High user adoption, and hence significant business impact, is the holy grail of virtually all SAP BI projects. All too often these projects fall short of this ideal.
As I started to examine the design principles behind UX (User Experience) design methodology, the concepts and reasons behind them resonated with my experiences in delivering multiple BI projects to business users over the years. Here I will take the five design principles and see how well they apply in the SAP BI space.
The five design principles
- Role based
This is about designing around the user persona rather than the EDW functional logic.
In the world of SAP BI, we’re often far too focused on designing the data model around the back-end (often transactional) structures, and then designing the reports on that basis. Better managed projects attempt to create the data model (often in classic terms the ‘mart’) around a concept of the business process supported, but the very nature of the BI design and build approach means that the mart has to be a complex composite of many different users’ perspectives of the process supported. In effect, it tries to support many personas.
If we adopt a more narrow, focused approach to Analytics alignment to users’ needs, not only do we increase the likelihood that users will engage with the solution more easily, but may find it easier and more secure to manage the access and security rights to the solution.
In the world of SAP BI, we need to ensure that a users’ interaction with a report is intuitive, effective, and error free.
This concept becomes ever more crucial as BI expands from the classic desktop and laptop, mouse and keyboard interactions, to the realm of Tablets, Smart TV’s and PC’s with Smart ‘touchable’ screens (i.e. Windows 8). Small, compact, icon-driven pull-down list selectors may work well on your 22+ inch desktop and mouse, but try using this interface without error on a 10” or even 7” tablet!
The worlds of good UX design and good analytics converge when both are founded on the SAP HANA platform. SAP HANA, as an open application development platform, provides the ideal foundation to develop a modern, highly user-focused interface and the In-memory analytics solution provides the high performance, effective reporting to compliment this experience.
This is one of the most common areas of SAP BI design that strays from user-centric principles.
If the principles of good UX design are applied correctly, reports should be as simple, and as focused to a given user persona, as is realistic. However, too often a SAP BI project approach is to coalesce like-requirements together so that a single report ‘supports’ a myriad of personas and user needs. This is often driven by the desire to minimise the number of objects that need developing and supporting. Consequentially, far too little attention is paid to the end-user experience.
What the project ends up delivering is a complex report with far more characteristics and measures than the main user (persona) needs, as well as a vast array of variables for the user to choose from to focus inon the area of interest for a specific role and task. This results in confused and frustrated users, information overload, and ultimately, unhappy users and poor adoption.
SAP BI projects need to focus far more on getting inside the day-to-day activity of the users, and use the UX approach to ‘de-compose’ the information needs down to the persona (role – task) level. Convergence of ‘like personas’ can drive good data model design, but the reports need to focus on a narrower definition of the persona.
In SAP BI terms, this is extremely relevant, even though it isn’t always given the attention it deserves.
By coherent, we mean that a users’ experience is consistent across all reports, information domains, and, if possible, analytical tools. For example, mandatory variables and prompts should always be at the top of any input list, and in the same order (i.e. date, organisational element, customer, and then product related inputs). Variables (prompts) affecting the same restriction should always use the same words and look identical.
Measures, and comparative date structures, should always be identical where the displayed element is the same. These are, and always have been, good design practices that effective unit testing and peer review should be used to enforce.
This isn’t a term one often hears in SAP BI projects. However, ultimately the BI project team is the supplier, delivering a product to its customers (the users). In the classical customer service world, there is less of ‘customer satisfaction’ and more reference to ‘customer delight’ in the pursuit of larger market share. This quest for customer delight is key to driving user adoption, which in turn can yield a better ROI as the project starts to deliver the business benefits its original business case was founded on.
Achieving ‘delight’ is founded on an effective combination of the design principles. The overall user experience should result in responses like:
“This makes my job so much easier”
“I was never able to see that relationship/correlation before”
“Now I can really understand what is driving that part of my process”
If the trends we see in the way we design and deliver business systems continue to evolve in the direction becoming evident today, then a users’ experience will become a collection of highly focused ‘apps’ enabling transactions with related analytics and measures in a process view of the world. So shouldn’t the reporting and analysis design methodology have substantive aspects in common with that used to deliver the related transactional apps?