TE: I’m a statistical economist by training, but when I left school, I wanted to see the world. I ended up working on an analytics project at Shell in New Zealand, with a makeshift system including a mainframe reporting tool, exports to Lotus 1-2-3, custom macros, and a pen plotter.
A few countries later, I joined Business Objects in 1991 as the eighth employee—prompted largely by the realization that the project that had taken me a month at Shell could be done in less than a day using SkipperSQL (as the BO product was then called).
I’ve had a variety of roles over the last 20 years, but one thing has been consistent: I’ve always been lucky enough to have a market-facing role, spending time understanding customers’ real-life information challenges. My role now is “technology evangelist”—I spend a lot of time at conferences and using social media, doing my part to explain new BI technology and how to achieve its full benefits.
TE: Clearly, we had the right product at the right time: organizations had valuable information locked away in their databases, and you had to have technical knowledge to get it out (although Oracle at the time positioned SQL as an “English-like language” for power users). We pioneered the notion of a “semantic layer” that let business people access information using standard business terms and which automatically generated the complex SQL required to get the data from the database.
BO also benefited from great management from its founders, Bernard Liautaud and Denis Payre. They were determined to create a “Silicon Valley start-up” on the outskirts of Paris that included rapid expansion and a global vision. There were a few stumbles along the way, but in the end, it became the second European software start-up (after SAP!) to reach a billion dollars (USD) in revenue.
TE: Clearly, the technology has changed a lot over the years, but the business requirements remain remarkably similar: cutting costs, finding new opportunities, beating the competition, getting closer to customers... The biggest change is perhaps that BI is now clearly mainstream—only a tiny fraction of organizations don’t have some sort of BI in place, even if it’s only using spreadsheets.
TE: The dream of most organizations is to have a BI solution that just works—people believe strongly in the benefits of BI, but wish that it were easier to put in place robust solutions.
A new wave of technology, including in-memory techniques, holds the promise of making some of the processes much simpler. But many of the most intractable problems stem from business processes themselves (organizations with 20 different definitions of “customer” etc.), and are consequently much harder to fix. There’s also the problem of expectations: as soon as you provide better BI systems, business people can (and should) move on to new and even more difficult questions. I don’t believe people will ever be completely happy with their information systems (and if they were, it might be a sign that they needed more imagination!).
TE: SAP HANA combines a series of technologies (in-memory, column stores, in-database calculations, etc.) to create a truly innovative alternative to traditional BI infrastructures. Early customers are now starting to reap the benefits in the form of radically faster access to large quantities of data. This is allowing them to make better decisions, earlier, and more often, and fix potential problems in real time, rather than analyzing what failed in the past.
TE: I think the biggest benefit of HANA isn’t actually its speed—after all, every new generation of databases has been faster than previous versions. What’s different about HANA is the way it “collapses the layers” between data and analysis, and radically simplifies the implementation of new BI projects. When business people come up with a new analytic need, IT should no longer have to say, “Come back in six months when we’ve managed to get the data into the data warehouse.”
An analogy for how BI has changed would be the move from film to digital photography. In the old days, you’d have to buy film, load it into your camera, take pictures, send the film off to experts to get it processed, and after three days you’d get your pictures back—only to realize that they weren’t quite what you wanted. A lot of enterprise BI today works along exactly the same lines—as a business person, I have to rely on experts to get the solutions in place, it’s slow, and what I get back is often not quite what I need.
HANA is like the digital camera—it’s faster, but the real benefit is getting rid of the redundant layers in the process. Today, I can take a picture without an expert’s help, and if it’s no good, I can quickly take another while I still have the subject to hand: we’ve all become better photographers because of digital cameras, and better BI will be the result of in-memory technologies like HANA.
TE: It’s clear that cloud computing is the future, and it’s also clear that on-premise installations are going to be with us for a long, long time. SAP’s strategy of “orchestration”—helping organizations make the best use of the combination of technologies (increasingly including mobile)—sounds like the right approach to me. From an analytics point of view, BO was, and continues to be, a pioneer in on-demand business intelligence with our OnDemand platform, which now offers HANA-based BI in the cloud.
TE: The “nomenclature wars,” as I call them, drive me nuts. People often conflate two things: the technology that is being talked about (this changes over time, requiring new terminology) and the underlying business needs that are being addressed (this doesn’t really change).
I read a lot of rubbish about how BI is “backward looking” and analytics is “forward looking.” I can assure you that BI has always been about actionable information. At the end of the day, what counts is using data to improve the way you do business. Call it whatever you like, but vendors in particular shouldn’t try to belittle what we’ve been doing for decades just because they have some new technology they want to sell.
TE: I present regularly on topics such as “why BI projects fail and what to do about it” and “how to implement BI competency centers.” Overall, there’s higher maturity in BI, but it’s widely scattered, with each new generation relearning most of the same lessons.
The key area of improvement that is required is to always remember that BI is about people and the business, not about the software architecture. The technology is, of course, often a challenge, but when BI projects fail, it’s almost inevitably a problem with organization, culture, expectation setting, and business alignment. If you run a BI project, you should be spending more time on these things than the underlying IT infrastructure.
TE: MapReduce, Hadoop, and related technologies have proven their worth in enterprise contexts surprisingly quickly, and every vendor in the BI space is busy providing tighter integration. The latest version of SAP’s Sybase IQ database has tight links with Hadoop, and you’ll see a lot more coming out this year, both in terms of integration and best practice (i.e., determining where it makes the most sense to use these new technologies).
TE: As I sat down to gather my thoughts about BI in 2012, I quickly came up with the same long laundry list of BI topics as everybody else: in-memory, mobile, predictive, social, collaborative decision-making, data discovery, real time, etc.
All of these things are clearly important, and we’re going to continue to see great improvements this year. But I think that the real next big thing in BI is what I’m seeing when I talk to customers: they’re using these new opportunities to not only improve analytics, but also fundamentally rethink some of their key business processes.
Instead of analytics being something that is used to monitor and eventually improve a business process, I’m seeing analytics become a more fundamental part of the business process itself. One example is a large Telco company that has transformed the way it attracts customers. Instead of laboriously creating a range of rate plans, promoting them, and analyzing the results, it now uses analytics to automatically create hundreds of more complex, personalized rate plans. The plans are then thrown out into the market, monitored in real time, and those that aren’t successful are quickly culled. It’s a way of doing business that would have been inconceivable in the past, and will be a lot more common in the future.
TE: When at Business Objects we used our own technology to run things, we put a French spin on the phrase “eating our own dog food” and claimed that we were “drinking our own Champagne.” So that’s my answer!