Launch a Mortgage in a mobile only bank
It's hard to believe that in a little over a year, we launched a mortgage proposition as Atom Bank, one of the latest entrants to the banking sector. As a start up, it was fast paced, agile, and reliant on a quirky UI which had been developed by our external agency partners. The idea was to re-use the patterns created for the Savings Account created in a 3D gaming engine, Unity.
1. Exploration and Research
I started at the beginning. There was a small team who were not yet mobilised and continuing to be involved in operational readiness plans for the launch of our savings propositions, which was actually a blessing in disguise. It allowed us a little breathing space to understand the user journey and the concerns and fears of mortgage applicants.
I got together with the HO Customer Experience, and we started by interviewing two internal users who were, rather helpfully, going through a mortgage application at the time.
Firstly, we were getting a handle on how they describe the process, we noted the language they used, the bits they were unsure about and the pieces of the puzzle which were a bit scary. After all, this is possibly one of the biggest financial transactions you are likely to make in your life, so yes, it should be hold-on-to-your-pants scary.
Then, we worked with the participants to help us fill in the process map. Really simply, we were putting together a mental model of users as they moved through the application.
We used this to start tackling the system requirements document. This helpfully was provided in a 146 page word document...so we did a little bit of BA work to uncover what parts of the process we were able to match to the users mental model.
Main finding from this was: Users have NO idea whats happening with their application and this really worries them.
2. User Story creation
I threw a bunch of stakeholders in a room, armed them with post it notes and sharpies, and we went through the mental models and system requirements. The BA and I facilitated the session and we were able to understand more about the intricacies of the process by going through it step-by-step. This was very useful, and we made a draft of the user stories which we should be able to deliver for MVP 1.
There were a number of constraints as we were working with an underlying 3rd party system, and in hindsight, we should have had their expertise in the room with us as we went, but we came out with some clear questions and some clear stories which we needed to fulfil.
3. Co-Creation workshop
Somehow, I managed to gather a smaller group of stakeholders together to understand some of the key screens in the in-app journey for a mortgage application. We wanted to pave the way for an in-app application, but were currently looking at a simple way to provide updates to the user.
As part of the initial research work, because we uncovered a common theme of users feeling left out of the loop when it came to the progress of their application, and we were limited in terms of the ability to initial make the application, we decided to concentrate on a way to keep the user up to date with what was happening, allowing them to feel in control of the process.
We described the level of information we would have available, and the expectations of the users at this point, and we went screen-by-screen through the application journey, and understood what we wanted to tell the user, and what the user needed to know (sometimes this is different)!
We took A4 pages, handed them out to everyone, and asked them to meet the requirements of the process. First, we needed the user to insert the code to view their mortgage. What would this look like? How would we get the code? Some people put notes, some people drew, some people did it in the style of the app and some people did boxes and arrows. Perfect.
We then went through them as a group to call out the ideas and similarities, coming up with a sort of amalgamation where we took forward the good, and left out the not-quite-there-yet.
It was actually fun, and one of the main reasons you might even choose to do this style of workshop is to uncover peoples pre-conceptions about what they're working on. Getting people talking is one of the easiest ways to learn.
4. Iteration 1
We weren't quite there yet, we were well on the way to needing a bigger picture to talk through. I sketched out a rough journey, and copied the pages we did in the co-creation session. I stuck them together on a wall in kind of an order.
Because we started to understand the elements on the page, I brought in our lead visual designer Ryan to go through the journey. He went through it with a fine toothcomb and was able to look at it with a fresh pair of eyes to uncover some questions I hadn't thought of. It's so useful to do this - I love to work as a pair from a UX/UI point of view - it helps you both understand the task, and offer different perspectives of the problem.
5. Updating the Team
We switched around teams a bit as people left, new people came in which is pretty expected in the startup space, but we needed to go back over a few things, and as we were working as more of an agile team, we were light on documentation, but had a journey we were now starting to put together with the current styling assets.
One interesting part of widening the net was that we have started to really get into the detail of other areas needs. For example, for those who have ever worked in finance, you know there is a hell of a lot of legalese-mumbo-jumbo that needs to be shoved in somewhere. But always, at the heart of this mumbo jumbo was a well intentioned sentiment, to try and protect the user.
Our fellow Risk and Compliance team let us know we had to put a message;
YOUR HOME MAY BE REPOSSESSED IF YOU DO NOT KEEP UP REPAYMENTS ON YOUR MORTGAGE
on every page. With the interface as it was, this would look awful, and probably go some way to induce a compliance style 'banner-blindness'.
In the spirit and tone of voice of Atom, I came up with the idea of a 'FUN Compliance Quiz'. This was a question, which we could add into the app which asked 'Why might your home be repossessed'? It had a couple of fun answers in, and obviously the correct answer of "If you do not keep up repayments on your mortgage".
I was surprised to get away with it, but we actually met the requirements even better than a simple message, as we were able to check the users understanding of the regulations.
5. Refining and testing
So by now, we had pulled a large chunk of the journey together; for logged in and logged out users. It was looking pretty slick. Alongside some constraints, we had been able to come up with a credible, understandable journey which addressed the main concerns of mortgage applicants.
Once a user has been to a broker and they have recommended an Atom Mortgage, they are given a code. They can share the code to their app, it will pick it up and link their application to their mobile app.
Once they have the code, they can view the details and will be notified by push, SMS or email (whichever their preference) that their application has moved on a stage.
We started to test this through with our Business Development Managers, as it was obviously important that we have access to some brokers who were going to be part of the panel to sell our mortgages.
We made a prototype on principle, which is excellent for smaller journeys which are full of animation, and made communicating this with our devs really easy. They knocked up a working version really quickly, considering we were re-using design elements already built, within a week, the flow was put together and some further questions answered.
Now we were equipped with a working version, I wanted to test test test! I sat down with about 10 different internals to do some guerrilla testing and again, uncovered some further expectations, some additional things we hadn't spotted and a couple of tweaks to the copy as we seemed to be using a little too much jargon.
We then set about calling in some real users to go through the main journey.
I was in the middle of setting up Atom's in house user testing whilst this was going on, so I was excited to be able to get to speak to real people outside the company. We tested a couple per day, and tweaked the prototype in between. Some users we were able to sketch out with them a different way to display things, some we found confused by the interface but it was good news, most people kind of got it.
Oh lord we weren't finished yet. Part of the fun of working in a startup for me is that you really get a war-time-esq feeling. Like, we're all in this together guys, we need to make this work. Cue long sessions through the summer eating ice lollies outside whilst the Intermediary Team Support Manager Cassie and I came up with the draft of the copy.
Long sessions with various 3rd parties helping to iron out kinks, helping them to prioritise what they needed to deliver and by when. Helping out with refining the user stories, presenting the updated journey to various committees, putting together showcases to take to third parties and working with our marketing team to refine the copy - just par for the course.