensability of all the questions within the existing user flow to enable NN to provide insurance, I identified the primary area for enhancement as the information architecture (IA) of the user flow.

Consequently, I embraced and advocated for an alternative approach that centers around logically grouping the questions, with each section strategically establishing a mental framework for the user. This approach effectively reduces cognitive load, allowing for a smoother and more intuitive user experience.

Over the course of multiple days, our team dedicated focused efforts to shape the desired user flow, with a specific emphasis on the “happy version” – an ideal scenario where everything goes smoothly for the user.
By implementing the user flow I proposed, we were able to refine the overall user experience and fostered an active and productive dialogue between the ING (UX) and NN (Technical) teams. During this collaborative process, we identified a couple of sections within the flow that required improvement:

  1. Determining car age: One area of concern was the inability to ascertain whether a car is more or less than two years old when the registration number is known. This limitation prompted us to explore potential solutions in order to minify the friction of the flow.
  2. The information we are able to extract from the VIN number (registration number). This side of the flow is covered by an API called INFORMEX. This API is able to pull only some information, that doesn’t break the EU GDPR rules. The first registration of the car is sadly, not an information we had access to.

Over the course of multiple days, our team dedicated focused efforts to shape the desired user flow, with a specific emphasis on the “happy version” – an ideal scenario where everything goes smoothly for the user.

By implementing the user flow I proposed, we were able to refine the overall user experience and fostered an active and productive dialogue between the ING (UX) and NN (Technical) teams. During this collaborative process, we identified a couple of sections within the flow that required improvement:

  1. Determining car age: One area of concern was the inability to ascertain whether a car is more or less than two years old when the registration number is known. This limitation prompted us to explore potential solutions in order to minify the friction of the flow.
  2. The information we are able to extract from the VIN number (registration number). This side of the flow is covered by an API called INFORMEX. This API is able to pull only some information, that doesn’t break the EU GDPR rules. The first registration of the car is sadly, not an information we had access to.
  3. Home ownership inquiry: Another section in the flow involved asking the user about their status as a home owner, as this factor contributes to the overall “rating” of the user. However, we recognized the absurdity of this question and we initiated a sub-track with NN to see if this question can be removed.

Explainer: In this context, the age of the car plays a pivotal role, shaping the unique treatment and considerations provided to users. The distinction lies in whether the car is within the range of approximately 2 years old or older. A car exceeding the 2-year mark is deemed a “second-hand car,” which entails distinct premium rates and specific information requirements compared to those associated with owning a car less than 2 years old.

The user test

and the conclusions

Due to a scheduling issue, we weren’t able to test the flow in both French and Dutch. As I mentioned in previous case studies, at ING we preferred to test both languages, because we found out Walloons think differently than Flemish people.
The test was done in French only.

  • 6 participants
  • 3 women, 3 men
  • Ages between 23 and 61 years old
  • 4 ING customers, 2 non-ING customers
  • Mix of digital and non-digital users
  • Mix of users with a new or used car

The test proved to be successful, with the participants saying the experience was very intuitive and straight to the point.
There were, however, some gripes the test highlighted and it was enough for the business team at NN to alter and push for some optimizations.

Home ownership inquiry:

While initially intended to gauge user trustworthiness, our user testing revealed the obvious. Users were not only surprised by the question but also found it to be an unnecessary distraction from the main flow’s purpose and an indiscretion. Armed with strong suggestions and compelling arguments, I presented my case, and it became a pivotal moment for NN’s decision-making. Recognizing the importance of maintaining user focus and minimizing disruptions, NN made the choice to eliminate the home ownership inquiry altogether.

The product fulfillment screen:

Given the nature of the product, the final screen also has requirements for a follow-up from ING and the user. This pivotal moment called for seamless coordination, involving the exchange of crucial documents that needed to be signed, received, or sent. However, during our user testing sessions, we discovered a slight disconnect. While users did notice the presence of a policy registration number, the precise steps and actions necessary for completing the process were not as transparent as we had envisioned.

Conclusion

and the next steps

As UX designers, we understand the significance of user testing as a pivotal step in the design process. However, it is important to acknowledge the inherent challenges associated with accurately interpreting and effectively applying the results obtained.

In the specific context of the user test conducted for this particular user flow, I am pleased to share that it was a success. We were fortunate to receive valuable feedback from users, which shed light on aspects of the design that resonated positively with them. Equally important, it also highlighted areas where further enhancements were warranted.

With these insights in hand, our dedicated design team is now poised to embark on the next phase of the design journey.

Accessibility issues

While ING (and most of the banks now) wants to cut the support calls and visits to branches as much as possible, insurance business does have its particularities. In this case, letting the user know it can contact the insurer is a must-have.

The reasoning behind this is that in the case of insurance, users prefer to see they can get help in any circumstance. It adds to the credibility of the product and increases the comfort levels of the user.

However, the solution that was designed by the tribe has several issues, besides accessibility

  • Accessibility: the contact section is below the CTA, so screen readers will mention it after the CTA was mentioned
  • ING Flow: The rules of ING Flow (part of the ING design system) clearly specify nothing can exist below the button
  • Scalability: even if now ING and NN offer just a phone number, in the future other methods will be implemented (eg. chat, video calling, Whatsapp etc). There is no room for any of this in this iteration.

There were no resources to change this, as it’s a transversal approach for all insurance flows, but it will be solved in a future iteration.

The product fulfilment screen

As highlighted by the user test, users don’t really understand what the next steps are. his can be very risky for the user, as in the case of an insurance claim, they may have the unpleasant surprise of not having all the paperwork in order, which can result in a rejected claim.

As far as I know, in the future an online signing feature will be implemented and users will be able to 100% complete the flow in one go.

Development of an OCR feature (Optical Character Recognition)

Even though the vast majority of users know where to find their car registration number (VIN), they still have to manually type the serial number. Given that 90% of the cars have the VIN number in the windshield, pointing the phone and letting it fill in the serial number would be a very nice feature.

Anything to eliminate friction in the flow!

Challenges

and obstacles in the road

Even though it was a pleasure working on this project, as it often happens, there are various external challenges that need to be overcame.

  • Extremely unorganised communication between ING and NN
  • “Surprise” requirements that pop up in the middle of the “solution fit” stage and after, potentially altering the flow and its logic
  • Diverging opinions developed by working in silos
  • Tedious approval process and micromanagement: the PAL wanted to approve everything, from colours to UX patterns
  • Product owners changed three times during this time
  • Sketch to Figma migration: ING started a migration from Sketch to Figma, changing its developer delivery process altogether, amongst other processes
  • Dev team involved very late in the conversation
  • Rushing for launch and not abiding by the ING way of working. It was a constant struggle to do WCAG validation, test, and iterate