What’s the Big Deal About Big Data?
by Karyn Hodgson
As industry buzzwords go, "big data" may not top the list for a majority of integrators today; but that could change very quickly.
What is big data and how does it benefit you? If you are like many security integrators the subject of big data may barely be on your radar yet. However, looking to the future of the industry, it seems inevitable that as more and more devices become both smarter and more connected (think Internet of Things), the desire to do more than just sit on all of that information will prompt users to demand more from their security system, their integrator — and, increasingly, their data that may go way beyond the scope of “security” today.
With a job position that literally looks toward “what’s next,” Rob Martens, futurist and director of connectivity platforms, Allegion, Carmel, Ind., says big data is poised to take off. “Big data means extremely large data sets that can be analyzed by computers to reveal patterns, trends and associations, especially relating to human behavior or interactions. It is literally terabytes of data. You and I would look at that and say ‘I don’t even know what to do with that.’ Yet we are collecting all this data right now that ultimately just sits there and wastes away. With the low cost of sensors today and the capability to generate more and more data, the key thing to understand is that processing power is so accessible and inexpensive we are now actually able to actively analyze data to see patterns in it that before just looked like chaos.”
While big data is relatively new in the security space, there are several reasons integrators should be looking into big data right now, industry experts say. Big data is only going to get bigger; security has a role to play in big data; and those that understand how to work in this space will potentially see a big return on their efforts.
What is ‘Big Data’?
Martens uses the acronym “ETL,” or extract, transform and load to describe big data. “Imagine a big tower full of information that we pulled from all these different systems and devices. Now take a hose and plug it into that tower and start to suck stuff out. That process is ETL. Extract is information from X time period or parameters. Then I am going to transform that data, clean it up and load it into a common format.”
In the world of big data, definitions abound, in part because it is a complex concept to understand. Two other ways of looking at it break it down into four-step processes.
One way to describe big data is “capture, process, analyze and visualize,” says Kevin Wine, vice president of marketing, video and situation intelligent solutions, Verint Systems Inc., Columbia, Md. Big data also comes in two flavors, he explains — structured and unstructured. Structured data is easily searchable using computer operational instructions. “Unstructured doesn’t have tables or anything like that. It isn’t even close to being useful. The weird thing about big data from a definition perspective is it has nothing to do with the size of the data sets. It is more about the complexity. A lot of information that is easy to process isn’t big data.”
Joe Leung, analytics marketing manager, HP Surveillance, Palo Alto, Calif., has a slightly different take on the four-step big data process. “Big data is characterized by the four Vs: volume, velocity, variety and value. There are massive amounts of data in vastly diverse formats and sources. These different sources (variety) are coming together and growing at an explosive rate (volume). Velocity is the speed of the information coming at you; there is a lot of value trapped in that data.”
Big data is not new, even for some in the security industry. Manufactures such as Verint, HP and others cited in this article all have or are working on some form of big data solution. But for the majority in this industry, it is new to them, says Dan Murray, director of sales, global key accounts, Bosch Security Systems Inc., Fairport, N.Y. “This is a relatively recent topic in our space. Part of that is a reflection of the changes in technology that have occurred in the last 18 month to two years. We have standards like ONVIF that are now trying to deal with metadata. There are massive volumes of structured and unstructured data that by its very nature makes it difficult to process in the world of traditional databases and software.”
Integrator Yilmaz Halac, vice president of client technologies, SDI, Chicago, agrees. “Big data … grows exponentially. It is difficult to mine and manipulate that with the old fashioned databases everyone is using today. A lot of data doesn’t always mean big data. It has to have intelligence on the back end to support the concept.”
Halac describes his company’s current approach to big data as preparatory. “A lot of what we are doing as a firm right now is getting clients up to speed and ready for big data. Big data is in its infancy stage right now. People don’t know what they have and don’t want to overwhelm themselves. But in the next two to five years I think we are going to see a lot of investment in this. It is a fairly new concept and it is our job to educate customers about the benefits (see SDI’s blog “Realizing the Potential of Your Existing Data” on SDMmag.com).”
What Can You Do With Big Data?
Big data is much more than collecting information. The real magic happens in what you do with it.
“When we think about big data and making it valued for customers, a lot of folks in IT use this word called ‘actionable intelligence,’” Wine says. This is truly the fourth (and even fifth) step in the process, he adds. “Now that you have captured, processed and analyzed, it is time to present it. We then recommend a next step: action.”
While video analytics and PSIM solutions have touched on this concept before, big data goes far beyond security.
“This is a much bigger game and drives a lot more for the customer about productivity, reduced costs, better response times, mitigating risks and ensuring compliance with operational procedures,” Wine adds.
In Marten’s ETL description, “load” is the part that turns big data into an actionable tool, he says. “That business intelligence (BI) tool allows us to map, graph, demonstrate, twist, turn and pull the data in a bunch of different ways. The larger the solution set, the more beneficial your solution is going to be.”
Big data solutions can potentially include information from systems as far flung and varied as security, social media, public call centers, internal machine data and even the dark Web.
“The value of big data is an interesting concept,” says David Humphrey, CTO, video and imaging for HP. “If you are watching CNN and a reporter in Philadelphia is talking about a train crash, it is probably true. They are not going to lie about a train crash. But if you saw one or two tweets about a train crash, you would have to be more cautious about that.
“There may be a ton of small information that, when put together, can actually produce a valuable result. Or there could be a ton of rubbish, but in there is that one gem. It is all about whether you can use statistical analysis to derive value from the lower level or go through the weeds to find one flower.”
Martens agrees. “You never know what you are going to find and that is where the most exciting processes occur. We will start to see more and more innovative ideas come forward. It will go from big data to BI to new ways to attack a problem nobody necessarily realized was there.”
Hank Monaco, vice president, marketing, Tyco Integrated Security (TycoIS), Boca Raton, Fla., breaks it down this way: “There are two components to big data I see — business intelligence and data intelligence. The first is around how the business operates and how to set up rules. But we have begun to do things that over time will create relationships around the data itself. That gives a different insight into the productivity or efficiency of your systems. It is in very early stages but we believe it will provide real benefits for our customers.”
What big data can do is limited only by the imagination of the people conceiving and designing the solution, Humphrey says. “It takes a level of imagination to step out of the current norm and say ‘If I had this, I could do that.’ It is a new field, and one that is coming quickly.”
The difference between PSIMs, video analytics and true big data is that the former gives you the ability to connect disparate sources of similar data; big data actually connects across silos. Because of this, multiple departments can benefit.
“When this big data platform is considered to be deployed, every single department will be part of it,” Halac says. “Everybody generates data and consolidates it. It will tap into all the departments — a single solution everyone can use.”
That was the philosophy behind 3VR’s big data solution. “It is a relatively new business for us, but as we looked at our core verticals we were good at catching the bad people; then we thought, what can we do to help customers learn more about the good people?” says Jeff Karnes, senior vice president, marketing and operations, 3VR Inc., San Francisco. “A lot of it is breaking out of old habits and opening yourself up to the broader use cases.”
Dan Cremains, director of product management, March Networks, Ottawa, Canada, says the retail environment is a classic example. Video cameras and security systems are being developed and used in that to do more than just security — a type of smaller-scale “big data” solution. “In our traditional retail world, for example, there are entities interested in other things besides security, including customer service, staffing, operations and compliance. People in charge of that part of the business are looking for tools to help give them answers.”
Charlie Erickson, executive vice president of product management, 3xLogic Inc., Westminster, Colo., adds: “We focus on showing you the outliers. We say, ‘Spend your time looking at this.’ If you have 8,000 employees and two analysts employed to look at theft, we can say, ‘Here are the top 10 most likely people who did X behavior across your company.’”
Retail is certainly an early adopter of big data, but even what is happening in that vertical is only the tip of the iceberg in terms of what big data is capable of. At its core, big data is a proactive tool to make business processes easier, better, or more profitable.
“When you are talking about big data, the sources of information can be very broad and that is the whole idea behind it,” Murray says. “It is much larger than a security technology discussion. But an important question [for the integrator and their customer] is how do you create additional value out of that security investment? A company has spent tens of thousands to deploy this technology; now what other utility can they find for it within the organization?”
How Do You Sell Big Data?
Just as most integrators have finally become comfortable with the IT department in the room, selling big data will add even more players around the discussion table.
“It adds new decision makers but it also creates opportunity, in my mind,” Murray says. “Where in most cases security technology is still looked upon as a cost center within most organizations, if you can extrapolate and pull additional value into other areas that are decreasing costs and adding operations efficiency, it is an opportunity to not only differentiate yourself, but also charge more for the services you are offering.”
It ties back to multiple stakeholders having multiple budgets, Humphrey says. “You have a marketing budget and a security budget. If you can get them collaborating, everyone will get a better solution at the end of the day.”
Sometimes the customer is the one asking for a solution, but more often the integrator has to be prepared to sound them out on what their needs are.
“It begins with us having a really good understanding of what their problems are and their unmet needs,” TycoIS’s Monaco says. “Who are the key stakeholders and how can they benefit? How can we be sure we are providing the largest benefit? If you only focus on security you run the risk of being sent to procurement. We want to be involved early on in the planning process. If we can become a better partner, if we can be viewed in a more elevated way and demonstrate that our systems and services can provide benefits beyond security, we can get access to those other stakeholders.”
Those stakeholders are often at the top of the organization, Karnes adds. “It is definitely an enterprise sell, a solution sell. It takes skill to do that. What we found when we launched our product was we were talking to the chief marketing officer, the chief data officer, even the CEO — a lot of chiefs were in the room and very interested in how we could improve their business. At the same time, loss prevention appreciated that more people were in the room because now maybe they could get the budget to get that IP camera they wanted.”
Cremins says the questions security integrators are used to asking can be adapted to this new, wider audience. “Traditionally they talk to an end user and say, ‘Tell me about your security issues and problems.’ They should still ask those same questions, but take out the word security. ‘What are your overall business concerns? What are you worried about in terms of growing your business?’ It turns into a whole different discussion.”
Halac points out that just because the audience may be “new,” the client doesn’t have to be. In fact, many existing enterprise customers may already be in a good position to look at doing some big data. And if they aren’t, it is the integrator’s job to educate them about what it is and help them get prepared. “The majority of our customers need education about what big data is as well as ensuring them they have the right stepping stones to get there. What kind of analytics are they running? Is their data prepped and ready to do a big data platform? If we sit down with them and educate them, they realize that a lot of the questions they are facing right now can be easily solved using this big data platform.”
Big data is definitely looming on the horizon, and integrators should be looking at it now, Erickson says. “I think it is only going to grow. We have all these devices. We are starting to collect data from where people are walking around in a building. There are all these sensors. The fact is we are already getting so much data from beyond traditional security. How do we deal with all these devices? Integrators will migrate to this because it plays well with cloud solutions and recurring revenue (see sidebar, page 76). People are looking less and less at the physical piece of hardware and more at where they are going to make their money. Big data becomes a part of a solution sale and less of an equipment sale.”
Martens adds, “People in our industry might not have thought about big data as something they are generating, but they are. There is huge value in it. We will start to see correlations that we have never seen before.”
SIDEBAR: How to Get Into Big Data
Big data can be an overwhelming concept for both providers and users; but those who have already gotten their feet wet have some sage advice for integrators looking to explore this avenue. Big data isn’t easy. But it can be well worth the time and energy to examine it.
“Go to the manufacturer for help,” advises Kevin Wine of Verint. “We are squarely in this game. We are always looking for the right relationship with a great partner who wants to think about this long term. If not us, do your research and plow into how other providers do it. There are actual solutions out there, so ask them to demonstrate what they can do. Understand the technical capabilities you will need to do it. It is not for the faint of heart.”
Solutions range from security-based (often video analytics) to cloud and Internet provider solutions.
“We are supplying big data solutions to customers around the world,” says David Humphrey from HP Surveillance. “But we are not experts in everything. The integrator needs to be able to bring in that last 20 percent that turns a generic big data solution into that customer’s preferred solution. In order to do that, you have to know your customer well. You need to bone up on the technologies, understand the problems with data. It’s not rocket science, but it is all important. If anyone says it is easy, they are wrong.”
You don’t become an expert in big data overnight, says integrator Yilmaz Halac, SDI. “You first have to understand all the components, the need, the opportunities and what you are getting into. You have to know the tools, platforms and how to consolidate all these components.”
Depending on what you are trying to do, going with a provider that at least understands the security space can have its advantages.
“There are lots of products out there made for looking at big data,” says 3xLogic’s Charlie Erickson. “But which ones have been applied to areas a security integrator will be in there talking about? They want someone who has developed a solution that meets the needs of their customer. Integrators should be finding trusted solution partners that can go in with them on joint presentations.”
Different sales skills are sometimes needed to effectively sell big data, says Jeff Karnes of 3VR Inc. “You are speaking a different language and need to have a different credibility. When it comes to security you have all the vernacular and understanding already. The credibility just oozes out of you. But when you start talking business intelligence and you are new at it, they can sense that very quickly.
“That is a transition that takes time. First you have to recognize it; second you have to hire it; third you have to align it and create a structure that is not self-competing.”
Dan Murray, Bosch Security Systems Inc., advises making sure your company’s internal staff and processes are up to doing big data. “First and foremost identify where the holes are in your current offering so you can identify your deficiencies and what resources you don’t currently have from a product or manpower perspective. Then there is the ‘buy or make’ decision. How long can you provide this using subcontractors or third-party consultants verses having that capability in house? If you take the lead tech in your organization and develop them over the next 18 months or so, then when there is a shortage of talent — and there will be — you will already have someone in place.”
“Know thyself” is a good tenet to keep in mind. “That is my number one piece of guidance,” says Allegion’s Rob Martens. “Start with the root basics. One thing that makes it an intimidating space is it is almost too flexible. You can get lost in the data. Have an idea of how you want to use it, what you are looking to get out of it to start with. Once you realize that one size does not fit all, then talk with providers about what they specialize in; figure out what you want and go have conversations with the people that have those specialized skill sets.”
SIDEBAR: What Can Big Data Do for You?
There is a lot of discussion around what big data does for the customer. But what does it do for the integrator? Many big data solutions are cloud-based, making them ideal for an expansion of RMR-rich managed and hosted services.
“There is a huge role for security to play in big data,” says Verint Systems’ Kevin Wine. “For those integrators that have crossed the chasm already concerning convergence with IT and are either already offering managed security services, they can now broaden that constituency base to other departments. For that integrator it is a natural progression to offer other types of professional services. Ultimately this could lead to things like service contracts that will allow them to know a system is experiencing issues before the customer is even aware of it.”
“Many larger integrators have already begun their own big data initiatives,” says Jeff Karnes of 3VR Inc. “A lot of that is wrapped around the RMR, Software as a Service and cloud that is very much tied into big data as a delivery mechanism. That is an interesting factor for a lot of integrators because it is recurring revenue.”
The IT-centric nature of big data also means that there is another player that is already in this space: the IT integrator.
“For the short term I see this coming more from that side,” says Bosch’s Dan Murray. But, he says, the fact that IT integrators will be out early with this just means the security integrator should jump on this now so they don’t miss the boat. “It is the only way to control their destiny.”
And the integrator does have a unique role to play, SDI’s Halac says. “There is a lot of siloed information an intelligent integrator can glue together and make easy to understand for the decision makers. It takes an integrator to play a big role in finalizing connecting the dots.”
Ultimately, each integrator has to decide if big data is right for their business, but Karnes advises to think seriously about it. “Do it. Don’t be afraid of it. Big data is here and it is staying here. It will be a very important part of the growth in IT. If you want to be part of that growth you will have to have a big data play.”