SAP AG general manager, business analytics, Steve Lucas, visited Australia recently to help kick-start the company’s go-to-market efforts for the High Performance Analytic Appliance (HANA). He spoke with Freya Purnell about the potential of in-memory computing, how mobility is changing the game, and the future for SAP’s analytic solutions.
FP: Why is in-memory technology considered a game-changer?
SL: We are in a day and age where people have a fantastic experience at home, where you surf the web using Google. You are the master and commander of your universe – you can query the world’s largest database and do it in real time, and it only takes half a millisecond, and it’s free.
Then you go to work on Monday, and you have no idea how to get information, it takes weeks and it’s not free. That is the core issue – why is it that in the consumer world, it’s free and the experience is amazing, and in the corporate world, it costs money and it’s not amazing?
What we’ve done internally at SAP with our HANA system is take literally all of our CRM information, our entire forecast – every customer, every record, every opportunity, every interaction – and put it in one system. I can open up my iPad and answer any question I have – by product, by quarter – it’s right there instantly available. Even over a poky 3G connection, I can still get this information in real time. That empowerment is what endears people to Google, and I think it is what endears people to HANA – you are empowered, you have all the information and the tools you need and it’s in real time, and that is very transformational.
FP: How will overcoming the current gap in the user experience translate to value for the enterprise?
SL: The scenarios with HANA are really limitless, but I’ll tell you what I see happening right now. One is real-time fraud analytics, from the point of the swipe of the credit card. That information whizzes through the line and ultimately gets dropped into an SAP ERP system. The problem is that the analysis on whether that was a legitimate transaction doesn’t happen until weeks later, because there is so much information and it has to be dumped into a data warehouse. We want to literally make that instantaneous, where we can identify fraud at the point of sale. This is a huge opportunity –not just for banks, but for retailers.
The question is, how does HANA do that? We take computing that normally happens on big, slow, spinning disks and big relational databases, and we compress it down to happen inside a computer’s physical memory. It’s like the difference between writing the equation 2+2 = 4 and thinking about the answer. Calculations, data storage, complex analysis is all 100 per cent in-memory, and that makes it unique in the market.It’s having these kinds of results – for the ability to do anything that has big data and the need to apply complex analysis to it, HANA is just a killer app.
FP: What type of market do you think there is in Australia for HANA?
SL: HANA is a database – a very unique database – and we want to win in the database market. So we are bringing a new and unique database technology to the Australian market, and it is cutting edge and next generation, but the beauty is you can discard a lot of your perceptions over the last 20 years about how databases work and how they have to be architected, and move to something new.
A lot of areas in the Australian market – such as retail, financial services and agriculture – gain huge benefits from the analytics scenario in general, but I think any application that you have that uses a database today will perform better and benefit dramatically by having HANA as that database.
FP: How do you see it working in the agriculture industry?
SL: I’ll give you an example of a company we are talking to right now. The current process is that company will sell a fertiliser that has some kind of insect repellent property plus obviously nutrient properties for plants. A good chunk of the time either the fertiliser or the repellent does not perform as advertised. The farmer will say, “Come and look at my plants”, so the organisation who sold him the fertiliser will visit and write an assessment. There’s a tremendous amount of information that is collected, and they’ll record it on a piece of paper and take photos, go back to headquarters, and put that information in a database. Then you have to go through the excruciating process of comparing that to the last 50 years of crop information to decide whether they owe the farmer something. There’s a lot of number crunching required to be able to do that.
This is where the pairing of in-memory technology with mobile devices really becomes the killer app. To be able to put in the claims adjuster’s hands the ability to go out, work with the farmer, take photos, enter the information, and have that real-time analysis done using HANA is an accelerant. I don’t think people associate agriculture and big data, but actually it’s pretty amazing how much information goes into making a simple decision.
FP: How is the mobility piece changing the game for SAP?
SL: The reality is that we are headed to a day and age where today it’s called the mobile device, but soon it will just be the device. If you think about SAP, we got our start in the technology world by innovating ERP and then we built these cool line of business applications, but that was really a data-in strategy. Companies needed a place to put their information.
The challenge is that getting data in is very easy – getting it out is not so easy. So we bought BusinessObjects, which became our business analytics organisation, and that’s our data-out strategy. Then we saw this great migration of people putting down their laptops and picking up iPads. It’s like what happened when we went from desktop to laptop. So we acquired Sybase for mobility, and integrated it into the other two components, so regardless of what you are doing you can consume it on a device. This is the new SAP; it is all about ‘information in’ as a system of record, ‘information out’ as a system of engagement, and delivering it wherever you need it to be via a mobile device. And we have of course now started innovating below that stack in the database world.
We want to be able to, as our CTO [chief technology officer] Vishal Sikka would say, free the screens – get them where people need them when they need them, make them compelling, engaging and highly desirable. The vision is a reality, it’s underway. But I think companies and customers need to appreciate that a real-time enterprise needs the right in-memory data model to deliver real-time information tied into your core operational analytic systems wherever you need them to be.
FP: Obviously that approach requires a certain level of cultural change within the organisations that are using SAP. How do you see that evolving?
SL: I think the cultural change that happens with our customers actually needs to happen at SAP first. It starts from within. Since 2008, when the housing market collapsed, I have felt there has been this general undertone of agreement amongst most businesses that they need to change. I think that whole mess was a big wake-up call for a lot of industries.Our customers are asking us to help them transform.
Bill [McDermott] and Jim [Hagemann Snabe] have really been driving this change [to thinking about SAP as a business analytics and mobility company]. We spent tens of billions of dollars transforming SAP through the BusinessObjects and Sybase acquisitions into something radically different than what the company was just three short years ago. Getting our people to believe that then translates into articulating it to a customer. Our customers are demanding of us that we help them transform into something new. If we don’t do it, somebody else will.
FP: Where is the integration with Sybase up to now?
SL: We have a tie-in between the core ERP and line of business systems and Sybase, and we have a tie-in between business analytics and Sybase. The construct of being able to build this integration is there, but it needs to get better, so we are looking at how we make it seamless, so the authentication system I have internally works on mobile devices. These devices are easily lost, so there are extra security tie-ins that we are working on right now, and we are updating and releasing new versions constantly. So I am really pleased with where it is.
FP: What type of feedback are you getting from the market about the BusinessObjects 4.0 release?
SL: It’s both good and bad. The feedback is typically, “I didn’t know you could do that”. It’s great, because it makes you realise what SAP is capable of, but the challenge is that we have to create more awareness with our customers. The feedback is generally positive, I think there is a lot of excitement and a lot of very positive reactions about it.
FP: Do you see any gaps in the analytic solutions market that haven’t been filled?
SL: I think predictive analytics is a place where we partner today, and we have very good partners. Would I like to fill that gap more aggressively? Yes, because at the end of the day delivering information to people has many flavours.
The predictive analytics market is smaller, but from my standpoint, it’s an area we need to embrace. We look at predictive as a product, but we also look at it as features. The ability to look at a report and use a generic predictive model to tell me what might happen, that’s predictive as a feature, not as a product necessarily, and we absolutely are interested in doing that. The question is, what do we do with predictive as a product, which is a smaller user base than predictive as a feature. I am also fascinated by the world of how you visualise data.
We are working on filling those gaps already with projects underway. If you look at the maturation of information, there are younger generations that I think relate more to infographics and other types of visualisation forms than perhaps pie charts, bar charts and things like that. I am intrigued by it, but I don’t know if there is a lot of money there yet or not.
FP: Will the partnership with iRise start to address that visualisation piece?
SL: Possibly. I think we want to address it natively primarily. From a company standpoint, we would like to make sure that a market is mature enough and customer advantage is high enough that they want to make an investment. So I am intrigued by them and by a few other companies, but ultimately we need to natively build those features.
FP: What’s next on the analytics scene?
SL: I have been really fascinated by two very simple things in the world now. One is that I always ask myself, why is it that Google seems to magically know what I am going to ask before I type it in. But here is why that happens – you do a query on Google and get an answer. So for example, you go on vacation to New Zealand, and then what happens after that is the most important part – you tweet about it, you put a post on Facebook about it, you blog about it, and that commentary goes back into the database, so the next person who comes along and asks the same question gets a better answer, because the data itself has evolved.
We don’t do that in the field of corporate computing. You look at two charts, and say, that one is good and you move on, and the information remains in your head. So we miss the brilliance of what people bring to the data, and that is what we need to capture in the system. So what I see in our future is a ‘like’ button, for example – the ability for you to say this information helped me get to my goal. That’s where we are thinking, how do we capture the brilliance of people and the conclusions they come to. You can make that conclusion better for the next person and accelerate, and I believe if you do that, you get a critical mass of the data relating to you in
ways it has not before.
So in the analytics space, I see sentiment, more mobility, and more social aspects of business intelligence.
This article was first published in Inside SAP Winter 2011.