How To Address The Issues You Cannot Know
Software development for the web is rather straight forward these days – especially with agile setups. We have experienced project managers, business analysts, feature architects, back-end-, front-end or full-stack developers, and we develop use cases, user stories and use best practices and frameworks to put everything together.
It’s still quite a bit of work, even with the help of automation and templates. You need to be thoughtful with the business logic and choose the right technology stack, depending on where and how your software is supposed to work, often integrated in other systems that cover issues like login security and data storage.
Still, users are often dissatisfied with the result. If everything is set up right, how come we keep running into issues with users we didn’t see coming?
A lot of it has to do with the user’s experience. As surprising this may sound – a great share of the user experience actually depends on different things than the user interface alone.
A Perfect Storm
2021 was the second year with Covid. I had taken more responsibility as Head of UX at DaziT. By then we had got used to working together remotely, for the most part. Once in a while, issues, big and important enough, meant a video call wouldn’t suffice. So we met in person, with the necessary precautions in place.
On a late summer afternoon, I was in a large meeting room, together with the programme leadership and about 20 additional people, ranging from top positions, unit leads, heads of architecture and development, to consultants and people outside the DaziT programme, namely the head of the Federal Office for Tobacco and Beer Taxes at the FOCBS.
I’ve had my share of meetings with a palpably tensed and anxious mood. This was a crisis meeting on a different level. The head of the Office for Tobacco and Beer Taxes was visibly nervous.
He opened the meeting explaining:
This was the transition year for the digital transformation of the Office for Tobacco and Beer Taxes. All beer brewing companies were expected to have created a digital account by August and completed their tax report digitally by September. That had not happened. By the end of August, 90% of all beer brewing companies had not completed the process of registration, let alone entered their tax report on time.
Confusion, Irritation and Frustration
The question was: Why? All companies had been invited early enough. A PDF with clear instructions had been sent out. The software worked as it should – it had been tested several times.
What had happened? Over the next 40 minutes, the DaziT project team was quick to explain their findings in more detail:
- The tax report Web app had been introduced in spring 2021
- Most users had dropped out throughout the process of registration, which had to be done separately, on two different platforms
- Users were uncertain where to begin, and insecure with every step of the process
- They often didn’t even know which legal entity in their company was able to do what part (admin and user rights)
- What was supposed to be a streamlined process, simple and effective, turned out to be a nightmare for customers (individuals and beer brewing companies)
- Some companies are big enough to have a custom API that directly connects with the Office for Tobacco and Beer Taxes. They had already transformed their processes with direct help from BIT, (the Federal Office of Information Technology, Systems and Telecommunication, FOITT).
The problems occurred with the majority of brewing companies – often private individuals or small family enterprises. Thousands of these breweries are registered in Switzerland, where everyone who is brewing more than a limited amount of beer per year needs a license. *
I asked what we knew about our customers. It turned out there had been no customer research and the applications had been built as island solutions, without any crossover of information. They were simply linked to one and another and users were supposed to figure out which step to take next, in which order, where to find all information, and how to find back to the application they originally wanted to use: The tax report entry application.
A Way Out Of The Maze
Luckily, there was one set of information we did have by then: hundreds of complaints, customer support calls, logs of calls and e-mail exchanges over the past five months. It all painted a clear picture: users were confused, irritated and frustrated. Most of them wanted to give up and naturally, the customer support team struggled to guide them through every step.
There were too many variables at play. Too many branches of interaction where a customer could get lost. Too many factors the customer simply didn’t know, because no one had told them what the next step was for and what would come after.
This was the pivotal element: clear and straight forward guidance and bringing the customer mindset on the same page, so they have more confidence in knowing what each step does and what they can expect as an outcome.
I suggested to revamp the whole path, change the UI to combine steps logically, be more consistent, more recognisable (each page has a familiarity to it, confirming to the user that they’re still in the same process), be more user-friendly, and, overall, provide help and support with every single step.
A Straight Path Forward
Had there been proper user research beforehand, we would have known about the concerns and frustrations of the customers. We would have been able to evaluate their savviness, when it comes to accomplishing official tasks online. It’s safe to assume most people who enter the brewing industry do this out of passion and rarely out of an interest in Web technology.
The problem was, we had developed software with proper UI, with a user experience that worked fine within the tax report application. But combined with the registration processes and the different legal entities involved, it all became too much and quite confusing. Only when you work in IT at the Federal Administration, you may have a clue how eIAM, the ID registration and management system, works. (Side note: eIAM has been improved since 2022.)
Over the coming two weeks we developed a consistent and harmonic user flow that made sense to users. Our solution aimed to fulfil a number of goals:
- From the start, users would get a clear idea of the tasks at hand: User account creation, registration of the brewery, filling out the personal details, assigning roles
- We would inform them using a clear, conversational language, letting them know what they were expected to do next
- They would be guided at every turn of the way
- Familiarity was a big factor: Each page should be a step, congruent with the previous one, even if the summary of tasks took place on different platforms
- The design should be friendly, accommodated by a help drawer that could be called up on any page
- Illustrations should help with guidance and indicate where the user stands in the whole process
We created an interactive prototype using Figma and tested it with four different user types. The results were promising: At no point did anyone fall out of the process or was entirely lost. We found a couple of small gaps and clarified the language. But overall, the results were stunning. It took about four weeks to implement the front-end changes and after testing it again, the changed process was implemented and rolled out.
Understanding Motive, Intent, Flow And Execution
If you’re looking at a project only as software development, just by summing up the parts of its content and putting together a plan how to build it, without regard for how it will integrate with other software, or fitting into people’s lives – you have no way of knowing what issues you are going to face. You may be meeting or hurting people’s expectations and your customer support lines will be swamped with complaints.
As UX Designers we are trained to look at these things. But we cannot foresee what we don’t know. To be able to address issues before they appear, we need to do user research and look at entire chains of interaction of users coming from various points in life, and map the information we find. Customer Journey Maps enable us to draw conclusions, which we can then apply in software development. What may not matter on a single page, where you fill out you registration details, may matter in the whole flow, the path a user has to take to get from A to Z.
We need to get a better image of what users are thinking, before and while going through a process. We need to learn about motive, intent, flow and execution to be able to adjust the flow of events.
Neglecting user research creates a lot of extra stress. Eventually, these issues will catch up with us, and that will cost us nerves, time and money.
This is where Human Centred Design (also within SAFe) can help you solve issues you didn’t see coming, even if you had the best intentions and all the tools to create great software.
Annotations
* As long as beer is sold and consumed free of charge and a certain amount of output is not exceeded (400 litres per private individual or 800 litres per year for clubs), there is no beer tax to pay and therefore no registration as a brewery is necessary.
This article was first published internally on the Adnovum intranet.