SAP very kindly let me participate in their analyst day on April 10th, and a dinner the night before in San Francisco Westin. I particularly enjoyed spending some time with Irfan Khan, who is the CTO of Sybase, and an SVP at SAP. SAP paid for my flights and hotel too. Here are my thoughts coming out of the meeting. As always, these are just my personal opinions and not that of my employer.
GA announcements were made for BW on HANA and BPC 10 (also works on HANA). SAP said they have 80 live customers – although I have not seen a number of how many have switched off their old disk based BW systems, if any. Vishal also showed impressive performance stats on a 100TB 16 node IBM system running HANA. https://twitter.com/#!/vijayasankarv/status/189764752694722561/photo/1
It was also impressive seeing how much performance improvement has been attained on the new Sandy-bridge processors . All of that is commendable – and something SAP should be proud of.
There are a few things I would like SAP to address at the field level. Customers want benchmarks – and SAP should give them some benchmark as part of certifying the hardware. SAP should also prove out the data center readiness of HANA – things like DR/HA, power consumption etc. I told Vishal Sikka that it will be awesome if he can pull the plug literally on HANA system during his SAPPHIRE keynote and show on stage that nothing will happen to data, and that fail over works. He seemed to be in agreement. Same with concurrent users – it is a common question on how performance will vary as concurrent users work on HANA. If these questions are quickly answered, SAP has a good chance of capitalizing on the momentum.
I was also pretty psyched to hear that Vishal had designed some parts of HANA himself – hats off to him. That is true technical leadership.
Steve Lucas explained SAP’s vision on being a database company – and reminded us that it is not just HANA, but also EIM products, ASE, IQ etc. First off – SAP could not have found a better guy than Steve to lead the charge on DB. He is terrific and has a great team behind him. I was not a big fan of SAP’s original message of trying to be the number 2 DB vendor on the planet by revenue. That is not only a rather unrealistic goal given ORACLE, IBM and MS will go all out and protect their turf, it will also weaken important partnerships SAP has with IBM and MS. Vishal played down the number 2 DB vendor message, and instead put this idea forward that SAP just wants to be the fastest growing DB vendor out there. Now that is a credible message, and I am sure the partners will not feel so bad about it. Nevertheless – it will be naive to think that the three other DB vendors will stand still. It will be an interesting play off.
SAP also announced a fund to subsidize HANA implementations which is worth 250 million Euros. The idea is that if a customer buys HANA, SAP will throw in free consulting up to some % of their license fees. I seriously think this is a terrible idea on many levels. Of course I don’t deny that I work for a big SI, and hence I have a serious bias.
1. It will shut off SI partners from HANA for next couple of years. And if they have nothing to gain from HANA, why would they help SAP sell it to clients where they have relationships?
2. If HANA is truly easy and simple – why does it need SAP to implement it? And after SAP walks out of the implementation – who will support HANA ?
3. If SI partners are kept out of HANA work, what will be their confidence in investing in training and IP generation for HANA and everything else SAP comes up with?
Obviously, SAP needs to capitalize on HANA momentum now before ORACLE and IBM come out with something that makes it less of an attraction. So I get why SAP has to go all out now. But I truly believe that taking SI partners along with them would have been a superior strategy. In any case, there is a chance that SAP might in turn “outsource” some of the free consulting work they give to clients to partners. I will be watching this action closely.
SAP has also come up with a VC fund for HANA based start up companies for $155M . This is a brilliant idea – and I hope many people will make good use of it.
Sanjay Poonen spent some time explaining the mobility vision. They are not going to use SUP any more – and stick more to SAP Mobility Platform as the branding. They also announced partnerships on mobility with 3 companies. I got a chance to talk to the CEO of one of these companies called Sencha. I readily admit I like the Sencha partnership more than everything else I heard during the event. If anything, SAP missed a chance to showcase them more prominently. They have a free opensource SDK for HTML5 based mobile and desktop app development. They have an OData connector too for accessing SAP systems. They have a flexible licensing model, and have millions of developers. Now – that does not mean those millions should be claimed as part of SAP ecosystem. But it gives SAP a chance to attract those people to build SAP apps. I would encourage everyone interested in mobility to check these guys out – I am planning to do a hello world on that SDK when I get some free time.
Not a lot of clarity was given on the SAP’s mobile pricing models, sandboxes of SAP systems for mobile developers to make sure their apps work etc – but we should know more from SAPPHIRE.
Here is a parting thought for SAP – at the next teched, could you get Sanjay Poonen on stage to do a bit of keynoting for mobile ? From what I have seen and heard at other events – I am sure Sanjay will surely capture the attention of SAP techies, and he is one of the most articulate people at SAP.
All things said – it was one of the best run SAP events I have been to. Well done, SAP
15 thoughts on “SAP’s announcements on April 10,2012 – and Vijay says…”
Greate writeup Vijay, I will just make some quick points…
1. Even if SAP suceeds with HANA, what guarantee is there that Oracle, IBM or MS would not build a better, faster, cheaper clone of HANA and sell it at discounted price, eating up the market that SAP created.
2. Columnar store and In-Memory technology has existed prior to HANA, so where is intellectual property in all of this, which SAP can monetize and defend.
3. Demand for fast, embedded analytics should have been driven from UI, which SAP does not seems to be interested in anymore as is the usual SAP cheering crowd.
4. Again even if SAP succeeds will ABAP current architecture would allow leveraging SAP HANA interms of parallelism and distributed computing. SAP Netweaver / Java could have but they are not really where business happens anyways.
5. If SAP fails, will they be able to recover
All very valid points, Ranjan – thanks for commenting
Great writeup. Thanks. I guess it took a while for me to consume the information provided in SAP event on Tuesday. I probably misunderstood the message communicated on Tuesday. After reading your blog and IBM’s statement on DB2-10 Vs SAP-HANA, one thing is becoming clearer: DB vendors it seems are becoming a bit uncomfortable with SAP-HANA. This is a great news for SAP.
Frankly speaking, I’m still wondering how they developed robust R/3 system 20-30+ years back. It was inconceivable at that time. And now they executed SAP-HANA in style. It seems-based on your blog-all left to do is political maneuvering. I’m not suggesting this is going to be easy;all I’m suggesting is that they’ve delivered technical master piece in less than 2 years!!!
Now your point on DR/HA: Even today-40 years after RDBMS book release and DB2 – DR/HA doesn’t work very well when needed. Even well-planned Business Continuity tests continue to fail. These tests are conducted every 6 months with the help of one of the largest SIs and I hear after every test, oh well, SRM test didn’t go well this time; we couldn’t get /sapmnt/ mounted or one of RDBMS shared libraries was missing or we couldn’t apply redo logs to roll forward etc. Something every time.
SAP has some of the best and brightest minds in the world. So I don’t believe they would consider the idea of demonstrating DR/HA at Sapphire. If they do, hats off to them. Even if they fail(most likely they will), I would applaud their confidence level and efforts.
Once again, thank you for writing this blog. This helped see SAP’s message more clearly and easy to understand format.
Thanks for sharing your thoughts.
IBM and ORACLE and MS will all take Hana seriously , but for now it will just be a nuisance to them than an actual threat in my opinion. Next year we can see how they are shared the market.
R/3 took many years to stabilize , so I am not sure if that is a good comparison . Further, R/3 had nothing else to be benchmarked against unlike Hana which has established competitors.
SAP has very talented people no doubt, but it will be a stretch to say they are smarter than colleagues at their competitors.
In short, I would give sap a polite applause for now , and will cheer them wildly if they deliver in the promise if Hana in next year.
If after decades, DB2 DR/HA is not stable, is it rational to expect SAP to do better ? I will be curious ..more so in DR side .
Thanks for your feedback. When I talked about R/3, I was not talking in terms of how successful it was. I was talking in terms of design sophistication, standardization, maintainability, modularity and layers. Basis layer for example doesn’t easily let O/S take control of SAP. So far I’ve never worked with something like SAP-this is just one person’s opinion. My experience include: PeopleSoft, Oracle DB, DB2 DB, Informix DB, SQL Server, TIBCO and other home-grown applications. I don’t believe that design evolved over a period of several years. I believe they followed the principles of software design from the get-go.
And I find a similar approach with SAP-HANA. (Hasso Plattner’s videos on how he is designing HANA application is available on internet).
When I talk about SAP in glorifying words, it is not because I’ve an agenda; it is purely based on how I feel. I’m thinking about writing a blog on what I like about SAP. That might help others to come forward with other products that work like SAP.
And DR/HA failure was not with DB2 but Oracle. I don’t believe the failure was due to issues with RDBMS. The Business Continuity team found issues in other layers.
As far as talent is concerned, it is everywhere. However it seems SAP uses it more efficiently than other non-start-up companies.
Thanks for your time Vijay. I’m glad you had a great time San Fran and Palo Alto. San Fran and Palo Alto are my favorite places.
Great Article. Interesting to see the point about Partner Ecosystem. SAP made its way to be where they are by strongly counting on Partners. Change ?
It is an interesting change. Maybe they feel now that partners get to eat more of the pie than SAP:)
very well written summary! thanks a lot!
Very good Sir. I always read all your articles
Good article Vijay and while I can see your concerns about the impacts to the BIG SI’s on the “free” 250M EUR of consulting and the details were very vague from Steve on who would be eligible “ie case by case” I still think it is an interesting approach as it gives customers and incentive and potentially reduces the overall cost. That said SAP is pushing HANA RDS and “Simplicity Becomes Best Practice” which could send a mixed message on why all this subsidized consulting is needed anyways.
I was most disappointed in mobile around the lack of major strategy change around pricing although hearing in back channels there are some developments and more will be announced at Sapphire so cautiously optimistic.
Yes, on mobile , we did not get as much info as we did on DB stuff. SAPPHIRE would be VERY interesting to say the least
Vijay.. good summary. I am a big fan of Sencha. With their latest version 2, they have made it enterprise ready (MVC, reusable framework etc). They are also probably only HTML5 which gets the UI part well. They give an almost native API look. Try it, you will like it.
Regarding developer tools, I heard SAP is working on some way to expose the SAP mobility platform to developers. Also, some tiered pricing model. I guess, we will probably hear about these more in SAPPHIRE.
Thanks KK – Now that I know you have tried it, I will ping you if I run into any issues with my hello world