SAP innovating with cloud, mobile and in-memory computing
Based on my attendance at SAP's SAPPHIRE NOW conference this week, SAP appears to be making major advances in three strategic themes of innovation. But to succeed, it needs to do two things equally well: reach new customers with leading-edge technology while at the same allowing its legacy customers to adopt these technologies in an incremental way.
I arrived in Orlando Sunday evening just in time for dinner with SAP’s co-CEO Jim Snabe and a small group of bloggers. It was a chance to get to know Jim up close and exchange views on SAP and the enterprise software market. The conversation was very frank, and most of it was off-the-record. Jim described his own roots at SAP as well as his vision for the future, which currently revolves around three broad themes of innovation: cloud, mobile, and in-memory computing. Not surprisingly, these turned out to be major themes throughout the keynotes and briefing sessions I’ve attended.
As Jim pre-announced with us Sunday evening, SAP just sold customer No. 500 for Business ByDesign (ByD), its pure SaaS solution for small and midsize businesses. This puts SAP ahead of its goal to reach 1,000 customers by the end of 2011, a point of pride. On Monday, I met in a small group briefing with Rainer Zinow, who is in a leadership position with both ByD and SAP's new line-of-business applications for customers of SAP’s Business Suite. The ByD platform, though originally developed for the small business offering, now serves as the platform for all of SAP’s on-demand applications, such as its Sales On-Demand.
On Monday, I also took time to walk out to the show floor for a test drive of the ByD application. I was impressed by the extent of functionality offered by the product, though I did find screens to be crowded and a bit difficult to navigate at first. Some of this might be resolved at implementation, as screens can be customized to reduce the amount of information displayed.
On Tuesday, I interviewed an early ByD adopter, who confirmed that they did find it helpful to configure screens to remove unnecessary information. This configuration can be easily accomplished by end-users. Nevertheless, it points out that once you get into implementation and change management, there’s not much difference between an on-demand and on-premise solution.
SAP took a big step into mobile computing with its 2010 acquisition of Sybase. Much of the focus since the acquisition has been on development of the Sybase Unwired Platform (SUP), which provides the infrastructure for mobile applications (for example, device management). Now, with the platform now in place, the focus is turning to mobile apps themselves. SAP is showing dozens of mobile apps on the show floor. I got a look at some of them on the floor and also in a blogger briefing with SAP’s Ian Kimbell. Kimbell reports that initially it took SAP’s team a week or two to develop the initial apps, but after coming down the learning curve, the team now only required a couple of days to developa new application. This sort of productivity will hopefully lead to an explosion of innovative mobility applications, not just to replace desktop apps but to enable new business processes.
Some of the new apps are quite simple, such as expense reporting or management approvals, while others carry deeper functionality, such as those for patient data reporting or field service management. Some are developed directly by SAP, while others are being developed by partners. Most are or will be available through the newly announced SAP apps store.
See the video I shot, below, of some demonstrations of these mobile apps.
Most enterprise software vendors consider cloud computing and mobility applications as the top two innovations currently disrupting the enterprise software marketplace. But SAP consistently lists a third innovation: in-memory computing. The strategic place of in-memory computing was made clear in my meeting with SAP co-founder and current Supervisory Board chairman Hasso Plattner.
In memory computing is simply an computing architecture where an entire database resides in main memory and all reads/writes are performed directly in memory, instead of in disk storage. (In SAP’s implementation, disk storage is still used as staging area for source data, for failover integrity, and to archive inactive data.) In-memory computing is enabled by the rapidly the improving price-performance of computer processors and memory, allowing main memory to reach into many terabytes, limited only by the number of nodes in the cluster.
SAP’s flagship use of in-memory computing is in its HANA in-memory appliance, although the technology is also being deployed in other applications, including SAP’s Business ByDesign, described earlier. SAP referenced a number of early HANA adopters during the keynotes, most of which focus on business analytics with very large data sets. Early adopters include:
Colgate Palmolive, which uses HANA to generate detailed real-time sales reporting for customers. The firm is also working with SAP now on a HANA application for trade promotions management.
Medidata, a SaaS-provider of clinical trials data management. The firm uses HANA to allow its customers (generally large drug and medical device developers) to analyze clinical trial data from around the world for trends in quality issues and to quickly take remedial action, potentially saving large amounts of money in each trial.
Caterpillar, the manufacturer of large earth-moving equipment, which uses HANA to analyze large numbers of product configuration options and design-to-order features. This application of HANA allows the firm to determine what configurations are feasible from instantly analyzing millions of rows of data.
Canoe Ventures, a joint-venture of several cable TV companies that uses HANA to customize delivery of ads to millions of individual viewers, allowing them to instantly request more information. This is an application that uses HANA for non-SAP data. Once the ad and viewership data is no longer needed, it returns to its sources. Nothing is held in a permanent data warehouse.
These are just four of the many case-studies presented by SAP board member Vishal Sikka and discussed further in my briefing with him after his keynote session.
As a data analysis tool, HANA competes with Oracle’s Exadata product line. In-memory computing itself is not unique to SAP. It is also employed by other enterprise system providers, such as Workday, as well as data analytics firms such as QlikTech.
Interestingly, SAP has at least two connections between in-memory computing and cloud computing. First, its cloud ERP offering, Business ByDesign, itself uses an in-memory computing architecture. Second, SAP plans to offer HANA as a service, running in SAP's own cloud, for customers that cannot afford or cannot justify an on-premise purchase of a HANA appliance.
What does it all mean?
SAP is one of the largest and oldest enterprise software providers. It has a large installed base. In pushing forward on these three fronts of innovation—cloud, mobility, and in-memory computing—it has two goals. First, to keep its installed base customers supplied with new technology, to keep them from looking elsewhere. Second, to provide new and interesting solutions to prospects that are not yet SAP customers.
These two goals can be seen in each of the three themes.
With its cloud offerings, SAP is aiming its line of business solutions, such as Sales OnDemand at its installed base, some of whom have been leaving the fold, choosing Salesforce.com for sales force automation rather than SAP’s own on-premise CRM solution (as evidenced in a recent deal I was involved in). At the same time, SAP is targeting its Business ByDesign service to new prospects--small and midsize companies--that are increasingly looking at cloud solutions, such as NetSuite. In addition, SAP thinks it has a winning strategy in selling ByD to small subsidiaries of its large customer base, which might otherwise look to Microsoft, Epicor, Infor, or QAD, for example, in a two-tier configuration.
With its Sybase Unwired Platform, SAP is trying to make life easier for its installed base customers, who otherwise would have to develop their own mobility applications or integrate offerings from a variety of providers to run on a variety of devices. At the same time, SAP thinks it can sell its mobility applications to companies outside of and separate from its installed base, potentially serving as an entry point for other SAP products. My associate Dennis Howlett thinks SAP should be even more aggressive in this strategy, by providing the Sybase Unwired Platform at low or even no cost, to gain a foothold in new organizations.
With its HANA in-memory data appliance, SAP is looking at providing supercharged data analytics capabilities to its own installed customer base, such as its Business Objects users. At the same time it has taken great pains to ensure that HANA plays equally well with non-SAP data, to position HANA as a general purposes data analytics solution, as shown in the Canoe Ventures case-study, outlined above.
Will SAP be successful in these three themes of innovation? It certainly has the resources to do so. Its large installed base throws off billions of dollars in maintenance fees that SAP can invest in new development. It also has a long and proud history of engineering excellence. And in its acquisitions of Sybase and Business Objects it gained the subject matter expertise and customer base to give it a foothold in mobility and analytics.
If there is any area where I might have reservations, however, it might be in cloud computing. Cloud solutions are an entirely different animal than on-premise systems, in terms of how they are developed, sold, delivered, and maintained. The revenue model is different, and as a public company, SAP is very sensitive to short-term constraints on its profitability.
When it comes to investing in the resources SAP needs to make ByD or the line of business applications successful, will SAP willingly pull resources away from its high-margin on-premise business or its multi-million dollar HANA deals? Those who are fans of Clayton Christensen understand the "Innovator's Dilemma." It is not an easy step for market leaders such as SAP. The leaders in the Business ByDesign and the line of business applications areas are certainly smart, talented, determined, and visionary. But will SAP as a whole stand behind their efforts when there is easier money to be made elsewhere?
Whether SAP as an organization can maintain the level of commitment needed to make them successful remains to be seen. The early results, with 500 new ByD customers is encouraging, but there are many more months and years ahead.
Note: SAP covered my travel expenses for this event.
by Frank Scavo, 5/18/2011 07:45:00 PM | permalink | e-mail this!
It will be interesting to observe the disruption HANA (and Sybase Replication Server) has on the datawarehouse establishment. Reduced reliance upon (inflexible) data cubes, without the adverse performance impact on source systems, is a giant leap forward for business users.
But for me, the most interesting change might come when HANA supports both read and write, unlike the current version which supports only read.
Will this take us back to the good ol' days of a unified enterprise database solution? Or at least drastically reduce the number of applications and databases in the Business Suite. By how much can the pain associated with complex interfacing reduce?
I think SAP has a real opportunity here because many customers are holding fast to their ERP system (as distinct from Business Suite) in a bid to reduce their exposure to interfacing.
This quote by Vishal Sikka on HANA:
"But perhaps even more important, we see the opportunity to completely rethink all of our applications on top of this new technology, and that means a lot of the applications that currently do the processing of that data become so much faster...which gives us the opportunity to completely refactor the apps themselves." Quote taken from http://www.informationweek.com/news/global-cio/interviews/229200069
Vishal is probably referring to the refactoring of applications within themselves, but I see this going further where two or more or even an entire suite of applications are refactored together.
But there will still be a need to connect different systems from different vendors, and older systems with newer systems.
Matt, thanks for the comment. I am in process of reading Hasso Plattner's book, "In Memory Data Management," and I can tell you that the reference model is puts forth there for in-memory data base is a full read/write implementation. If HANA only supports "read," it must be by SAP's choice at this point.
In addition, as I mentioned in the post, ByD is also built on an in-memory database, and obviously that has to be a read/write model.
Regarding simplification of the full suite, I think you are on the right track. Rrom my briefings it is clear that SAP's vision is to incrementally migrate functionality of the full suite to this technology. That will take many years, but if successful, it would be a real transformation.
Send tips, rumors, gossip, and feedback to Frank Scavo, at
I'm interested in hearing about best practices, lessons learned, horror stories, and case studies of success or failure.
Selecting a new enterprise system can be a difficult decision.
My consulting firm, Strativa, offers assistance that is independent and unbiased.
For information on how we can help your organization make and carry out these decisions, write to me.
My IT research firm, Computer Economics provides metrics for IT management, such as IT spending and staffing benchmarks, ROI/TCO studies, outsourcing statistics, and more.