I recently heard from the Price Waterhouse partner who was instrumental in our moving to Tampa in 1990. It was nice to catch up and it took me down memory lane. In many ways, my group was at the epicenter of the emerging client/server application market. In 1992, we surveyed all the new players like PeopleSoft and incumbent ones like Dun and Bradstreet Software and that led to the publication pictured above. It also led to a full feature spread in InformationWeek and other press coverage. That in turn led to the recruiting call I got from Gartner which I joined in 1995.
I remember the conversations with SAP about its client/server product, R/3. There was a quiet conviction that its 3-tier architecture would scale even as the market was unconvinced anything on UNIX boxes could scale. There was a fiery determination that stored procedures and other code embedded in database were evil. From my experience with R/2 at a previous assignment in London I knew if R/3 was anywhere as functionally broad and globally competent, it would squash the competition. Its integrated nature would allow customers to de-fragment their application portfolio.
Sure enough R/3 did all that
I could also see at Price Waterhouse and much more clearly later at Gartner that SAP was clueless what happened to its code after it was shipped. It could not see or chose to ignore the havoc its partners and its upgrades were causing at its customers. 300+ pages in SAP Nation describe that pretty vividly.
So, when S/4HANA was announced last year, I thought it would continue the good that R/3 brought the world, and offset the evils it also wrought. Since SAP had seen cloud vendors like Workday and Salesforce, and its own cloud acquisitions adopt a much more efficient delivery model for over a decade, I thought it would adopt and improve their approach.
Now, I am not so sure
Unlike with R/3 there is no firm conviction about the architecture. Now SAP tells customers we have multiple models – take your pick (see the 5 editions it recently announced pictured below).
It has given up its commitment to database portability. Code embedded in HANA is efficient it says, not lock-in, not evil. Systems management tools around HANA, oh they will come eventually. The in-memory switches of other databases – don’t believe them.
And unbelievably, SAP, the one which helped companies de-fragment their application portfolio, has now become the fragmenter-in-chief. It has tens of un-integrated acquisitions, with no clear, long term road map of how S/4 will be rolled out. The disparity between the 5 editions will keep getting worse as some will have 4 releases a year, others one or fewer and customers will adopt them at their own pace.
I am sure somewhere in Walldorf, someone can keep all this straight. My only hope about all this is since S/4 is the anti-R/3 on so many fronts, that it will also kill the massive, bloated ecosystem that R/3 created.
I know – you want some of what I am smoking