Skip to Content

Intro:

To understand SAP’s Cloud Strategy, it is important to retrace our steps back to 2003 and look at the journey till HANA. Another key factor in the understanding would be to look at this journey from Gary Hammel and CK Prahlad’s “Competing for the future” context, which clearly specifies Strategy, tactics and Operations as distinctive elements in execution. SAP’s strategy to the Cloud remains unchanged, the tactics change rapidly and rightly so and the nature of Operations is largely driven by the SAP Customer and Partner Ecosystem.

SAP starting 2003 till 2011:

Allow me to retrace our steps and ramble a bit on SAP’s Cloud strategy starting from Jan, 2003, when SAP Netweaver was launched during Shai Agassi’s days. The concept of ‘ESA’ was unveiled and Shai hinged the entire concept on interoperability by addressing the world of web services coupled enhanced with business semantics to drive the redesign and development of the what one would call the ‘Mother-board’ known as R/3. (Am only starting from core R/3, post 4.6c). This meant turning around all the components of SAP R/3, the extended-enterprise components and create an open platform that would follow specific standards from UN/CEFACT and derived into a methodology known as CCTS (Core Component Technical Specifications). This meant driving a common understanding between systems and enterprises to have these services understand the semantics of the business information to be known and understood with the help of a common understanding of semantics at a conceptual and a contextual level for standardizing the information dissipation across the enterprise that would touch upon the very definition of CDTs (Core data types that would be brought about from the contextual realm into a logical realm and have the GDTs (Global Data Types) defined.

Simply put, SAP took upon the massive conversion of all its objects and underlying data-models to drive a unified meta-model strategy in place that would set right the enterprise wide service meta models for A2A and B2B. This was the inside out approach which was needed for ABAP-based systems, since ABAP systems were not scalable enough and required a huge amount of interfacing between ABAP-systems themselves. Since XML alone would never be sufficient as the silver bullet to resolve the challenges around interoperability which would enable machines and humans to understand the content delivered in the proper context, agnostic of hardware, software or operating systems. These meta models would be populated inside a UDDI 3 compliant registry on-premise for the entire SAP landscape (read ABAP systems here) , where SAP or the Customer would have the option of turning this around upside down and have a cloud based model, very different from the typical ASP model. While the ‘River’ Project solely focused on supporting lightweight extensions to its on-premises ERP thereby trying to abstract out the customizations and black-box them into insulated pieces for easy maintenance, starting off with Carbon 5.0 which was for Amazon’s Elastic Compute Cloud . (We shall leave SAP’s licensing approach out of this). Although SAP indicated that the acquired the e-Commerce vendor ‘Frictionless Commerce’ in 2006 would play a central role in its on-demand strategy, yes, that changed. Again, this was a tactic change, not a strategic change since it considered the time not right to run mission critical applications on the Cloud. I shall blog more about certain topics in detail later.

A critical reason that saw Henning ahead of his time in terms of his vision, but the market didn’t view it that way. The tactic (I refer to it as ‘Tactic’ simply because the ‘Strategy’ is to have the complete portfolio – ABAP and Non-ABAP systems on the cloud in a systematic manner. What goes on the Cloud first is purely a tactic and not a strategy. Extending this logic when SAP was being viewed as an organization which was trying to turn itself into a platform vendor like IBM (which had its Service Registry/Repository before Oracle bought out  Systinet for Design-time governance and Amber-Point for Runtime-governance of the services and SAP was yet took a beating and had to change tactics with more build. So SAP now had the Net weaver Platform and it was slowly converting all objects, Inside out with ARIS process modeling for ABAP Based Systems and then had the ‘River’ project more or less prove a show-case for all ‘Powered by SAP Net weaver (PBNW)’ solutions with the Composition Environment,  which was offered to customers to re-platform on a Java-based environment on the Dual-Stack Application Server. Oracle hit SAP below the belt by buying out SUN, which again led to a change in tactics.

All the while, SAP’s strategy on the cloud remained constant with the Business Suite Offering customers an option process-based Services and the flexibility to extend with CE, instead of bastardizing the core SAP which would need to be treated as the mother-board of applications and all services would still follow the CCTS-enabled GDTs to populate extensions with the usage of PI (XI’s name changed as Business Objects Acquisition led to the need) with the help of ‘Outide-in” services which would be needed for SAP’s Cloud based strategy which demands maximum interoperability with ‘PaaS’ and the satiation of ‘ISVs’. No change in SAP’s Cloud Strategy, but more on tactics. The hitch is of course, the Master data, which couldn’t be ignored while trying to set right the industry-wide Business Semantic adoption while Cloud-based Applications really require to be scalable enough to process large volumes of data on the Cloud. And so now comes HANA with that drives the in-memory technology with SAP’s High-Performance Analytic Appliance (SAP HANA) software for flexible, multi-purpose, data source-agnostic in-memory appliance, which is an evolution (not a revolution) when one considers SAP’s BI Accelerators rolled out for customers who were looking at the plug-in models as high-performance add-on hardware components for dropping aggregates, loading large volumes of data into the main-memory and toying with the idea of ‘On-premise-Cloud-adoption’ demos (If I may use the term). HANA is simply an extension of a stable strategy with the tactics being rolled out one by one. Master data is key here and therefore one must laud SAP for sticking on to its Cloud-strategy, adopting various and varying tactics inspite of recessions-recovery, law-suits, dual-CEOs and what not. There is much more to be read in terms of what DUET is aimed at (reasons both financial and technical). I have not yet touched upon SAP’s evolution on the BPM with CE space in this blog, but will surely do so in the next few blogs.

Outro:

SAP’s Cloud ‘Strategy’ has remained unchanged and heads in the right direction. The ‘Tactics’ and ‘Operations’ have changed and will continue to change wrt Business, economic, technological and other environment factors. This is just a view-point and I am not speaking on behalf of SAP, this is just a perspective from a geeky bystander and I would like your views/comments on the same.

To report this post you need to login first.

12 Comments

You must be Logged on to comment or reply to a post.

  1. Mark Förster
    Hello Kartik,

    thanks for the compilation. I read it, and tried to understand it, and the whole time I was wondering “SAP and Cloud Computing” does it really mix?

    I really like the concept of the BWA, and the prospect of HANA (and especially IMCE) sounds great. However, IMHO these have nothing to do with Cloud Computing. Looking at the SAP software portfolio, I really have to search which SAP software is compatible with the ideas of Cloud Computing at all.

    – simple, transparent scalability?
    – pay per use?
    – on demand?
    – SAP as a service? Honestly?

    ESA sounded like a really geat idea, however in the last years I didn’t hear much about it any more.

    From all software I know of, SAP is the least compatible with Cloud Computing, and it is really
    hard to spot (in real life!) some first tiny traces of Cloud Computing in combination with SAP. Although I am really looking for it!

    Regards,

    Mark

    (0) 
    1. Kartik Iyengar Post author
      Hello Mark~

      1. Strategic: Mentor Thorsten Franz’s Blog catches the idea of HANA and the Cloud while starting a nice debate at http://tinyurl.com/5sr8bqv . But that will have to be SAP’s direction (Strate
      2. Tactical & Operational: CIO Blog has Vishal talk of HANA here and now more at http://tinyurl.com/4z2nbb3 as In-memory computing business is here and now for immediate Business value (Quoting HP) at http://tinyurl.com/3a3nj78
      3. Touted as the ‘Oracle-Killer’ (Tactical), SAP says that “the strategy of SAP to the cloud Many an SAP ERP instance runs on Oracle Database. Those same Oracle sales representatives are likely making the case for Exadata, which the company has called the most successful new product in its history. Earlier this year, it emerged that SAP was working with Oracle to certify Exadata. AP does not view Exadata as a competitor, as HANA’s architecture and intended purposes are too different, Sikka maintained. (This is a separate topic for discussion). Now comes the real deal with Vishal when he goes on to say, “HANA will also be aligned with a series of specialized applications aimed at specific business problems. The new Strategic Workforce Planning application based on HANA was created in 70 days, according to SAP. Managers can use the software to simulate and determine what sort of staffing changes might be needed in the event of an acquisition or foray into a new line of business.”

      The points that you mention as the following – are brilliant:

      a. simple, transparent scalability? (Why not? Eg. HANA on the Cloud?)
      b. pay per use? (Why not? Social CRM like Salesforce?)
      c. on demand? (Why not? Managed Services that are typically outsourced to countries like India today? Or CRM on demand?)
      d. SAP as a service? Honestly? (Maybe not in the current, monolithic form, but leaner, meaner applications around WMS and CRM?)
      And you are right in pointing out what seemingly seems like SAP’s current form of existence wrt suite of products, that the concept of ESA/ESOA was primarily meant to slice it up. Which is there with BS7 in terms of processes. The question is not ‘How’, but the question is ‘When’. The strategy remains the same, the tactics change.

      Best regards
      Kartik~

      (0) 
      1. Mark Förster
        a. simple, transparent scalability?
           (Why not? Eg. HANA on the Cloud?)

           First of all, a BWA needs to be located
           really close to your BW system with a  
           1 Gbps network connection to be useful.
           So you cannot move a BWA into the cloud.
           Scalability would be no issue, you could
           easily add blades.

           For HANA I would need to see real-life
           examples. Does it make sense to have a
           SAP instance in the datacenter and use
           a HANA instance from a cloud? Won’t you
           lose the performance gain due to latency?
           We’ll see.

           As soon as your SAP instance is running
           on IMCE directly, separating them doesn’t
           make sense at all.

        b. pay per use?
           (Why not? Social CRM like Salesforce?)
           One of the major selling points of “Cloud Computing”
           is to only run the systems when they are needed,
           so you can shut them down for some time and save money.
           There aren’t so many SAP systems which can be
           shut down for days. Once they reach a certain
           size, then 7×24 is a must, you are always on.

        c. on demand?
           (Why not? Managed Services that are typically outsourced to countries like India today? Or CRM on demand?)
           What about SAP customizing and interfaces?
           Maybe SAP Business by Design is different,
           but SAP Web AS is really deeply intertwined
           with the platform. No company ever succeeded
           in providing SAP instances with some automated
           process and I doubt this will happen anytime soon.
           SAP software isn’t a good example for providing
           software on demand.

        d. SAP as a service? Honestly?
           (Maybe not in the current, monolithic form, but leaner, meaner applications around WMS and CRM?)
           (And you are right in pointing out what seemingly seems like SAP’s current form of existence wrt suite)
           (of products, that the concept of ESA/ESOA was primarily meant to slice it up. Which is there with BS7 in terms of processes.)
           (The question is not ‘How’, but the question is ‘When’. The strategy remains the same, the tactics change.)
           That was my main point, I cannot see it yet. And I even
           cannot see any signs at the horizon. I am really looking
           forward to any change on this, just wondering when this
           might happen.

        (0) 
        1. Kartik Iyengar Post author
          Excellent points highlighted in detail, that’s really appreciated. No, I doubt it that anyone would be so operationally highlighting that or providing answers to the Qs you pose – and rightly so. But the picture remains clear, back to the point with SAP’s game plan. There are enough brainiacs sitting in WDF watching and analyzing this to make a reality. HANA is on-premise today (as an example) for the four points of today. The ‘Strategy’ remains to getting it up on the Cloud, when technology is ready to host it up (your points). Finally, it might be just another three years when mission critical applications might be hosted on the cloud (by not looking at technological limitations of today).

          I specifically drive my blog as a journey (tactics and operations) to highlight the strides that SAP’s making toward moving an elephant to dance in the cloud. That remains consistent and clear (to address Kesker’s  blog, point no 4 as the precursor to the blog). And that’s why I highlight that SAP’s strategy with Cloud is clear, it’s about the baby-steps.

          And you just fed me more points for research (thank you for that :-))

          Best regards

          Kartik~

          (0) 
          1. Arun Bala
            → #1. THE ‘STRATEGIC’ IDEA!
            Not to deny facts by Holger Kisker. I’m quite adamant on the fact all these assumptions wouldn’t be completely “strategic”. Thorsten Franz’s Blog depicts similar points, but not entirely ‘strategic’ idea! As discussed on the blog: “HANA will also be aligned with a series of specialized applications aimed at specific business problems”. So, here she comes, Whooo.. Whoooo… HANA… And, everyone eagerly wants to test-flight this road-engine? Awesomesauces! 1st will HANA be affordable? I like the way Vitaliy Rudnytskiy has a take on this blog with some very practical examples which relates why DW must be exercised (→ http://goo.gl/Zzz8Q). Technologies: HANA is created on existing technologies: MaxDB, TREX, P*Time, Sybase. A new product & it naturally inherits many of the existing features. I’d say a big NO for a bad modeling that may have impact comparing to good design? If so, what are we trying to offer as a lowered TCO? Practically at this infancy when we can’t even predict how HANA can behave with BW & other array of new products released, it’s in fact just a “fact-based discussion” till date. Question was “Will SAP dismiss ✩ STAR Schema” → What is the future of Infocubes? → http://bit.ly/hzBOPG  There was a question on this like if the data model is “bad”, will we need to worry on in-memory capabilities? HANA must not be broken into separate pieces, as we really need to read data from RAM! Applications using In-memory capabilities today like NetApps believe to harness “locality” of data & memory capabilities (also ExaData). Will appreciate any links to learn more! Not all business problems & models are consistent, coherent & follow the same model. I’d suggest everyone to have a dig at this blog on HANA which describes ERP runs atop IMCE? Pay per Use! Aligns with #Salesforce on PaaS though! Talking a few general case studies & trends what F100s use, we can’t assume business can’t have deviation! GDTs: Blessed are those organizations that understand & ‘practice’ the real semantics of the business information! Due to enterprise irregularities, the applications have to figure a way to align with these, but how many do & don’t? (Viz. we may also compare this even with some existing complex custom developments & “crude” ABAP batch-jobs, LSMWs, BAPIs etc, will migration suffice?)

            → #2. HANA on Cloud!
            Blades are not cheap! You are speculating in the right direction. But in all the above figures, HANA as appliance we can see that is going to eventually become our in-memory underlying database. Love to take a dig at this blog → http://bit.ly/eUkzrO for distinction between HANA and its database. Yes, HANA using in-memory greatly reduces time for reads because of Disk IO bottle neck elimination. But consider cases, when we are talking about a Mio records: HUGE latency. HANA can be best described as a Ferrari car. On a contrary, if we happen to notice certain ideas, I’d consider it something “out-of-the-box”, as like the Ferrari to run on “bullock-cart” wheels! Who knows what the future holds! Excellent digg on → http://goo.gl/LplyK I sincerely ‘slide & go’ with Mark Foerster’s & every other fellow readers’ comments who has a “cloud” on his head thinking > Does it really “MIX”? Q: Why mix oil with water! Accepted both are liquids!  That sounds interesting, but is it really viable, secure, marketable, economic? I’m not an expert on these facts, but I can relate its not what we have been seeing usually. Recently Dennis Howlett had blogged on the necessity for not mixing On-Demand Vs Clouds! There was a healthy discussion on this topic & of course several of his ideas exchanged with Dick Hirsh as why there should be a separate entity for ByD developers apart from SDN. @rhirsch: “Guerrilla Tactics for SAP’s OnDemand ‘GotoMarket’ Strategy” http://dlvr.it/F9Xm0  dahowlett ByD response: http://dlvr.it/FChNm
            → Moral of the big story short: ‘Cloud & ERP’ is like ‘Chalk & Cheese!’
            There was another blog on #IW Information Week which argues how clouds contribute to only a trifle portion of SAP’s profits. Of course major revenue for #cloud vendors isn’t via SaaS purchases, but through subscription renewals & % of royalty via apps store! Clouds are best suited for medium size industries. That’s it. If HANA has the capability to scale & process 1,000,000,000,000 records & that a company has successfully purchased HANA-ON Demand (*assumption), can we think any SME will opt for it considering these aspects?

            → #3. ELEPHANTS on-board!

            I’m sure Kartick has blogged a lot here & have heard the strategy of SAP for Business ByDesign w.r.t ‘DATA’ management. Lot of the readers here may be aware of this fact that ByD has made several restrictions, (1.) # of users (2.) Seperate costs for Master Data. The more the data & # of users, the more the costs $$$$$! Rightly pointed out & ofcourse ‘Master’ data is key here, so I’d price it high… This post reminds me of the book by Louis Grestner (Former CEO, IBM) titled “WHY THE ELEPHANTS CAN’T DANCE!”. SAP ‘strategy’ to shift to #Cloud + Mission Critial Applications? It’s then a SALES pitch!! Is it a joke or sincerely what practicalities can it hold? “Gateway” was a brilliant example using Web services. Something slightly aligned with the way XI/PI WSDL also is consumed! That was indeed a good thing to be appreciated!
            Any organization implementing HANA would surely think of having these factors for any ERP/Solution!
            1. On-Premise
            2. Safety (+Uptime of course!)
            3. Updates! Writes! Changes! RW capabilities

            Not considering here in these comments, the fancy of using ‘Augmented Reality’ using BI capabilities, but strongly believe unless an application & a requirement deems ‘real-time capabilities’ Vs  ‘history’ data, we need to use some “intelligence” → Something similar a data mining engine can serve. “Built for Business Intelligence and NOT Finding Restaurants!“ http://goo.gl/32bK2 . And there’s one more great stuff “Workhorse” on IMCE → Data Loads + Frequency, ETL & Recons? Love you’d give some thoughts on this one. Do we absolutely believe HANA-ON Demand would allow for faster WRITES? I’m not arguing the gross-‘facts’ w.r.t schemas etc. As size grew, processor speeds & in-memory capabilities had to be increased. Coloumnar Storage from my view point allows faster READs than WRITEs!

            →  #4. ‘HANA’ NOT OKAY PLEASE!
            Its just some capability which is available & just released last 2010. More over, I can figure it this way to best describe it is → A Bugati X600 worth $$$ released in a car show! Capabilities are huge, but here comes the Qs! → 1. Is it sustainable? Multitude of probabilities & assumptions bring in one nice aspect: Why outsource Memory? I’d want to port 1Mio Data tuples on-board the ‘On-demand’ HANA, will it be a viable option. Take for example ByD/River where there’s no strategy to migrate ‘legacy’ data! Except using some flat file “business templates”. Now hiring: ‘DW’ Architects → 2. Is it worth the salt? They say a saying ‘Penny wise & Pound Foolish’. Are we ready to pay $$$$s to every byte of ‘Master’ (Viz . HR based Info types etc) on our #CLOUD! Come on… →  3. Will there be takers? I recall one such example on a massive strategy for #consumer #trends: At 1st a popular wireless network carrier introduced 3G & IPTV in India, due to very poor turnout & lack of consumption markets, now it suggests all its customers to go back & switch to EDGE/GPRS! Pitiful isn’t… In fact when iPhone 3G was introduced in India, there was no 3G! Imagine…. Just a simple example which came to my mind! Does SAP have the flexibility to leverage all essential usage features it has to clouds? (Q: Why’s that most of the BI-based planning applications hosted on Web J2EE migrated to ABAP NW, when none’s much interested in using them? Does this latency take a dig on application scalability or is that Java based stacks aren’t capable of performance? Why is that the tools are not ALIGNED with existing usage or legacy BPC?)
            2. Social CRM like Sales force? I guess everyone would love this blog on 1/9/90 Rule : → http://goo.gl/qMgI
            The term Social CRM has been heavily used by marketing & sometimes oversimplifies the way people can relate to. Are we ready to invest such a CAPEX budget for being “social”. There was a good comment recently on being “social”. A frequent social networking addict isn’t really “social” (he’s a loner), but only can ‘project’ his image as being one!  Jokes aside, my second take is that we have witnessed the capability to changes, how Facebook/Google has grown a popular using applications that leverage social-centric data, how text data processes do the complex math of extracting sentiments & generating grammars based on it. We can’t just be “social” of us who long ago realized that a “busted flush” Sales/Marketing storm leads the marketing way . Internally, it is meaningless, but fact accepted these give value insight to management strategies for business decision-makers iff we adapt new business models to make it. None speaks louder for you than your customer. (Who speaks candidly & in unfiltered fashion to your potential customer – From → http://goo.gl/JF1Rp). More great reads from the author on “Enterprise apps, Lotusphere and the Hollywood era” → http://goo.gl/is9fl Amazingly true!
            “The collective squeals of delight could best be summed up as: “Yay – Kevin gets it…yay – social business…yay – social anything…yay, yay, yay.”” Maybe we must list them for architects to consider! 🙂

            → #5: MINIMUM VIABLE PRODUCT!

            I’d noticed this term recently during an enterp week & this caught my eyes. #MVP! (Minimum Viable Product) — Wonderful term! MVP for Dummies (like me!) is it has just those features that allow the product to be deployed, and no more. I’m thinking to blog some salient points on MVP for Innoday… Errr, sorry Innojam 2011 on #SCN (Need some help though!). Check this article > How does your company continue to realize value from their investment in Business Intelligence? → http://focus.com/c/FIj/ Will this idea be a MVP which you can market? General method of “deploy 1st” and “code later” is akin with TDE (sort of agile methodology followed by some crude programmers like us!)

            The ideas are something like unit tests are written before and fail until the code is written, asas before releasing and assessing the impact HANA-On demand would be solely with #SAP who only can ‘monitor’ & reconfig this brilliant feature! MVP is a real eye-opener at many events including GRATNER! The question goes like this > If we are marketing something like this, will my organization be able to afford it? 2nd > How many products are you releasing! Why’s that everyone wants to have a bad time in deciding products like as with 250+ products of #ADOBE (No pun intended!) 3rd Q? > Why are you asking me to buy software that only F100 companies need? Typos regretted! Interesting PoV Kartick! Just a few general comments though.
            →  The power of ‘AND’… Mr. Sikka, I heard you!

            (0) 
            1. Kartik Iyengar Post author
              As said earlier, will say it again. Arun, your thoughts are excellent albeit opinionated. You might want to cut out the rough edges as this is a public forum. Do understand the sensitivies involved for this is a debate, it pays not to be abrasive and impersonal. My mail id is kartik.iyengar@capgemini.com and I look forward to meeting you and hearing your views.

              Best regards

              Kartik~

              (0) 
  2. Kartik Iyengar Post author
    Interesting opinions by Arn Bala Ganesan, seems to have been deleted by am puuting this back on. Arun: I truly appreciate your passion and opinions and it’s good to have a healthy debate. I must confess and put it quite bluntly that you have a plethora of informaation, but somewhere, the train of thought got derailed. I would be grateful if you can pass me your coordinates and I’ll work out something to have a debate around the topic. But hats off to you for having been candid and informal. It would be my good fortune to hear your opinions and thoughts. You can reach me at kartik.iyengar@capgemini.com

    Subject: ‘HANA’ Not OK Please!
    Author:  Arun Bala Ganesan

    Text of message:
    → #1. THE ‘STRATEGIC’ IDEA!

    I’m not denying facts by Holger Kisker, but I’m very admant on the fact the assumption wouldn’t be completely feasible. Thorsten Franz’s Blog depicts similar points, but not entirely ‘Strategic’ idea! Do you wanna check the reality! As said : “HANA will also be aligned with a series of specialized applications aimed at  specific business problems”. So, here she comes, Whooo..

    Whoooo… HANA… And, everyone eagerly wants to test-flight this road-engine? Awesomesauces! 1st will HANA be affordable? Question was “Will SAP dismiss ✩ STAR Schema” → What is the future of Infocubes? http://bit.ly/hzBOPG

    Not all business problems & models are consisent, coherent & follow the same model. I’d suggest you to have a dig at this blog on HANA which describes ERP runs atop IMCE? Pay per Use! Aligns with #Salesforce on PaaS though! Talking a few general case studies & trends what F100s use, we can’t assume business can’t have deviation! GDTs blessed are those organizations that understand & ‘practice’ the real semantics of the business information! Due to enterprise irregularities, the applications have to figure a way to align with these ISVs, but how many do & don’t? (I’m relating this even with some Custom Development & Crude ABAP Batch-Jobs, LSMWs, BAPIs etc, Will migration suffice?)

    → #2. HANA on Cloud!

    I sincerely ‘slide & go’ with Mark Foerster’s & every other fellow readers’ comments who has a “cloud” on his head thinking > Does it really “MIX”? (Q: Why are you mixing oil with water! Accepted both are liquids)

    That sounds interesting, but is it really viable, secure, marketable, economic? I’m not an expert on these facts, but I can relate its not what we have been seeing usually. Recently Dan Howlett had blogged on the necessity for not mixing On-Demand Vs Clouds! There was a healthy discussion on this topic & of-course ideas exchanged why there should be a seperate entity for ByD developers apart from SDN. I’m attaching the link, incase you’d need to know. @rhirsch: “Guerrilla Tactics for SAP’s OnDemand ‘GotoMarket’ Strategy” http://dlvr.it/F9Xm0 + dahowlett ByD response: http://dlvr.it/FChNm

    Moral of the big story short:  ‘Cloud & ERP’ is like ‘Chalk & Cheese!’  (original comments by dahowlett )

    There was another blog on #IW Information Week which argues how clouds are only a trifle portion of SAP’s profits. Of course major revenue for #cloud vendors isn’t via SaaS purchases, but through subscription renewals & % of royalty via apps store! Clouds are best suited for medium size industries. That’s it. If HANA has the capability to scale & process 1,000,000,000,000 records & that a company has sucessfully purchased HANA-ONDemand (*assumption), do you think SME will opt for it?

    #3. ELEPHANTS on-board!

    I’m sure you’ve blogged a lot & have heard the strategy of SAP for Business ByDesign w.r.t  ‘DATA’ management. Lot of the readers here may be aware of this fact that ByD has made several restrictions, (1.) # of users (2.) Seperate costs for Master Data. The more the data & # of users, the more the costs $$$$$! Rightly pointed out & ofcourse ‘Master’ data is key here, so I’d price it high… Your post remind me of the book by Louis Grestner (Former CEO, IBM) titled “WHY THE ELEPHANTS CAN’T DANCE!”. SAP ‘strategy’ to shift to #Cloud + Mission Critial Applications? Its then a SALES pitch!!!! Is it a joke or sincerely what practicality can it hold?

    Any organization implementing HANA would surely think of having these factors for any ERP/Soltution!

    1. On-Premise
    2. Safety (+Uptime ofcourse!)
    3. Updates! Writes! Changes! RW capabilities

    I don’t conside here in this notes, fancy of using ‘Augumented Reality’ using SAP + BI capabilities, but strongly believe unless your application & your requirement deems ‘real-time

    capabilities’ with ‘history’ data, you need to use BI > something similar a data mining engine can serve (“BI is’nt where you find Where’s my restaruant”) > Good post by Centigon Soltutions @

    Ryan Good Man. And there’s one more great stuff > “Workhorse” stuff on IMCE -> Data Loads + Frequency, ETL & Recon Jobs? I’d like you to give some serious thoughts on this one.

    Do you absolutely believe your HANA-ONDemand would allow for faster WRITES? I’m not arguing the gross-‘facts’ w.r.t schemas etc. As size grew, processor speeds & in-memory capabilities had to be increased. Coloumnar Storage from my view point allows more faster READs than WRITEs!

    #4. ‘HANA’ NOT OKAY PLEASE!

    Its just some capability which is available & just released last 2010. More over, I can figure it this way to best describe it is >> A Bugati X600 worth $$$ relased in a carshow!

    Capabilites are huge, but here comes the Qs!

    1. Is it sustainable? If I say I’d wanna port your 1Mio Data tuples on-borad the ‘On-demand’ HANA, I’m the world’s #2 fool. The #1 is the guy who suggested me HANA for this. I would simply put a smiley & say “God only knows!” & place a call hiring ‘Data Warehousing Architects’

    2. Is it worth the salt? They say a saying ‘Penny wise & Pound Foolish’. Are you ready to pay $$$$s to every byte of ‘Master’ data records (Viz . HR based Infotypes etc) on your #CLOUD!

    Come on…

    3. Will there be takers? (I recall one such example: At 1st a popular network carrier introduced 3G & IPTV in India, due to very poor turnout & lack of markets, now it suggests all its customers to go back & switch to EDGE/GPRS! Pityful isn’t… Infact when iPhone 3G was introduced in India, there was no 3G! Imagine….)

    → HANA can be best described as a ferrari car. But if you happen to notice your ideas, its something geeky, bit impractical, as you want this ferrari to run on bullock-cart wheels!

    Does SAP have the flexibility to leverage all essential usage features it has to clouds? (Q: Why’s that most of the BI-based planning applications hosted on Web J2EE migrated to ABAP NW, when none’s much interested in using them? Does latency affects scalability or is that Java based stacks aren’t capable of performance? Why is that the tools are not ALIGNED with existing usage or atleast legacy methods like BPC?)

    → #5: MINIMUM VIABLE PRODUCT!

    I’d noticed this term recently during an enterp week & this caught my eyes. #MVP! (Minimum Viable Product) — Wonderful term!

    MVP for Dummies (like me!) is it has just those features that allow the product to be deployed, and no more.  I’m thinking to write some salient points on MVP for Innoday… Errr, sorry Innojam 2011 on #SCN (Need some help though!). Check this article > How does your company continue to realize value from their investment in Business Intelligence? http://focus.com/c/FIj/

    Will this idea be a MVP which you can market? General method of “deploy 1st” and “code later” is akin with TDE ( sort of agile methodology followed by some crude programmers like us!) >

    The ideas are something like unit tests are written before and fail until the code is written, asas before releasing and assessing the impact HANA-Ondemand would be solely with #SAP who only can ‘monitor’ & reconfig this brilliant feature! MVP is a real eye-opener at many events including GRATNER!

    The question goes like this > If you are marketing something like this, will my organization be able to afford it? 2nd > How many products are you releasing! Why’s that everyone wants to have a bad time in deciding products like as with 250+ products of #ADOBE (No pun intended!) 3rd Q? > Why are you asking me to buy a software that only F100 companies need?

    → → → Typos regretted! Sorry for bogging some of your #ideas down Kartick! But let’s shape it ‘together’. (The power of ‘AND’… Mr.Sikka, I heard you!)

    (0) 
    1. Vijay Vijayasankar
      I think we can have a very productive debate on the topics, Arun – you raise some excellent points. Some of them probably have good counter points, and I think every one here will benefit from such a debate.

      All I would request of you is to organize your thoughts into a blog for the rest of us to understand your POV better.

      Looking forward to it.

      Cheers
      Vijay

      (0) 
      1. Arun Bala
        Thanks Vijay & Kartick. You both have been Amazing SAP Mentors to review the comments very early in the morning! I actually deleted a few notes here & there & re posted it at the bottom. I’d been a sort of bad critic though. Guess I was a bit drunk sleepy also when posting ;-). Needed some corrections & edits! So only.
        (0) 
  3. Kartik Iyengar Post author
    A note of gratitude to one of the most informed and noted bloggers – Holger Kesker. Do check his posts out, needless to say, he is spot on with his predictions.

    “Excellent blog
    Comment from Holger Kisker on Thu, 02/03/2011 – 05:41
    Hi Kartik, I did read your blog, an excellent reflection on SAP’s technology journey! I guess it comes to the point what you/I consider tactical or strategic. I simply think that the consolidation of its cloud platform on ByD and River, while a year ago Frictionless was still announced as the LoB OnDemand platform, is more than just a tactical step considering the investments from both SAP and its clients. Thanks for your insights and thoughts!
    Best regards, Holger”

    (0) 

Leave a Reply