The last and probably toughest part of our efforts at SAP was to simplify the consumption of SAP’s software by the customers. Simplification can be viewed as the attempt I mentioned in past entries to eliminate “IT tax” imposed on the whole industry by, well… by the industry. Every industry goes through stages where every rule is up for re-evaluation – these are the transition stages, such as the time where the industry went from big-iron to distributed Unix boxes. After those periods of great change, we tend to fall into a period where inertia drives continuous innovation with more and more extensions along an established architecture and business models. This is the classic innovation model detailed so well by Clay Christianson in his innovation series.
The last wave locked into a model where the application comes to the customer as “almost-done”, with various vendors allowing the customer (usually through a system integrator) to continue and modify the solution to fit their needs. No matter how many times we tried in the past to convince the customer that the offering is “complete” or “out-of-the-box” there is just too much inertia in using all means to fit the solution to the customer processes and not the other way around. I find an industry that spends a high multiplier of cost for implementation and operation to the original cost of the license as an industry that is ready for a disruption – in the classic Christensen model.
The effort this industry needs is a strong analysis of where the money goes to (in the true Total Cost of Ownership) and a concentrated attempt to design the costs out by changing the product, go to market, and eco system so that they just cannot accept the cost any more. Doing so successfully would not only make the customers happier – it would release many funds currently captured in the operational cost structure back into capital investment in the form of innovation.
Take as an example the BI accelerator (BIA) that SAP shipped almost a year ago now. BIA was one of the smartest software innovations that were created at SAP in the last decade – an algorithm that allows a system to store data warehouse in columns instead of rows – leveraging the low columnar variance to compress the data and distribute it across many CPUs. The result was a multi-billion record data warehouse machine that can retrieve any query in sub seconds. Amazing innovation. The team came to me with the software, and won an innovation competition inside SAP (in a wonderful battle of the geeks during our Developer kick-off). To install the software, you needed to be an expert in so many different domains – from hardware optimization to Software setup and analytics. I asked the team to finish the product – “get me a box that can be setup in one day with one on/off button”. Without aiming for that kind of solution we could not have scaled the deployment. The team delivered and BIA actually got the 100x performance boost promised and then some.
The interesting result, we did not understand up-front, was the elimination of the complex human element required to operate a data warehouse. You see, the highest cost in data warehousing is not the disk or cpu, rather you need to figure out how to eliminate the need to constantly “lift the database” by optimizing aggregates and improving query response times. By storing the data in memory you eliminate the ability to speed data access (Well, you also eliminate the need by getting to sub-second response). The result you should see when BIA takes hold will be a shift of that cost structure away from base technical maintenance into smart analytic applications (that and less database usage - which is an added bonus).
I have shared in previous keynotes projects such as Muse for simplification of UI, but there should be more such efforts eliminating costs out of installation, configuration, testing, general operation, etc. SAP announced some products aimed at similar simplification during Sapphire – such as the Duet appliance, or even Duet itself. I believe the next massive industry disruption is coming, in the form of massive TCO reduction. Most of that reduction will be in the packaging and system integration around the product – and some of it will be engineered into the product itself. SAP is in the best position to drive that change – you see, SAP doesn’t have a problem of trying to figure out how to connect lots of pieces together to make a solution happen. Instead of that SAP can work from a single blueprint and affect the change horizontally across the entire suite, and for all industries.
Comments