CRN roundtable: the lowdown on big data

By on
CRN roundtable: the lowdown on big data
Page 3 of 4  |  Single page

Part 2

CRN: Do we feel that the conversation about Big Data is still pretty much limited or exclusive to CIOs, or is it seeping beyond CEOs and other decision makers catching on to the dirty words and calling for action?

Chris: I think that you’d have to be on a desert island not to have heard the term.  But I still think people are wondering how to deal with it and work out what to do.  We are sitting around the table talking about we know what the tool set is, and that’s good, I think undoubtedly we do know what the tool set is, but the CEO of a small business down the road doesn’t, and even if they did know what the tool set is, they couldn’t harness it. I still think it’s a really niche thing to do. 

That’s why there are still people in the research sector who are thinking about it.

Rob: I 100 percent agree with that. It’s all about making the data relevant. You can capture so much data, but it’s about making it relevant, and marketing departments are going to want to have certain visibility, and certain pieces of information and linkages between certain data cells as to actually give them reporting or insight that they’re not currently getting.

But then you’ve also got other areas within business which are going to need to see the different pieces of data mining and analytics, for example security departments within organisations which need to see different pieces of converge data to give them the information they need to make their decisions. I think that’s where it’s actually sitting right now.

CIOs are hearing enough, they’re starting to understand that there are opportunities business-wise to utilise all this data, but it’s now time to get down into the real specialist areas within businesses, where they are going to actually be able to make good business use out of them. That’s where I think it’s heading right now.

CRN: Chris you touched on this point via email, the reported distinction between analytics and reporting. It speaks to the point of massive amounts of data and having to do something with it, but then the difference between extracting information and doing something with it.

Chris: That’s right, a couple of people have touched on the subject today. We were talking about getting the report from the data centre. We do get lots of information thrown at us, and I think one of the things that have come out from some of the people around the table today is the opportunity to take that knowledge or that information that we’ve got - sometimes it’s digestive information, not even big data information - and use it. 

I just wonder sometimes if we have got to a point in our businesses, where we are running them so lean and we’re pressing people so tightly, that they don’t even have time. I keep thinking about the accounts department in our office. They’re constantly dealing with invoicing and purchase orders, do they have time to give me the business intelligence that I need?

Peter: That’s the point. One of the issues right now is that IT departments have been asked to cut costs. So cut costs in the IT department, not use IT to cut costs in the business, so I think there’s going to be a bit of resistance there. 

It’s actually a bit of a strange thing, because IT was brought in to actually save the business money, and now it’s seen as a cost centre in itself. That’s going to right some difficulties in the IT department actually providing funding, so when you mentioned earlier who should be approached, I think it’s the CEO and CFO, not the IT department, because they’re going to still be defending having to cut their cost, but if I look at it from a purist perspective, information technology has been, until now, a bit about how do you process things efficiently, the processing.

Now it’s analysis and I think it’s the era of analysing. Things have been processed to death.  You’ve got all sorts of business applications, you can do it all electronically, we’ve done all that – now it’s time to start to analyse, and I think it’s going to be the era of analysis. 

CRN: Do you think a number of CEOs could be justified in turning around and saying this sounds a bit like Y2K, like you’ve got to do it, but we don’t really know why, but you need to invest, otherwise you’re going to fall behind?

Peter: I think like what John said, if you get a couple of organisations who say ‘I’ve done something here and it’s saved me money or it’s got me more sales’ then you’ll actually start to see a domino effect.

Jonathan: I totally agree, but it will be people following the money though. So you’ll get a hedge fund that has the ability to trade and arbitrage in XYZ by analysing all this data and making money, people will be going ‘well that’s exciting’ – or a utility going ‘I’ve found this cool thing’ ---- or the CIO getting a better real-time understanding of risks in whatever, and that’s not only revenue, but I think people will do this. People will make money. 

Google Analytics are a great example of those scenarios and I think that’s what is happening in most of our businesses in different levels of maturity in different scale, but I think we are analysing and getting insight, and those businesses that are faster and better will be more successful and there will be case studies that people start to follow.

Tiberio: So in a way that’s precisely what has happened with internet property. Some large companies were extremely innovative, like Google, like Facebook, and they just broke a few paradigms in terms of which technology they should be using and how they should be using it, and guess what happened, they demonstrated value.

As soon as they demonstrated value companies tried to come along and play the same game. There is no reason why this is not going to happen in other industries.

John: I spent a number of years in the military, and the military has been a leader in the evolution of technology generally. If you go back to the Prussian wars, artillery and all that sort of thing, the military has been using big data for battlefield analytics for a long time. Now that started with humans in data and analysts and all the rest of it, and you can throw in the FBI and CIA in there. 

But if you think about the complexity of warfare now, and what they have been using in big data to get human intelligence signals intelligence, all the other type of intelligence fields into a central command area to determine what is the target, or what are the targets and process that very quickly, with all of that data coming in which is real time unstructured data problem – and then they’re doing that now. 

How did they catch Bin Laden, and you can go on with lots of other examples, and the speed of warfare, and we’re talking real-time.  We’re not talking about holding that for a few weeks and then processing it at a later date, and then doing something with it. I think a lot of the technologies are now starting to come out of there into commercial segments and problems as well. 

CRN: It’s like looking back to the origins of the internet isn’t it?

John: That’s right, it is. Putting my telco hat on here big data needs fibre. Our ability to have fibre around the place is important in our going forward although a lot of people go backwards and say ‘well you don’t need fibre because we’re doing okay’. You’re starting to push big data around, and the way you’re pushing it, it’s going to go to the cloud, it’s not going to be closest inhouse, and you’ll need fibre and stuff.

Chris: That’s a really interesting point. You want to be able to actually use obviously the computer today. You’ve got your own data centre and the big challenge for us is not the ability to do that from a technological perspective, it’s the fact that I’ve got two petabytes of data and I need a fast link to the cloud to do the processing.

Nobody can sell me that at a competitive rate, or at a rate that would make it attractive. So we actually have done the analysis on this, and we worked out that in fact nothing’s going to stop us using the cloud as an adjunct to our hardware, and it’s about the price of the coms.

Scott: It’s ironic really, because you’ve taken a new cost effective processing tool and made it cost ineffective.

Chris: Exactly and I think the whole term ‘communications’, how do we shift the data around, that is actually with current lockup.

Scott: Another form of data, it’s still early days and that’s video analytics. There are some leaders in that space that can do very effective analytics in crowded situations, and the potential there, because now you’re introducing another sense in human terms into the whole big data conversation, and the enhanced context that could give you around what you’re doing with all those boxes is amazing.

But how are you going to shift that video around? If it works well it’s going to be working well at a low resolution, because it’s going to be more easily adoptable, but it’s still bigger than a text file.

Stuart: I think it’s almost the opposite. Big data would solve that problem, because it would solve data getting big  So a lot of the stuff we’re talking about here is how you move to these micro transactions, and location-based services. 

That’s not big data right, it’s very small amounts of data, but the problem is that we’re storing it, so we’re doing post-processing, as soon as we turn it into real-time, you will find that data has a value over time, and you will put a value on it, and you’ll either use it or you’ll get rid of it, but you won’t keep storing it, because a lot of the stuff we’re talking about has only value for a very small period of time.  A lot of the people I speak to around the region aren’t going to store all the data. 

They just can’t afford to do that, even though disk is cheap, it’s still not effective for them to store it, and so it’s really saying, well do I capture everyone who came to my website, or do I only capture the  people who made a transaction? There’s information you can build around both of those, but it’s just working out which has actually got value to you.

Chris: It may go beyond just extracting or filtering the data. It may go to the point where you really need to look at local processing to extract some intelligence from that data, so what’s the meaning of the information you were talking about?

What might you do instead of transmitting video to a central place and having processing there? You build a network camera that understands that’s a person, that’s a briefcase, the five points of facial recognition, you transmit the face.

Tiberio: Just to really subscribe to your point, it’s much cheaper to move computation around than to move data around, that’s something we’ve learned with the internet, again. So the internet is really a good teacher for us. 

Previous PageNext Page
1 2 3 4 Single page
Got a news tip for our journalists? Share it with us anonymously here.
Copyright © nextmedia Pty Ltd. All rights reserved.
Tags:

Log in

Email:
Password:
  |  Forgot your password?
By using our site you accept that we use and share cookies and similar technologies to perform analytics and provide content and ads tailored to your interests. By continuing to use our site, you consent to this. Please see our Cookie Policy for more information.