As the CTO of SAP, Vishal Sikka's job is to figure out what SAP's future applications will look, feel and act like for its customers. He does that by experimenting with untested, emerging technologies; innovating with his team and co-innovating with customers and partners; and developing new ways to think about enterprise architecture and governance mechanisms.
To him, the goal of his work is to provide clarity when customers see complexity, application harmony instead of integration challenges, and easy-to-follow roadmaps that offer direction at unfamiliar forks in the road.
That's no small task. Traditional on-premise ERP vendors such as SAP are under attack from smaller software-as-a-service (SaaS) and on-demand vendors. The value of maintenance fees is being questioned like never before. And the buzz around next-gen cloud-computing offerings has also cast gray skies over SAP's established business model.
[[ For more on Business ByDesign, see "SAP's Business ByDesign: A Riddle, Wrapped in a Mystery, Inside an Enigma." For more SAP innovation, see "Innovation Dreaming Won't Stop Enterprise Software Screaming." ]]
In fact, SAP has been poked and prodded over its much hyped and much delayed Business ByDesign offering-its suite of on-demand applications for the SMB. With the analytic mind of a computer scientist (which he is, holding a doctoral degree in C.S. from Stanford), Sikka takes umbrage at SaaS vendors propagating cloud computing platforms that are not technically accurate.
Based at SAP Labs in Palo Alto, Calif., Sikka is the vendor's first-ever CTO. His position was necessitated by the retirement of luminary Hasso Plattner from day-to-day business operations; he was SAP's de facto CTO for decades. In the Office of the CTO, Sikka has more than a thousand employees working for him.
Sikka recently spoke with CIO.com Senior Editor Thomas Wailgum about SAP's own cloud efforts, his "timeless software" initiative, and what customers really need to know about Business ByDesign.
CIO.com: Given all the change and consolidation in the enterprise software market, it would appear that your job as it specifically relates to SAP's product roadmaps is hugely important now because customers want to know where SAP's portfolio is headed. Do you agree?
Sikka: It is an extremely challenging thing because of the scale that SAP operates at right now. We cover a very broad spectrum-not only across industries and countries but also within the solution portfolio, going all the way from core transactional applications to business-user solutions: BI and analytics, new user experiences, mobility and so on. It covers a vast spectrum.
To preserve the integrity [of the portfolio] and to evolve this in a way that we can constantly absorb innovation is a very a fundamental challenge. That's why we established our technology strategy with something I call timeless software. It is, essentially, a set of tenets we live by to make sure that we are always evolving our products to bring the best innovation-that actually happens increasingly rapidly-but do that without costing our customers or costing ourselves incoherence.
CIO.com: How much of "timeless software" is SAP marketing and how much is ingrained in what you do?
Sikka: Many of our large customers have deployed all of our products, so we have to be able to preserve two things at the same time.
First is to make sure we are delivering the best innovation-whether it's mobile-user interfaces or breakthrough new analytic solutions or business-user products. On the other, there's the business process platform, where you run your mission-critical, core business processes: financials, procurement, supply chain management and so forth.
To be able to do all of that in this entire spectrum [in a way] that is coherent, that the cost of operating this for our customers is low, and that the cost of maintaining it for our maintenance organizations is also low [is our goal].
So that's the duality: We used to think in the industry that this was a divide; that you either had innovation or integrity. What we have established in the last two years is that it is possible to architect software in a way that you can consume innovation that comes in, include innovation in [products] very rapidly, and you can constantly do that without breaking your integrity.
CIO.com: What's an example of that?
Sikka: With Business Suite 7, we now have a little more than 3,000 enterprise services. These make the entire suite essentially programmable from the outside. Even though we have tens of thousands of screens on this suite and we have hundreds and hundreds of business processes on the suite, there are always more user interfaces and processes that customers or partners build. And that process has gone further in other products, like in CRM and Business ByDesign.
There is a service layer for that entire UI, so the UI is totally decoupled. That means that we can add mobile UIs very rapidly, as we did that with RIM and the BlackBerry, and now with Sybase. And we're doing it with Apple and the iPhone, though we haven't announced those yet. We are doing work in the labs to bring those new UIs in without changing the underlying application in any way.
CIO.com: What about the difference between customers who are two upgrades behind and don't need any innovation versus those who expect it continuously-how do you balance that?
Sikka: The way SAP looks at the relationship with customers in the timeless software paradigm is that the relationship is what is timeless; everything else changes. Every one of our customers has a multiyear relationship with us. We have more than 100 customers who have been with us for more than 30 years.
When you talk about innovation and leading edge versus non-leading edge, I think of it this way: There are customers for whom innovation is a necessity, and there are customers for whom innovation is a comfort.
For those for whom it's a necessity, we deliver [innovation] as it comes. For the others, they take comfort in the fact that innovation is there when they are ready for it.
But we do have customers that are so leading edge that we don't yet have products for them. We work with them on customer innovation engagements and on things that aren't even products yet. In fact, on things that might not even become products because there might not be enough scale underneath [to support] them. We can even take it into production with those customers in a one-off way.
CIO.com: There's been so much made of "The Cloud" and where SAP's products and strategy fits into cloud computing. What's your take?
Sikka: Overall our cloud strategy is in four parts.
The first building block of our cloud is to make sure our software is ready to be consumed by our customers. Today, we have hundreds of customers that use our software to deliver mission-critical cloud services to their customers, partners, employees and suppliers on far bigger scales than some of these SaaS companies.
For example, one of our large banking customers in Germany has 40 million customers and they're running our software for transactional banking. There are days when they do 8 million transactions in one hour. These are mission-critical transactions; this is not like updating your pipeline status or contact address, or something like that. We work with our partners-like HP, IBM, VMware, Citrix-to make sure our software is ready to provide cloud-based economies to our customers and their business networks.
Secondly, we augment this cloud offering with new services that we can deliver ourselves in the cloud. These are narrow applications-like talent management, salesforce automation and pricing optimization. We will deliver or are already delivering these in the form of integrated and complementary services that we offer in the cloud, and those clouds are again powered by our partners and are operated by us.
The third piece is Business ByDesign, which is an entirely different category of software: an integrated suite for running an entire midmarket company, which is something nobody does. It is not like salesforce automation or talent management, or what Workday or NetSuite does, which is largely in application areas like HR and so forth.
Business ByDesign is a piece of software to run your entire business: financials, manufacturing, procurement, purchasing, CRM, integrated analytics-the whole thing.
You have to understand that most of our customers have very substantial investments in their data centers, and these investments are not going to go away any time soon. In fact, my assumption is that for the foreseeable future, because of these investments we will see a dual, hybrid world of private and public clouds.
So at these companies that have these massive data center investments, what's happening is that companies like VMware, Microsoft, Citrix, HP and IBM are already delivering services to these data centers to turn them into private clouds.
What that means is that they can bring in excess capacity from Amazon or one of these public cloud providers as a natural and seamless extension of their private cloud. And that is our fourth piece. We are working with them and some of our customers are already doing these things to bridge the public-private cloud divide that has emerged.
CIO.com: As a technologist, does the marketing hype and related confusion around cloud computing bother you?
Sikka: I sometimes find it amusing. When [vendors] call their little salesforce-automation application a "platform," that does actually bother me, as a technologist, to be honest with you.
CIO.com: There's a perception that SAP hasn't yet delivered on the cloud, due mainly to delays with Business ByDesign. Is that bothersome?
Sikka: When I engage with people on SaaS and what the cloud is, I say that SAP has been offering a subscription model to customers since the 1990s, for some 15 years. We've had multitenancy in our software since R/2. Actually, most of our suite is already multitenancy, and we have customers that run it in a multitenant mode.
So, multitenancy? Yeah, we have that. Subscription? Yeah, we've had that a long time. Well, guess what? If you tried to do analytics on Salesforce.com, you've got to export your data and go do it somewhere else. And if you run a 50 Terabyte data warehouse, like some of our customers do, and you run that in memory, you cannot make that multitenant. That's because it is an incredibly computationally intensive operation that you have to perform on large amounts of data that sits in main memory.
That's not multitenant, and I think customers don't want it multitenant. For example, we have large manufacturing and CPG companies that prepare their production plans, and some of them have thousands and thousands of products that they have to plan the manufacturing of. That process itself takes hours to days, on machines that are several hundred thousand dollars and with Terabytes of main memory on them.
When you run a portfolio of mission-critical products like we have, there are patterns of utilization that cross all kinds of assumptions, and one size just doesn't fit all. So when you talk to people about that in kind of a hype situation, people will say, "Oh you are not running it like this or like that," it's largely a reflection of not understanding the breadth that we cover and scale at which we operate.
What's one thing you would want people reading this to know about Business ByDesign and SAP's plans for on-demand software?
Sikka: That it's a mission-critical, integrated application suite. To get it to the point where [a customer] is running that as a service, you need to ensure that the first requirement of customers-which is integrity and stability-is met. You cannot afford an 8-hour downtime in a service or some security lapse. You're dealing with financial, payroll or customer data, and if things don't work, people can go to jail.
You have to get it right-and not only from a cost and go-to-market perspective, but from an integrity perspective. For us, it is far more important to roll this out in a controlled way to make sure the customer comfort is there and grows with the software, rather than to go out there and meet some arbitrary definition of some guy's take on cloud computing or SaaS.
So we will get it right, and we will take our time. Yeah, it took a little bit longer than we thought, but we can afford to.
Do you Tweet? Follow me on Twitter @twailgum. Follow everything from CIO.com on Twitter @CIOonline.