It reminded me of Larry Ellison launching the Fusion effort at Oracle (I think 2004) – and he was radical in its commitment to standard back then – e.g. admitting and calling venerable Oracle PL/SQL proprietary, not the future of enterprise software and that Oracle would commit to build Fusion on standards.
So equally kudos to Hasso Plattner to put down the thoughts and design principles for S/4HANA – even more in detail than Ellison did back with Fusion (at least as I recall).
 |
| The SAP Product Eveolution |
So let’s look at SAP’s S/4HANA launch – the video recording is here and after a 10 minute intro by CEO Bill McDermott Plattner takes stage.
The #1 Rule Plattner put out was to have a DB with 0 response time – and what is possible to innovate business applications with that design paradigm. Of course this means in memory, columnar, combination of OLTP and OLAP aka analytics today etc. Plattner gives a short run up how SAP got to S/4HANA from via research at the HPI in Potsdam and so on.
But let’s jump in when Plattner walks us through the qualities – about 18 minutes in:
 |
| S/4HANA qualities as discussed in this blog |
Built on HANA
MyPOV – Central to the S/4HANA DNA is HANA – with all the rationale on why SAP built HANA in the first place and the related benefits. But DNAs have strong and weak parts, and the weak part that concerns me is that the in memory nature of HANA will not allow SAP to address all business relevant data. SAP counters that with compression and falling memory prices it can is by itself rebutted by business relevant data growing faster than memory prices fall – so SAP needs to be careful not to be at the losing end of this trend.New UI (with Fiori)
MyPOV – SAP has long time been criticized on the quality of its UI. While supporting power users well, it was never a great UI for casual users. Kudos to Plattner to admit that SAP was afraid to change the UI. But now Fiori provides a good new alternative and the project is the largest UI innovation SAP has done since ‘enjoy SAP’ – sometimes 15 years ago. Many questions remain but it’s a good start.Guided Configuration
MyPOV – This is the more fundamental change than even HANA in my view. Business is accelerating and people in business need to move faster than ever before. And that will not change, so empowering business users to setup business applications is one of the strong points of S/4HANA. Kudos to Plattner again for admitting that SAP now has ‘what everybody else has’ – but I think Plattner is a little too tough on SAP here – as self-service setup (as I like to label it) is not (yet) pervasive in enterprise software.Subscription
MyPOV – The right move no question, challenges for SAP’s traditional licensing model and revenue streams but that is another story.1/10 of data footprint
MyPOV – SAP’s efforts and breakthroughs of reducing storage are impressive, but SAP risks at optimizing what was a cost and issue of the past – high memory costs. Memory costs are falling, SSD costs are falling even more and HDD cost is close to 0 cost. So the benefits of compression are commercially less and less relevant, but with SAP monopolizing HANA to run on the most expensive storage medium (RAM) it has to focus and optimize on this aspect more than its competitors. Providing subscription / SaaS it of course becomes even more relevant for SAP as it determines the TCO to operate S/4HANA. Again kudos for Plattner to raise that openly.3-7 higher throughput
MyPOV – Performance is always good for business. So well done by SAP, nobody does not like speed and doing the same faster. It actually becomes a mandate in the 21st century.Up to 1800x faster analytics
MyPOV – No question a columnar, in memory design beats tables on HDD aka spinning rost, but fair of course for SAP to raise the performance advantage. What SAP still needs to do and do more is how that performance advantage helps their customer run their business better. SAP had done good work on this, but still needs to do more to make the business case for HANA clear and tangible.ERP, CRM, SRM, SCM, PLM in one system
MyPOV – Plattner stated because of compression all these former ‘satellite’ systems can now come together in S/4HANA. Kudos to share that 40% of the SAP install base IT load is data exchange between SAP ERP and SAP CRM. What Plattner does not mentions is, that it was an explicit design decision to separate these system at the time, in order not to disrupt customers. But certainly one single system is the right approach. SAPs needs to explain though if it means a single database system (something SAP has achieved as (almost) all SAP products run on HANA) or a single enterprise system (S4HANA?) which is a little more out.No locking, parallelism for throughput
MyPOV – Definitively a strongpoint of HANA – the insert only paradigm has massive advantages and is a key design technique for modern next generation applications, good for S/4HANA to have it.Actual data vs historical data
MyPOV – Plattner cites co-founder Wellenreuther as the one responsible for a schema separation efforts in the early SAP days between actual and historical data. SAP did the same for S/4HANA – seeing performance gains between 2x and 5x. Definitively a good design paradigm that has come over from the old enterprise world.Unlimited workload capacity
MyPOV – By replicating between actual primary system and read only system(s), throughput goes up, a good design principle. What Plattner did not address is – what are the design / scale limitations of the primary (data entry) system. I am sure SAP has thought of that, but I don’t know the answer yet.Predict, recommend, simulate
MyPOV – A new technology platform like HANA enables new ways of doing business like simulation, prediction of outcomes and recommendations. Again kudos to Plattner for openness and historic tidbit, that former CEO Kagermann (while being a development leader) tried to do simulation in the 90ies, but technology would not allow to do pull that off. In 2015 SAP needs to show more how predict, recommend and simulate enable next generation best practices.HCP Extensions
MyPOV – Not sure why Plattner skipped over this – but a PaaS (HCP) is a key quality for a modern business application. Not only for hard core developers, but also for customers – ideally business users – who need to be able to change certain qualities of the system to adapt and differentiate the business process they need to run. No trivial feat to find the right mix between enablement and closure, but SAP with its vast experience of enterprise applications should be in a unique position to get this balance right. More to learn about this would be great.HANA multi tenancy
MyPOV –Kudos for Plattner to share that multi-tenancy is not that crucial for SAP with ERP than it is for the smaller footprint network applications. With that Plattner shared that SAP maybe operating with 2 ‘to be’ tenancy paradigms which is good to hear from a pragmatic standpoint, but more complex to operate and engineer for.All data, social, text, geo, graph, processing
MyPOV – A key aspect to raise that HANA can store and process all these data types for better insights, but if we e.g. pick social, we don’t see enterprises using HANA as their exclusive social data storage and processing platform. Back to the argument of cost of the storage medium.Overall MyPOV
SAP deserves credit for building (finally) its next generation application and sharing qualities openly. My concerns remain on cost to operate, limitations on use cases given super large amounts of data e.g. in social, IoT, BigData in general cannot be held in memory, in HANA as the platform for S/4HANA.SAP has done well on the roadmap side for Financials, but enterprises need much more than Financials (HCM, CRM, SCM etc.) and SAP owes customers, prospects and the ecosystem a roadmap how and when it will get to bring all these together with the S/4HANA qualities. Sapphire May 4-9th 2015 in Orlando will be the next stop to find out.
And more on overall SAP strategy and products:
- First Take - SAP Cloud for Planning - the next spreadsheet killer is off to a good start - read here
- Progress Report - SAP HCM makes progress and consolidates - a lot of moving parts - read here
- First Take - SAP launches S/4HANA - The good, the challenge and the concern - read here
- First Take - SAP's IoT strategy becomes clearer - read here
- SAP appoints a CTO - some musings - read here
- Event Report - SAP's SAPtd - (Finally) more talk on PaaS, good progress and aligning with IBM and Oracle - read here
- News Analysis - SAP and IBM partner for cloud success - good news - read here
- Market Move - SAP strikes again - this time it is Concur and the spend into spend management - read here
- Event Report - SAP SuccessFactors picks up speed - but there remains work to be done - read here
- First Take - SAP SuccessFactors SuccessConnect - Top 3 Takeaways Day 1 Keynote - read here.
- Event Report - Sapphire - SAP finds its (unique) path to cloud - read here
- What I would like SAP to address this Sapphire - read here
- News Analysis - SAP becomes more about applications - again - read here
- Market Move - SAP acquires Fieldglass - off to the contingent workforce - early move or reaction? Read here.
- SAP's startup program keep rolling – read here.
- Why SAP acquired KXEN? Getting serious about Analytics – read here.
- SAP steamlines organization further – the Danes are leaving – read here.
- Reading between the lines… SAP Q2 Earnings – cloudy with potential structural changes – read here.
- SAP wants to be a technology company, really – read here
- Why SAP acquired hybris software – read here.
- SAP gets serious about the cloud – organizationally – read here.
- Taking stock – what SAP answered and it didn’t answer this Sapphire [2013] – read here.
- Act III & Final Day – A tale of two conference – Sapphire & SuiteWorld13 – read here.
- The middle day – 2 keynotes and press releases – Sapphire & SuiteWorld – read here.
- A tale of 2 keynotes and press releases – Sapphire & SuiteWorld – read here.
- What I would like SAP to address this Sapphire – read here.
- Why 3rd party maintenance is key to SAP’s and Oracle’s success – read here.
- Why SAP acquired Camillion – read here.
- Why SAP acquired SmartOps – read here.
- Next in your mall – SAP and Oracle? Read here.
And more about SAP technology:
- HANA Cloud Platform - Revisited - Improvements ahead and turning into a real PaaS - read here
- News Analysis - SAP commits to CloudFoundry and OpenSource - key steps - but what is the direction? - Read here.
- News Analysis - SAP moves Ariba Spend Visibility to HANA - Interesting first step in a long journey - read here
- Launch Report - When BW 7.4 meets HANA it is like 2 + 2 = 5 - but is 5 enough - read here
- Event Report - BI 2014 and HANA 2014 takeaways - it is all about HANA and Lumira - but is that enough? Read here.
- News Analysis – SAP slices and dices into more Cloud, and of course more HANA – read here.
- SAP gets serious about open source and courts developers – about time – read here.
- My top 3 takeaways from the SAP TechEd keynote – read here.
- SAP discovers elasticity for HANA – kind of – read here.
- Can HANA Cloud be elastic? Tough – read here.
- SAP’s Cloud plans get more cloudy – read here.
- HANA Enterprise Cloud helps SAP discover the cloud (benefits) – read here.