Rethinking the business loan application

A business owner who has banked with Westpac for years applies for a loan. They’re asked for their transaction details, financials and documents, all over again.

Westpac wanted to make it simpler to borrow online, from application to drawdown. An external review of the existing end-to-end process had counted more than 4,300 steps, and found the main friction point: capturing the application. Staff keyed information in by hand several times, which caused errors. Customers were asked for information the bank already held, again and again.

This was a future-state design project. In Phase One, our job was to show what a better application could look like. That gave Westpac something real to test with customers, and a set of patterns it could use in other lending journeys.

The prototype tested well, and “save and continue later” helped with a long application. But the biggest finding was about trust. Many customers didn’t want to share personal and financial details about their business online until they’d spoken to someone at the bank.

Two smaller findings pointed the same way. The help panel with live chat was mostly ignored, because people thought it looked like advertising. And some non-Westpac customers worried that “save and continue later” meant they had committed to the bank.

This changed how we thought about the proposed journey:

• A person, early. A fully online application wasn’t what many customers wanted. The journey needed a clear way to talk to someone before asking for sensitive details. Our “nature and purpose” questions already worked out early whether a business could apply online or needed a banker.
• Help inside the flow. Guidance had to sit inside each step, not in a side panel that looked like an ad.
• Clear commitments. Saving an application needed to say plainly that it didn’t sign you up to anything.

It also opened a wider discussion at the bank about where digital helps, and where people need a person.

I co-led the Phase One concept with a service designer. I led the UX work across research, journey design, prototyping and testing.

• I owned the concept, the research with bankers and staff, the journey and its variations, the high-fidelity screens and the prototypes.
• I worked with a content designer on prototype content, an external UX agency on testing, and product, technical and lending teams on what was feasible.
• I brought in senior leaders to watch customers use the prototype, so they could see the experience through customers’ eyes.

I wanted to understand the problem from the banker’s side as well as the customer’s. So we built experience maps from what relationship managers and staff told us, then linked every pain point to the journey step and the system behind it.

That changed the conversation. With pain points sitting next to the systems they came from, you could see how much of the customer’s effort was really the bank’s own process, handed over to the customer. One banker put it simply: “A set of financials will get inputted manually 5 times.”

That question gave us a simple way to sort every piece of information. Some of it the bank already held. Some it could get with the customer’s permission, from public records or accounting software. The rest only the customer could tell us, so those were the questions we kept.

Workshops with service designers, technical teams and product turned that into four features. NZBN lookup finds the business and its directors, so the customer confirms instead of typing. Accounting integration with Xero or MYOB brings in financials. Nature and purpose logic works out early whether a business can apply online or needs a person. Conditional offers give an indicative offer based on the answers and data quality.

We benchmarked banks and fintechs in New Zealand, Australia and the UK, mostly on form structure and logic. Then we compared two ways to shape the application.

We chose the conversational flow because it takes a lot of weight off a long application. I designed the core journey and its variations for different kinds of business, then built the flows and screens in our design system.

We ran four days of in-person testing with business customers, changing the prototype between sessions. Senior leaders watched from an observation room. They saw for themselves how customers responded to the future experience, including where customers held back from sharing business details online. They also saw how each design decision was tested, and changed when it didn’t work.

After each day, breakouts and dot voting with the leaders decided what to fix next. That level of involvement was new for the bank.

The full Phase One concept did not ship. The internal systems and processes needed to support it weren’t ready. But the work carried forward in other ways.

• A customer lab. The project helped establish Westpac’s Customer CareLab. Design leadership used it to show the value of testing with customers.
• Reused form patterns. The conversational form patterns went into Westpac’s design system. They were reused in simpler credit card and overdraft applications.
• A reference for later work. The research, journeys and findings became a reference for later lending and application improvements.

In Phase Two I explored multi-director and joint applications, and an “Application Hub” concept. We tested these with customers, but technical limits paused the build.

I went in thinking about a form and came out thinking about trust. For a decision this big, people want a conversation first, and guidance has to sit inside the flow, not in a sidebar that looks like an ad.