This BeyeNETWORK spotlight features Ron Powell's interview with Michael Corcoran, CMO, and Jake Freivald, Vice President of Marketing, from Information Builders. Ron, Michael and Jake talk about the latest technologies available to corporations today for self-service BI, big data, analytics and mobile.I would really like to talk with you two about the trends youíre seeing in business intelligence and big data. Letís start by talking about self-service business intelligence (BI). That seems to me to be a very big focal point in most organizations today. They are really trying to get their information to a wider audience and make it easier to access.Michael Corcoran:
I think there is a lot of interest and activity, but I think there is also a lot of confusion about what is self-service for business intelligence and analytics today. I think a few years ago we saw a large community of people focused on how to make BI available to a broader community. And, at the time, they were somewhat relegated to the power user tools that traditional BI vendors were providing. And what you had was a very real gap between the more technical power users and business analysts versus the less technical, operational people Ė employees and partners who might have needed information. So you would tend to have this gap between people who had very sophisticated interactive tools and other people who were dealing with very static information. When those people, as they normally would,† started to say they wanted more information and wanted it to be more interactive, they were forced to give them the complete tool. Then youíd get a lot of shelfware because it was just too complex. Today you see a lot of activity. People talk about discovery tools, and again weíre seeing a trend where tools that are really nice for some analysts are now being bought on a larger scale for more people. I think youíre going to end up with a lot of confusion because these tools are really designed for standalone analysts who work with their own sandboxes of data. Now youíre going to try to deliver that to a broader scale of people, and I think youíre going to end up with a lot of inconsistency of information, islands of data that are out of sync, and youíll have people going out to the boardroom with different answers to the same question.
Where I see the real opportunity and the way we view it is that there are people who need tools, and if you think about the iPad or Apple concept, there are people who need apps. An app is something that you can deliver. Itís highly functional and highly interactive. It could be very sophisticated, but have you ever downloaded an app and used the user manual to figure out how to use it? You havenít Ė theyíre naturally intuitive. Our approach to BI self-service is about figuring out who needs tools, giving them everything from complete development tools to ad hoc tools for the power users, etc., and the added capabilities for visualization, predictive analytics
and location intelligence.† So allowing us to leverage all of that and deliver highly functional, interactive apps for the masses. Those could be people who are operational† employees, partners and even end customers outside of the firewall.Jake Freivald:
A key element there is that you have to be able to access the data because the people who need that information say that the data lives over in their operational systems, and you may not have a full-blown data warehouse built out of it. You need to be able to provide a subliminal way that they understand; and, as Mike was saying, thatís not typically going to be something that is drag and drop because thereís always the question of what happens when I let go of the mouse at this point. What happens when I drop? So it really needs to be the kind of thing that you would see more on an eBay or another website of that sort rather than the kind of tool that they have to be trained on.That sure makes a lot of sense, and really it leads into my next question about mobility. You recently talked about extending to the iPad and multiple devices. Can you comment on your mobility strategy?
We actually took a very early stand on mobility a few years ago. At the time it was considered a little bit controversial, and probably less of a leading strategy. When the iPad and some of the newer tablets came out, we decided not to go out and build a native environment for each platform. We decided weíd invest very heavily in the early HTML5 technology, and push that technology as far as we could. Over the last couple of years, weíve really seen that evolve. So we were able to implement a mobile platform strategy that would allow us, from our server, to build intelligence and deliver content that would act like a local app, it would look natively on an iPad like an iPad app, take on the native gesturing, the native menuing look and feel, but all of that essentially delivered from a server. Weíre not typically in the business of just building an executive dashboard for five or 10 execs. At that point, you could buy them all iPads Ė that is a simpler deployment strategy. For us, we have these very widely used operational applications that can touch upwards of millions of customers or users, and at that point you canít assume what devices theyíre using. You canít dictate it, and you canít buy them the hardware to build it each time. So if we had to have a strategy that would say we donít know if theyíre going to use Apple, or Android or Microsoft technology to view this or whether itís going to be mobile or on a desktop or laptop. So it had to be all-encompassing, and it had to be intelligent enough to deliver it in a way that had a more localized look and feel.Jake Freivald:
I think everyone talks about BYOD (bring your own device), which is a big deal inside the enterprise, but so many of our applications go out to customers and partners like Mike touched on. So you can provide people analytics to look at their relationship with you or how you are doing at fulfilling their orders or whatever it is, but you have no way to control what device theyíre using. You have no way to train them. You have to focus on something that will go on literally any device. So thatís a big reason for the† HTML5 strategy. Weíve proven that it works in many areas now, and I think itís going to be something that our competitors try to adopt† in order to catch up to where we already are with those kinds of deployments. Itís interesting that you bring up BYOD because itís becoming a headache for a number of companies from a security standpoint. Youíve addressed it by being very open. The ability to put the data out there on any device, following a standard format, confirms that you picked the right strategy.Jake Freivald:
It was a little bit controversial when we started with HTML5, and we got beat up a little bit by certain people. We donít get the kudos now that say we really picked the right strategy, but our customers appreciate it. And we do have security concerns, of course. You have to make sure that if the device is lost that everything can be wiped. You need to put on password protection and things of that sort to make sure that somebody receives it, knows that it is private information and that they can open it. We have strategies for that, and we work with all of our customers to make sure that is taken care of.
In talking to your customers, one of the key points they brought up is the fact that you can connect to anything. And itís that whole connector strategy Ė the ability that if thereís data anywhere, even big data, that you can go out and get that data for them and work with it.
I think youíre absolutely right that getting access to data is critical because a lot of the people who are using mobile devices now are not analysts themselves. An analyst might go back to his or her desktop and thatís where they do their deep thinking. But when youíre talking mobility, it is somebody who is in a truck, itís somebody who is on a plane or it is somebody who is doing operations on a factory floor. And because of that, you have to be able to get the information from those operational systems, you have to be able to pull information out of the cloud, or sometimes itís a combination of those things. So you need to get information from the cloud and enrich it with local operational systems. All of that information has to be put together and brought to bear at the single point where people are making decisions Ė and thatís on their mobile devices. So thereís no disentangling the idea of data access from the idea of mobile computing. Those things have to go hand in hand, and being able to achieve that is an important part of our mobile strategy.Michael Corcoran:
When you look at the requirements that people have today in real-time as part of their business model, they have to assume that they can get to any data. And with our technology, they certainly can. For us itís not just about reach anymore. Itís about latency. Itís about how timely you can deliver that information to people. Even analytical systems are becoming very operational. We have real-time decision making. Weíre seeing that in business, government and law enforcement. Healthcare, obviously, is a very time critical sensitive area for having patient information delivered to the people who need it. So, itís about reach, and itís about latency.
The other area that weíve invested in very heavily is data quality
because now itís become very apparent to organizations that the quality of the data is equally important. There is too much risk at stake here. There is too much legislation and liability that people cannot just assume their data is good enough. Data has to be accurate. It has to be validated. It has to be consistent for everybody. With big data and with mergers and acquisitions, itís becoming a more complex data architecture within every organization. Those are areas where we continue to invest, and theyíre equally as important as the reach.
In my 35 years in this industry, Iíve never seen as much hype as I have with big data. What are you seeing and where do you feel big data is today?Mike Corcoran:
I just watched a presentation by Wayne Eckerson. He calls it ďbig data, big schmada.Ē I wish we werenít as focused on the data as a technology. Letís just assume weíre going to collect more and more data because we are. And just because we collect it shouldnít mean that we should be focused on that. I think if we can learn anything from the 1990s and early part of this century, we had built very big data warehouses in lots of organizations but with very few users. A lot of the big data emphasis Ė whether it might be analyzing machine data or analyzing social media data Ė itís still information thatís going to be relegated to a relatively small number of people. It could be valuable; but, for the most part, when information touches very few people, you tend to struggle to understand the real value and ROI of that investment. I hear a lot of people talk about implementing Hadoop. There is a concept of ďhadumpingĒ Ė just dumping data into Hadoop and more history can be caught. What are you really going to do with it? That will be the question. Does it have an operational value? Does it have a managerial value? If it does, I think those are the areas where youíre going to see that return come into play.Jake Freivald:
You know, itís interesting, Mike. You mentioned that one of the things that you have to do with big data is share it. If itís locked up with a single person or a small number of people, youíre not going to get the value out of it. Well, the way that you share big data isnít by sending somebody four petabytes of information or even a couple of gigabytes of information down on their iPad or mobile device. So itís clear to me that the way to share the value of big data is to condense and condense and condense and extract a small amount of data from it. Something that can be shared such as a transaction that is likely to be fraudulent. Itís not petabytes of information. The petabytes give you enough information to be able to say, ďHereís a fraudulent transaction.Ē But the thing that you share, the thing thatís interesting, is that single thing that we can put in a sentence. Thatís small data. So you have to bring to bear all the analytical capabilities that you have in order to get that. You have to do trend analysis, you have to be able to do textual analysis to pull out sentiment from social media data, and you have to be able to put large amounts of data into a bigger columnar data store, for example, to be able to extract information from that in a fast way to get a single answer. And thatís the data that gets shared, and thatís the data that brings high value to everybody from the senior executives to the contact center
to your customers and partners. Thatís what really brings benefit.Michael Corcoran:
Jake brings up a very important point. First of all, chunking it down to something meaningful and then making it part of the other things youíre doing and where it plays a role. From my perspective as a marketing executive, when I run a campaign I donít look at social media analytics as a separate function where someone in my organization is looking at what the customers are saying and whatís the sentiment analysis for it. If Iím launching a campaign, I want to see the business results. That could be information coming out of my sales automation system, my CRM, my ERP, or my financial systems. But then if something is dropping and is tied to a certain area, can I find social media data that can point me to why I might be having problems. So bringing this data all together is becoming increasingly important.
The last major trend Iíd like to cover with you is the cloud. In fact, the cloud was the big buzz before big data. What are you seeing with cloud?Michael Corcoran:
I think a couple of years ago, for those of us in the BI software industry, a lot of people were invested in the cloud. Some of our competitors made the entire platform the cloud for everything they were ever going to do for the future. And I think what we saw was a big ďho humĒ for a couple of years. Very few organizations Ė most of them that would typically put their BI infrastructure, their platform on premise Ė theyíve left their BI data and platform on premise. They did not move it to the cloud. Weíre seeing an uptick now. This year I have to say thereís now more interest. I think people are starting to look at the real opportunities. There are still a lot of questions to be answered. If I am going to have a BI environment or application that might be integrated within my portal from Microsoft, for example, and I have a data warehouse provider over here. Well, if I am going to host that, am I going to host it all in the same place? Are vendors bringing their own hosting services? When I run a query, am I going to have to run it across three different environments across the Internet? You have to plan that out.
But I do see a definite uptick. Where weíve seen more dramatic usage and adoption is by people who build a very specifically purposed application for a software-as-a-service model in our technology where cloud hosting makes a whole lot of sense because typically itís going to be a multi-tenant used application by many of their customers. Weíre seeing a lot more activity for us where other ISVs in the software industry who might have provided an ERP
on premise. Theyíre now moving their application to the cloud. They need a BI platform to plug in that will run in that hosted model, and we provide the best platform for that hosted software-as-a-service model now.Jake Freivald:
Itís funny. I think the conversation weíre having now would have been different a year or two ago because of the definition of the cloud or what people expect from the cloud.† When I first started looking at cloud computing more seriously, people were really looking at software as a service. I swipe a credit card and I can get access to expense reporting. One of our partners, Chrome River, does that. But now more and more as people are talking about using the cloud strategically, thatís almost code for ďIím going to be hosting infrastructure outside.Ē So one side of cloud computing is how do I not have to get IT involved in order to do expense reporting or email or whatever it may be, and on the other side, itís very IT involved where they say Iíve got to deal with having the infrastructure in place so I can run these specific applications in the cloud, and itís cost reduction or whatever it may be. We have to know what weíre talking about, and you end up with very different conversations with different customers. Some people say they really want WebFOCUS to run in the cloud, meaning they want it to be easy and not have IT heavily involved. Other people say they need to integrate WebFOCUS with all the other activities that they are doing. Salesforce.com is kind of the consummate cloud application. So now how do I integrate Salesforce.com in the cloud with my infrastructure in the cloud. Or how do I integrate Salesforce.com in the cloud with my BI Ė which could be† hosted on premises or could be hosted in the cloud. All of those questions come together, and you really have to have the conversation with the prospective buyer to find out what theyíre looking for and what benefits theyíre trying to gain because all of those examples provide completely different kinds of benefits.Michael Corcoran:
Two points that Jake just brought up Ė two examples of things that weíve learned over the last couple of years since these conversations started are in the area of Salesforce.com integration. What weíve learned is that Salesforce.com customers donít want to leave the Salesforce environment to do BI things. And, typically almost every solution has pretty much required that. So we actually took a very different approach over time, and now weíre very tightly integrated. So you log into that environment, and then within that environment we bring WebFOCUS and we bring our dashboards and our reporting directly into that environment with direct data connectivity. So itís a much more seamless integrated experience in the cloud for the Salesforce.com user. And we become invisible. We look like a part of Salesforce at that point. Thatís driven up adoption for that solution. I think, in general, we definitely see that there is that level of integration. It has to be seamless. As Jake pointed out, it has to be easy. So weíve taken over responsibility and now offer a full-service offering. Itís not just saying you can host this on Amazon. We can run there, sure. People really want a managed environment so we have serviced it to manage that software in the cloud for them so itís completely seamless and turnkey for them now. From a customer perspective, with all of these trends, are you seeing some unique use cases from your customers?
Mobile is becoming almost an assumption for us Ė that every deployment is going to be used on any device. And we see a lot of evidence of that with our customers. I think whatís really cutting edge is when a number of these technologies come together. So weíre seeing customers do really interesting things where, sure, it looks like a BI dashboard and it looks pretty, but thereís location intelligence, integrated mapping, and embedded predictive algorithms built in for scoring. The information and decisions that are provided are being much more guided, but itís a convergence of a lot of these technologies. Behind the pretty screens, there might be a combination of structured and unstructured data sources coming together. Thereís a search component, where search is not just a separate function or tool. Search is integrated into that BI process allowing users to drill down, find the data they need and then analyze it. So it is the convergence of these technologies Ė predictive, visualizations, location intelligence, mobile and big data. I think itís exciting when it comes together because the role of IT and the role of information technologyís success is measured by taking what you have and whatever you can leverage and whatís coming and whatís new and just making it work together. Ripping and replacing and waiting for the perfect architecture never arrives, does it? Michael Corcoran:
Itís the convergence that matters more than any individual trend.Thank you. It has been a pleasure talking with both of you about how Information Builders continues to address BI and big data.
Recent articles by Ron Powell