I was fortunate to attend an SAP Screen Personas Customer Event in London on April 24, 2013 run by theSAP Imagineering team. This was part of my investigations into how we might improve the experience for many of our SAP users. I was expecting a demo of the capabilities of the software and the opportunity for some hands-on time to investigate its capabilities for myself. I got both of those and more besides. This blog is a quick summary of the content from the day that was important to me. For more information, including FAQs, how-tos and video demos, look at the SAP Community Network Wiki - SAP Imagineering - SAPScreenPersonas wiki.
The first few sessions of the day addressed not just Personas but UX issues more generally, and gave lots of good advice about how to approach a UX improvement project. A few interesting things from those sessions:
- UI design is a design skill and not a technical skill. If you are embarking on a UX improvement project, make sure you have design skills on the team. This is no different from using PowerPoint, where you can do lots of clever things very easily with no real technical skill required, but you can just as easily create awful PowerPoint slides as nice ones. This certainly resonates with me. I'm not short of technical skills, but my design skills are almost non-existent! My PowerPoint slide decks are almost universally bad
![]()
- Don't focus just on the casual users. When you do surveys of user happiness with SAP, the casual users are generally less happy than professional users. If you use SAP through SAPgui day in, day out, not only do you get used to it, but you appreciate the power the comes from the UI complexity. If you use it once a week, you will always struggle to remember how a transaction works. It is tempting, therefore, to focus UX improvement effort on casual users, but that is usually a mistake. Not only are the overall productivity gains higher when you improve things for professional users, but if you neglect them the professional users feel neglected and end up less happy as a result! Casual users should probably be your first target, but don't forget the heavy users also.
- Speaking of productivity gains, while that's an obvious thing to measure when looking at the success of a UX project, it isn't the only thing, and maybe even not the most important thing. Think carefully about how you are going to measure the success of a UX improvement project.
- You should have a long-term strategy for UX improvement, and not just a one-off project to improve the obvious few transaction and leave it at that.
So, to Personas itself. We were given a demo of the current version of Personas, version 1. We saw how to turn the standard SAP login screen (transaction SMEN) into something like a launchpad for users that need to use just a few transactions. We also saw how to take an existing transaction such as VA01 and greatly simplify it by removing unnecessary fields and tabs, and by merging content from multiple tabs into a single one. Alongside this there's functionality for changing the visual appearance with different colours, backgrounds, images, etc. so that you can end up with a screen that looks nothing like the original starting point, and even nothing like SAP if that's important to your users! There's also a scripting capability allowing you to perform multiple actions with a single button click, including chaining transactions together. The wiki mentioned above has links to examples of all of this.
One thing that concerned me about Personas was its longevity. From what I knew of it technically, it could easily have been a product that would be around for 2 or 3 years to address a specific requirement and then quietly fade away. Investing in such a thing obviously isn't very appealing. After hearing about the plans for future versions, those concerns have largely gone away. The current version has an architecture like this:
In essence it takes the output of the ITS webgui as its input. It allows you to modify that, and then renders the result in the client web browser via silverlight. It feels very much like an afterthought, an independent layer that could go away as quickly as it arrived, and that's what prompted my concerns over its longevity. A consequence of this is that version 1 of Personas does not affect the layout of screens in traditional SAPgui or the webgui. It is, in effect, a third gui. Note also that the Personas functionality exists almost entirely in a stand-alone ABAP add-on, although it does require a small amount of kernel functionality that exists only in the 7.21 kernel.
With subsequent versions, two major changes are happening. First, silverlight is being phased out as the presentation layer in favour of HTML5. Secondly, the screen re-writing functionality is being moved from the browser into the NetWeaver kernel. An interesting consequence of that is that it will then be able to re-write webgui and SAPgui screens! This moving of functionality into the kernel is what persuades me that this is a long-term product. This is a big change for the kernel and I don't believe it would be made for a product with a short expected lifetime. Both of these changes should be available in a version expected to be shown at Sapphire 2014, but some may make it into version released later this year.
I haven't mentioned the administration functionality for all of this. Tailored screens are knows as "flavours" "flavors" and you can control which users have access to which flavors, which flavors a user gets as a default for a given transaction, who can create their own, etc. We only got the briefest of demos of this aspect but what I saw seemed pretty comprehensive.
Overall, the event was a great introduction to what seems like a great product. I left with my head full of ideas for how I could use it back in the office. Now I just need to arrange to get access to it...