The paradox in designing for expert users
When it comes to designing interfaces used by expert users of business applications there is a lot of ambiguity as design has to perform like a consumer application. However, the expectation driving the plan is that expert users will generally find their way around hence they would be far more forgiving to design flaws. This ambiguity can cause major risks for multi-million internal business system transformation programs.
Enterprise applications have been around since the world of mainframes and have been the backbone of all support and transactional processes that need to do be managed at scale and with integrity. These applications form the backbone of all modern enterprises and organisations and the interfaces of these applications require a certain prior experience and specialised skillset to work on them. Alongside increasingly complexity of the applications the interfaces themselves have evolved from punch cards, hot keys and mouse to something which is a multi-sensorial experience involving voice, touch and automated AI bots these days
Reflecting back on the days when I used to develop CICS based green screens, the focus was more on maximising the availability of business process functions rather than how users would interact with it. The same story continued with developing native UI using powerbuilder, .Net or Java where the business process was more important than the user. Building applications using JS technologies and HTML is a very recent event in the history of enterprise applications and here is where modern UX practices come into direct conflict with the approaches for building enterprise applications.
Business process vs the User
In all transformation projects that I have worked, the business process has initially reigned supreme and key stakeholders have always been blindly led into believing that the new User Interface just needs to follow what the business process says. To convince stakeholders to understand users context and factor that feedback is considered as regressive and a friction to achieving delivery timelines. The key assumption behind this is that our users know what they need to do and any change to our business process would be hugely disruptive. However, where research was done with these users of enterprise applications, the feedback revealed asteroid-sized holes in the argument above. Some of the findings have been fairly consistent across different industries and organisations
Even expert users who have been in the domain for years have said that they had to find roundabout ways to deal with the inflexibility in their business processes presented via the interface. Over the years, unofficial and informal best practices have developed in the organisation that are considered the de-facto business process
Non-intuitive UI - We found that even the expert users had loads of guidance sheets and cheat sheets put on their desktop or the work boards to understand which screens or buttons to click for a certain use case. This shows that they rely on help and guidance to navigate through the UI
Feature overkill - this is an interesting area where we found that over the years product teams keep on piling up layers of features often to find that 80% of the application is rarely used. It also doesn’t help that there is never an attempt or thought to redesign the navigation to surface these new features or capabilities.
Design challenges
Design for expert users is like walking over coal, as it is about finding the right balance between designing highly pedantic vs too busy complex screens. Neither extremes are good, as enterprise KPI’s are very different from those for consumer applications. The design team has to consider productivity, efficiency and expertness of the user while retaining the key UCD principles in the design.
There are some material design guidelines and templates available but the risk is that designers often blindly adopt them without considering upfront user research or any specific usability testing. Some of the common types of interfaces where we often see these challenges are
Dashboard - how do we present user choice and present key information to the user without making it overtly complicated with too much information
Reports and visualisation - there is lots of guidance available around how to design and render plots but very little guidance on what to do when their are many plots and visualisations to be presented to a user. It becomes even more complex to design when multi-dimensional data has to be presented to a user.
Case management workflows - back office users sitting in customer support or transaction handling often work with case management whether it is in finance, procurement or HR. This is the domain which is now well served by AI, BPM and automation tools and we might soon be heading to a complete UI-less world where users will trigger workflows on Teams or Slack. However, at the moment the design challenge here is around designing a leaner journey that is simple yet faster for the expert user and doesn’t create additional support overheads for the organisation
Forms - are we digitising or digitalising is the key question for designers. This question often ends up ruffling feathers with the business process side as efforts to change the form ultimately lead to changing the underlying business process. It also has a direct impact on the data strategy that the organisation wants to pursue. These are all comfortable aspects yet very important points that we have found many organisations to almost completely skip in their planning until it is too late. From our user research findings we know that many of these forms are so old that neither the users nor those who analyse the data exactly know the purpose of collecting many of the data points in the form. This shows the need to refresh the design of the form from the perspective of user and business need rather than directly translate a paper or legacy form into a new UI.
The AI effect
As expected AI is proving to be a huge disruptor because it directly addresses the data that sits in these business processes and makes it far more easier to redesign journeys than the traditional design approaches. AI also removes the human element or potentially changes the current user journeys to assisted or human monitored journeys. Hence the design focus is expected to shift towards identifying and crafting assisted journeys that complements and enhances the AI-led journeys. Hence, expert users will need more of journeys where they can get the intelligence and the control to monitor and intervene where required in the AI-driven automated business processes. It is a big paradigm shift for designers and requires the community to develop more understanding of the emerging needs of expert users in light of shifting technology landscape heavily influenced by AI and automation