Adam Binnie, global VP and general manager of Business Intelligence Solutions at SAP, has worn many hats during his 20-year IT career. Binnie’s contributions include SAP StreamWork, BusinessObjects Explorer and Crystal Reports for Java. His background includes Business Objects and Crystal Decisions and he sits on the NSERC scientific council that advises and researches analytics at universities across Canada.
Binnie’s mission right now is to deliver an overall portfolio of analytic solutions to SAP’s customers. In this interview that took place during a July analytic gathering at SAP’s U.S. headquarters, he talks about SAP’s market positioning and moves into a broader discussion about business and data analytic trends in the marketplace.
Before we get the business trends driving analytics, I’d like to talk about SAP’s message to the market. With SAP’s ERP portfolio and all the acquisitions, Business Objects, Sybase, SuccessFactors, I am curious if you see analytics or business intelligence or something else as the current entry point for discussions with existing customers and prospects.
I am part of one of the businesses so to an extent that question belongs to someone else, but SAP as a whole has five businesses. If you went to our enormous customer base I think they’d say the on-premise business applications remains the center of what we do. In our on-premise business, something like 60 percent of the world’s GDP goes through our software. We have 17 million people operating in the cloud every day. We have a mobile business we are building very quickly that is going to be transformative in how all those other assets get delivered. We have a database and technology business and we have the analytics business which I operate. Each of those five represents a whole business that would be successful on its own but is more successful in context as you realize these things are all connected.
That much breadth is hard to center up in a message. I asked because a lot of messaging we’re getting from SAP now seems to be centered on the HANA in-memory database and analytics. Is that just a constructive place to engage businesses?
In the analytics business we think there is a plain connection between the database and technology and our mobile business with HANA and the transformation of the data environment to new architectures. Mobile technologies are the method by which content and insight and interaction moves outwards to the edges of the organization but you could also put the business suite in the middle of that picture, with the data marshalling in this real-time environment and applications being delivered out to business users that are connected to that SAP core application context. But I think our core focus is still very clear. We are about making organizations run better. It’s on the signs.
How are database customers responding to the analytic message?
Data warehouses and the core database part have been around a long time and we see that industry in the midst of a huge transition. There is in-memory technology coming through, new architectures [where] we think of analytics as taking all the information of value in core assets and the differentiator is how the business uses that information to make a difference. That also underlies everything we do. You have to run the actual mechanics of the business better, you have to operate IT more cost effectively and deliver many insights that come together in one mission.
We’re hearing of many companies that want to decentralize business intelligence and of growing customer demand for self-service BI environments. How does that reconcile to the traditional use of enterprise software?
I agree with the observation. I was chatting with Howard Dresner before our event this morning and we were talking about this interesting duality around agility, time to value and self-service versus governance and the laminar flow. The truth is there is the duality of needing to do or understand something right now and the knowledge that if I just do the “now” stuff all the time I will end up with chaos. So there is a wonderful ongoing tension between short, quick, reactive and long, slow and efficient and analytics is always going to be like that with new ways and new insights always waiting along that adoption curve. Clay Christenson talked about crossing the chasm and we do that every week. We are always doing something for an early adopter in an organization and something else for a late adopter. Both need to happen.
Yet we still hear that large enterprises are too slow to support immediate needs and that sends business users back to their own devices.
IT tends to focus on the long and slow like a big plan or data warehouse no one needs, and business has the short focus where they create the spreadmarts that can’t be sustained and we end up arguing whose spreadsheet wins, not whose data is right. That’s still true but the questions my customers ask me are usually not about technology or how to build a BI system. They want to know how to apply the BI analytics to some problem or challenge, and that conversation is also focusing on what the business is trying to solve and the multitude of strategies they have. On the back end, they might need a flow of operational metrics through sensors in trucks that service coal mines in Australia, right up to an executive and a dashboard on his cell phone.
Are cell phones and tablets becoming core parts of the new process execution platform?
You have to think business users will more likely operate and make more value of information on their tablets than on their laptops. I think in most organizations the first consumers have been the executives. That’s why the iPad has been so dominant. I met an education board in Australia whose primary driver right now is how to get analytics to every school principal. An interesting thing about a school principal is that he probably doesn’t have a laptop. He might have a computer, but he will almost certainly own an iPad. The reason educators are buying iPads is because as a teaching tool they are unrivaled. Our North America sales head likes to talk about how his three-year-old daughter tried to swipe the television set and declared it was broken.
How do you see this phenomenon happening globally?
First of all, it is happening globally. Here in North America I might offer you a cell phone application to help you adjust to mobile banking or whatever. We have a customer in South Africa that delivers mobile services as a primary business mechanism because most of their customers in the townships have a cell phone contract but most don’t have an address. They literally sign up with their phone and that is the vehicle to pay, receive money, scan payments, and do their business.
In the corporate world, users also seem more attached to their device than their address. Is that a similar kind of account management?
In the corporate world we can tap into that same kind of sweetness. But mobility has always been waiting for the analytic technology and availability to get there. Now we’re on another tipping point, so getting 20 executives an iPad is fine. But getting every employee an iPad is still something not under consideration in many corners of the universe where companies don’t have the security, maybe they don’t like Apple or something else. But more often the payoff for mobile is coming apparent for guys who drive trucks, in a mining plant in the middle of nowhere. It’s not just the road warriors and executives flying around but people who are never in the office, never near a computer or never had one. Suddenly their own cell phone they are going to carry anyway can be a vehicle for them to interact with timesheets, equipment tracking, health and safety, a photo or something else. Many parts of the software industry are experiencing generational transformations to social networks, people like my niece who just don’t use email.
Generational or otherwise, doesn’t this feel like an important question of timing for when companies ought to make a broad leap like that to full mobility?
The challenge for organizations is when to jump on the path, not a question of “if.” They will no longer be competitive if they do not change. Realistically, I am most conscious of the mobility [issue] even though we know it’s not even really going yet. When it does it’s just going to accelerate, especially on a global stage, places where mobile infrastructure is in place but digital infrastructure is not. In that setting mobile becomes the only technology. The other half is the data environment. Particularly when you look at organizations trying to connect data, not just their data but everyone’s data, so supply chain networks, value chain networks, they are starting to share data a lot and bandwidth is the enabler for that sharing.
People are already supporting working processes with powerful external apps and information streams. How can businesses create to that productivity?
In-memory analytics is proving to be one thing companies are finding very powerful and interesting. We have a new demo I will show you [Binnie opens his iPad and shows a demo], a contiguous map called Your World. We took 70 million rows of basically demographic info about the U.S., loaded it into a HANA box. It’s not a huge amount of data, but what’s interesting is that it is nonaggregated. Because of that I can swipe all around this map of the U.S. and get statistics automatically based on the exact display I see on the map on my iPad. So, I can see 1.9 million live here, 3 million 3 million live there, I have a whole wall of metrics we can access. I can see the percentage that is male population or registered voters or income as a calculated aggregate metric taking the population and dividing it by whatever. Had I done this in the classic business intelligence way I would group it by county and then by state and I could see the total for any given state.
That’s immediately more useful than a table of statistics.
Sure, [in a classic approach] you’ve predetermined what is useful but you might not be interested in just one state or a group of states. I might be interested in how many people are in a region not divided by boundaries. If I had transactional log data on a customer who went to my store I could provide this exact app for a retailer and they could look at distributions for people and determine an optimum distribution center location. You can do that if you don’t pre-aggregate. It seems like a small thing but it changes the entire game.
Is that the kind of in-memory proposition you’re offering to customers?
We call it HANA and not in-memory DB because it’s actually six or seven developments together. In memory is one of them, parallelization is another, the way we use column storage, the way we load it, several things aggregated together form this transformation. When you get a certain distance down a path, 1,000 times performance or 100,000 times performance, you get to these transformational points. Instead of having to do a batch run of suggested offers to you now I can make you offers when you walk in the store. If you have given me permission to know your cell phone number, when you touch my WiFi I know you are in my store or walking by. I can connect to things you like or that matter to you, offer you 20 percent off while you are here.
Do you believe most marketing managers are thinking that way now?
I talk to marketing managers, one especially who talks about people who don’t watch TV anymore, so he’s concerned how to build brand awareness. He used to have a rule of thumb that if he spent a certain dollar figure on TV ads he’d get a certain amount of lift, but now those metrics don’t work anymore. His favorite topic now is how to create a new metric.
That’s going to be a whole new frontier of work in a field where some say we are long on data and short on useful conclusions.
And data is still messy and dirty. It’s like horticulture, like a garden that is constantly going out of whack. All you’re really trying to do there is give it some guidelines. You can’t fine tune chaos, it never stays perfect. The sense of organic reality rather than inorganic reality is why data is looking more like a field outside than what we have tried to make it.