We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

The Importance of an Enterprise Analytics Platform: A Spotlight Q&A with Dan Lahl of Sybase

Originally published March 12, 2012

BeyeNETWORK Spotlights focus on news, events and products in the business intelligence ecosystem that are poised to have a significant impact on the industry as a whole; on the enterprises that rely on business intelligence, analytics, performance management, data warehousing and/or data governance products to understand and act on the vital information that can be gleaned from their data; or on the providers of these mission-critical products.

Presented as Q&A-style articles, these interviews conducted by the BeyeNETWORK present the behind-the-scene view that you won’t read in press releases.

This BeyeNETWORK spotlight features Ron Powell's interview with Dan Lahl, Senior Director of Product Marketing at Sybase. Ron and Dan discuss predictive analytics and why an enterprise analytics platform is crucial for analytics success.

Dan, why is it so important to an enterprise to have a comprehensive platform strategy for business analytics and big data?

Dan Lahl: Well Ron, I think that's the $64,000 question that businesses are really asking themselves today. The current strategy that dates back to the late ‘80s or early ‘90s goes one of two ways – either create an enterprise data warehouse that has everything in it, tries to get everything in it cleaned up, and then you answer questions out of that, or spin up data marts that are application specific. And we've been struggling for a long time about how to integrate those two delivery strategies. But at the end of the day, it really doesn't take everyone in the enterprise or all the silos of data across the enterprise into account.

For example, the needs of a data scientist are vastly different than the needs of an executive, and those are vastly different than the needs of a second line manager in retail who has to make a decision in the next 15 or 20 minutes about a product strategy at a store. The problem is that you have all these transactional systems, but they're not tied together into one comprehensive platform for analytics and non-transactional big data. What enterprises need is a comprehensive strategy. What we're talking about at SAP Sybase is a comprehensive platform strategy for actually taking all of the data, structured and unstructured together, that exists in the enterprise, across all the silos that have different data latencies and different user requirements, taking into consideration all the different users’ needs from building predictive analytics, to operational dashboards, to recurring reports, to ad hoc analysis, and getting all of that in one comprehensive platform so you can service all the users with all the data across the enterprise. In a nutshell, that is why we think enterprises need a comprehensive strategy that enables them to meet all the analytics needs of all the users across the entire enterprise.

That makes a lot of sense. Many vendors have taken the approach of adding a Hadoop distribution to their data warehouse offerings. Sybase has taken quite a different integration approach between Hadoop and your data warehousing offering. Why is that?

Dan Lahl: Well, it goes along the lines of the last question. Hadoop is just another data silo. So now, in addition to an enterprise data warehouse, you have a set of data marts, maybe one or more spreadmarts and then you have a Hadoop distribution to add to this complexity. What we think needs to happen is you need to take the data that you have in Hadoop and integrate that with your analytics platform. So it's not just spinning up some free version of software like Hadoop, but actually making sure that the data that's important to the enterprise is integrated from Hadoop into your analytics platform strategy. We think it’s important for enterprises to actually integrate the data across all of those three environments – data warehouse, data marts/spreadmarts, and Hadoop – to gain a holistic view of information and access.

Hadoop is really another data source so to integrate that into an entire system makes a lot of sense. We're seeing most enterprises today are working hard to make business intelligence and analytic capabilities available to a lot more people within the organization. Is that what you're seeing, and how does Sybase IQ 15.4 help with that?

Dan Lahl: Well, that's a great question, and that's exactly what we're seeing today. Back in the early 90s, you built the system and you had the two smartest people in the company doing predictive analytics for the next round of trends that you were going to see. You probably gave a report to the executive team every day, and that was it.

Now we're seeing that people across the enterprise want ad hoc queries and self-built reports, to be able to interact with analytics the way they interact with Google. They want to be able to do a search on their piece of the business with the information that's pertinent to them. Now, not only do you have the data scientists doing strategic and predictive analytics, but you also have the executives running strategic daily or weekly analytics, operational managers that need information that maybe they need to make decisions on in the next hour, and then you may actually have applications that need alerts that have to take place in seconds or milliseconds, and, in the case of Wall Street, maybe microseconds.

So we're seeing that, again, this enterprise analytics platform idea has to be able to serve the needs of a number of different user constituents. And again, I think it's all driven by what we've seen as this kind of Google effect. You want to be able to type in something and get an answer that is pertinent to what you need to do run your business. I heard Geoffrey Moore say something really interesting. He said that in our personal world on the weekends, we can get access to any kind of data we need, and so we're superstars on Saturday and Sunday. Then because our company doesn't give us access to the right information and the right tooling to help us do our job better, we're idiots on Monday morning. We go from being superstars as consumers interacting with the Internet, but then become less than superstars on Monday mornings at our job.

What we have to do is provide them the right data assets at the right time with the right tools to make the right decisions at whatever level people are in the organization, whether that be frontline worker, executive, or data scientist – and embedded directly into business processes. Sybase IQ 15.4 does that. With 15.4, you can actually take subsets of your data and subsets of your analytic platform environment, and you can dedicate these to different groups. So let's say the executives really want to deal mostly with high-level financial information. You can set that aside so only the executives have access to it. Similarly the marketing department wants 360° views of customers. You can set aside that information within Sybase IQ, and you can set aside the compute resources that group needs to answer those types of questions.

An example of that would be one of our customers in the UK. They have a call center application that runs 24/7, and that application has a ton of resources because there are 3,000 people in the call center. But then they also have data scientists that are looking to give the call center people better options to offer to customers who call in. Well, the data scientists have their own compute resources that access the same customer data tables, and they actually can run their own analytic algorithms without impacting the call center, which needs sub-second response time. Sybase IQ 15.4 helps to do that through a concept that we call virtual data marts. You can set up different application areas within the 15.4 environment and optimize for different workloads, different applications, and different people needs. And that's all in one analytics data platform with Sybase IQ 15.4.

That actually leads into my next question. With more and more people using Sybase IQ 15.4, do you see more organizations doing advanced analytics like data mining, data discovery, and predictive analytics to grow their business? What is Sybase seeing, and what are you doing to enable these types of new analytics?

Dan Lahl: We really see that as one of the big moves within the analytics space. We still see that people need to do these bread-and-butter reporting applications, and they need to have some ad hoc query capabilities so they can ask these Google kinds of questions. We also see the need for the application-to-application stuff on the low data latency side. But we see a big move in the marketplace toward advanced analytics, data mining, data discovery and predictive analytics. We see the number of data scientists within enterprises growing dramatically. The data scientists are the ones who troll over huge amounts of data looking for trends and anomalies – for that needle in the haystack – so that, for example, they can recommend to the company that they take one product, map it up with another product and the customers will buy those two things together for more money.

We see a lot of that happening in the marketplace today. And the way that we've addressed that with Sybase IQ is we've added a native MapReduce capability directly into the Sybase IQ product. A lot of these data scientists are using MapReduce techniques. What Hadoop uses, we've actually created that capability right inside of Sybase IQ.

But in addition to that, we've also added the capability through partners like Fuzzy Logix and KXEN to run the predictive analytics algorithms directly inside the database. The old model with predictive analytics tools was to extract a subset of your data into the tool to do your predictive analytics. The problem with that model is that, first of all, the extract is slow. Second, you don’t have a full dataset, and you're not able to really analyze as many attributes because you're doing a lot of data movement into the workbench tool. If you can do the advanced analytics or predictive analytics directly in the database with these algorithms from partners like Fuzzy Logic and KXEN, you can actually run these algorithms much faster over much more data and many more attributes. The bottom line is that you can get better answers more quickly.

So we've added partners to the mix, we've added the in- database analytics to the mix, and we've also allowed the capability for people to write their own algorithms. So if you want to write your own algorithm and run it in inside of Sybase IQ, you can do that as well. Data scientists, the quants on Wall Street and the insurance actuarials like that capability. So again, we've added a lot of capabilities in Sybase IQ 15.4 to address this wild new world of predictive analytics, advanced analytics and data mining.

Well Dan, when you look at analytics, obviously Sybase IQ was one of the first databases for doing analytics and has had a long history. We are now seeing an acceleration of adoption of analytics by enterprises, and we're moving into a new phase of innovation. What advice would you give enterprises looking to build out a data management strategy to support analytics?

Dan Lahl: You're right Ron. We have been the leader, and I would say, back in the early days, kind of the odd duck as one of the big database companies to have an offering specifically for analytics. Back in 1996, we first delivered our column-oriented database, Sybase IQ, that actually organized information not by row, which is how most traditional databases have organized data, but actually organizing by column. And if you think about how you do transactions, you insert rows into your database to capture transactions. But when you want to analyze the information, you actually are analyzing the attributes or the columns of the information stored in the database.

So way back in 1996, we had a view that this was really the right way to do analytics. We recognized taking a system designed for one thing – for transactions – and trying to make that an analytics database was not the right way. And we've seen the market actually come to us. In the last couple of years, there have been a number of the other database companies actually coming out with column-based compression. They're really not doing pure analytics on column systems as we would define it with Sybase IQ, but they're actually organizing some of their data by columns for compression purposes.

But to get back to your question, when deciding on a strategy for analytics, enterprises need to look at a system that is architected from the ground up to do analytics. So don't look at a traditional row-based database that's great for transactions, but has to do a lot of gymnastics to do analytics. The other thing we believe you ought to look for is something that has true column integration that really does, from the inside out, look at the information in a column-based manner.

But similarly, we at SAP Sybase think you ought to look at one platform that has not just the column-based capabilities of Sybase IQ, but also integrates in-memory processing with SAP HANA and sharing information between SAP HANA and then Sybase IQ, and then also look at low latency products that are used for quick decision making like Sybase ESP, a complex event processing engine. An analytics platform should encompass those three things as well as flexible integration with Hadoop. The advantage of the combined SAP and Sybase is that we have interaction between all of those different products between ESP and Sybase IQ, between ESP and HANA, HANA and Sybase IQ, and so on and so forth. So start with the column base, go to the in-memory analytics with SAP HANA, and then low latency with Sybase ESP. We think a great strategy for enterprises is to implement those across all of those different data latency needs as well as user needs.

And everybody's needs, especially from an end-user perspective, differ so it makes a lot of sense to transition from one to the next, and it seems like you have that strategy well in place.

Dan Lahl: That's right – and the interactions between those products as well. We've already built data feeds between Sybase ESP and Sybase IQ, Sybase ESP and HANA, and HANA and Sybase IQ as well. Those data feeds are already there. Pick those products for the specific needs, and we can tie them all together to meet, again, the data latency needs and the user requirements across those three products.

Well Dan, I want to thank you so much for taking the time to talk with me today about Sybase and the benefits of a comprehensive platform strategy for business analytics and big data.

  • Ron PowellRon Powell
    Ron is an independent analyst, consultant and editorial expert with extensive knowledge and experience in business intelligence, big data, analytics and data warehousing. Currently president of Powell Interactive Media, which specializes in consulting and podcast services, he is also Executive Producer of The World Transformed Fast Forward series. In 2004, Ron founded the BeyeNETWORK, which was acquired by Tech Target in 2010.  Prior to the founding of the BeyeNETWORK, Ron was cofounder, publisher and editorial director of DM Review (now Information Management). He maintains an expert channel and blog on the BeyeNETWORK and may be contacted by email at rpowell@powellinteractivemedia.com. 

    More articles and Ron's blog can be found in his BeyeNETWORK expert channel.

Recent articles by Ron Powell



Want to post a comment? Login or become a member today!

Posted March 14, 2012 by Anonymous

Thanks for the interesting article.  I really like the Geoffrey Moore quote.  It made me laugh outloud because it is so true.  Companies must learn to evolve or face the fate of Kodak.  Thanks again 

Is this comment inappropriate? Click here to flag this comment.