Fixing the root cause

We started by focusing on the customer side, since automation depends on receiving clean and accurate data.

The first issue came from a section called Customer Partnership. Customers were always asked to state which union or partnership they belonged to — even when we already had this information in our systems.

Many didn’t know the answer, leading to confusion and unnecessary manual verification by advisors. Worse — advisors spent up to 5 minutes per case checking or correcting partnership data, which wasn’t even part of their responsibility.

After several research sessions with advisors and a legal review to validate data handling, we redesigned the logic:

• The partnership question now appears only for customers whose data is missing.
• When customers enter new partnership data, it’s automatically sent to the correct internal team, not the advisor.

Impact: Advisors saved several minutes per case, and customer confusion dropped immediately after rollout.

In the example below: on the left, the partnership question sits inside the general information section. On the right, “Union and partnership” has its own section, a logical decision since it doesn’t belong in general information, with a simple dropdown that lets customers find their partnership.

Reducing miscommunication through smarter forms

Next, we tackled the loan change request form, where customers specify why they want to modify their loan.

Change requests were divided into simple changes (like changing payment date) and complex changes (like lowering monthly payments). Depending on the request, customers always had to state the reason for the change, and nearly 1/3 of all customers, 28.5%, were choosing the option “Other”, forcing advisors to manually read long comments to understand the actual intent.

That single choice made automation impossible.

To fix this, I partnered with our business analyst to analyze a few thousand “Other” cases. We manually read through comments and discovered that most confusion came from terminology — what felt obvious to us internally wasn’t clear to customers.

We reorganized the form:

• Reworded unclear options into plain language.
• Added missing categories customers frequently described.
• Highlighted specific common reasons for changes.

After releasing the update, the percentage of users choosing “Other” in 3 months dropped from 28.5% to around 17%. This directly enabled us to automate more of the advisor workflow and reduce manual investigation time.

Besides this, we noticed an increase in other categories like “Paying other debt”, which is always a rejection when it is stated as a reason for the change.

Building the advisor platform

Once the customer experience was improved, we turned our focus to advisors. At the time, advisors were juggling multiple disconnected systems: customer portals, Excel calculators, email threads, and document generators.

Together with another UX designer responsible for loan applications, I led the design of a new Change Request System for advisors, built to align visually and structurally with the loan application tool. Our goal was to reuse familiar patterns and design components to keep development efficient and maintain cross-system consistency.

Key decisions included:

• Aligning API connections so all customer inputs now flow directly into the new advisor system.
• Replacing advisor emails with in-system case communication.
• Structuring the UI around the logical order advisors follow, so data scanning and case handling became faster and clearer.

Dozens of co-creation and validation sessions were held with Finnish advisors to challenge their workflows and refine our design until it felt effortless and reliable.

Scaling automation and finalizing the system

By the end of the three-month phase, we had delivered two key outcomes:

• Cleaner data from customers, enabling automation.
• Streamlined advisor tools, reducing dependency on outdated systems.

The advisor system is still under development, but early results already show that the new structure improves efficiency, reduces friction, and saves significant time.

As these flows continue to roll out, we’re expanding our design logic into other Nordic markets, gradually standardizing how Danske Bank handles loan changes across all regions.