CRN roundtable: the lowdown on big data

And so capturing all this data as well as capturing or sourcing additional data, over and above what you’re generating, you’ve got a lot of data to deal with obviously, and you need the tools and techniques to actually utilise those data to get effective outcomes.

For example projecting how much power might be used or be required next week, or the week after, because of the high temperature situation or something like that.

This client, from accessing all this data, interrogating it, and developing the reports and everything, was able to actually make vast amounts of money by consuming off the grid, rather than producing power. So with being able to access all this ‘big data’ and having the appropriate analytical tools to actually get that information and provide the appropriate reports, the ROI on that project was absolutely phenomenal.

We are actually seeing a lot of that, but it’s really just about the data exhaust that we’re finding. Storage is cheap now and capturing that data, keeping it, you might not have a business use for it right now, but having that historical capability to then start to use a lot of these emerging tools on the market to actually interrogate that data against what you already have, is actually going to give you a lot of business benefits in the future. That’s what we are finding has been the most profound change.

Some of our customers are capturing some 100, 200, 500 gigabytes of data a day from machinery that they’ve never actually captured data from before in the past. That’s a lot of data, and so you need very clever tools to be able to actually interrogate that to get what you need.  So the tools are the next step.

Scott: The interesting question for me is what gave them the impetus or insight to actually start looking at the data tools?

Robert: Large power stations they have sensors and transducers, scatter software currently is very expensive to deploy, but scatter software is more of a control function to control scatter devices rather than to actually monitor and provide analytics. So capturing that data and storing it, gave the capability to actually look and see what is actually going on.

Scott: They were capturing it, but what made them think to start looking at it, when previously it was regarded as exhaust or not actually that valuable or low value density, and they weren’t really looking at it, what made them start? Are they the first ones that have done it that you’re aware of?

Robert: With this particular product yes. Splunk is the technology that was actually utilised in this case. And it was more about stumbling across the information that actually changed their business. There wasn’t a business directive involved in the project. 

They just wanted to capture the data, so they have lots of analytical people within their organisation.  After playing with it, they stumbled across a very important piece of information that business was missing out on, and the outcome of that was very positive.

Scott: It would be interesting to see whether you could control some of that stumbling shall we say.  That seems to be the thing, people stumble across these insights, and that drives them to invest further – and then they start to wonder ‘what if’, but the tricky thing is how do you get them to deliberately, intelligently, architecturally start to stumble?

It’s interesting when you talk about big time, load stuff and do it yourself. I think one of the issues that’s happened in the past is that end users have relied on tools like Excel to make their own interpretation, whereas in the old days, they needed to get these big reports out of the data centres, and they sent each one out, and now they say ‘well I can do something in Excel’.

But Excel has a limited view of the data, and it would be good for them to say ‘can’t do that anymore, but here is something else’. But they still want to be able to do it themselves.

Peter: Yes, why should I change the way that I get this information and why should I change the way I solve problems.

Scott: Within the next couple of years, 25 percent of marketing managers will have transferred to digital, so that will be at the expense of some of the media. Now if they’re doing that, they’re going to want to make sure they’re getting a return on investment, and so that will be an impetus for how can I get more insights.

Chris: There are some examples of that, I think some of the simpler things with Google Analytics for your website -  it sounds trivial, but it’s actually a non-trivial problem if you’re a big company and you’ve got a very large number of hits on your website.

You can get some very good insights out of that and it’s not in your own shop, but I think it still comes back to this issue of the guy who runs this business is the one who wants to be able to make the decisions, particularly if he has to pick the strategy.

Peter: Marketing people are going to be more accountable because you can in fact measure it. If someone says ‘okay I’ll put that over here, I want a lot of people seeing it’, you couldn’t tell whether they could or not, whereas you know whether they’ve looked at the page, you know they’ve flicked through and there’s a lot more information that you can analyse. So there’s a large accountability here that wasn’t there before.

Scott: Yes, it also makes the brand more accountable, right, because you have a more immediate link with your consumer.

Jonathan: I think you’re seeing the industry emerge, if you look at the SEO in the industry which reverse engineers the ‘big data’ to provide results and then analyse that to put key words then as an example. Look at online shopping or some of the online gaming industry which is very sophisticated online. You just need to follow where the money is, and dare I say it’s probably in the porn.

I’m sure it’s in the online gaming, where people are releasing new games very quickly, looking at the ‘swarming’ effects of who’s actually playing the games and working up ROIs, and shutting down those that don’t meet that criteria. So I think there’s a lot of insight.

I think your question is really interesting around how to control the stumbling. It’s really back to business basics, which is you’ve kind of got revenue and costs, and you understand you’ve got a cost issue or a revenue opportunity, you can then go and explore and say ‘maybe I should go and explore around here’ and ‘maybe I can get the right people with the right tools, because the tools are there.

John: One of the interesting things is because you’ve got the digital capability, you also have a lot of digital knowledge, so people have been hired to go through and do the click throughs, and you’ve got to have intelligent systems that can get beyond the knowledge, to know is your site really popular, or is it someone telling you it’s popular. There’s a lot of that stuff starting to happen more and more and you won’t really be able to understand it unless you’ve got the right tools to be able to handle it.

CRN: So do we feel that humans have essentially become stupid in this world?

Tiberio: No, humans are essential. You have to turn the power on. I agree with Peter, basically it’s a lot about the intelligence of the tools. I agree that the tools are there, but the tools we’ll see in five years time will look very different.

It’s essential that we keep on inventing the tools, simply because there’s lots of value in improving your tools and the system, and so whether there’s an improvement of tools or I guess there’s a lot of energy put on improving the tools themselves – so there are the two ends of the spectrum so to speak, improving the tools and actually how do you actually make the tools useable by the business. Which is this point about helping the customer.

John: The other thing about stumbling, with each of the market segments what we’ve got to do is find the first movers and help them get a competitive advantage, and as soon as they get a competitive advantage, the others will move pretty quickly. 

Scott: Everyone is playing a waiting game of when do we get into this, we know we need to invest, but we’re just not sure as to where that value’s going to lie, but then the risk is if you wait too long, then you will suffer from others using this to their advantage and you’re the client. 

I think the classic example is that little app that Amazon offers on your phone, which gives you the capability to walk out of the store, because they’ll offer you that product. 

So you use their app to do a search on the product, and they will guarantee you 5 percent cheaper or whatever it is if you walk out of the store and buy from them. Now what are they doing with that information? That’s big data in real-time.

Jonathan: With big data the insight is slightly obscure because you’re trying to find it, or stumble across it. On the other hand, and this gets back to my ROI question, there is data or information sitting in organisations that people aren’t going to access there, and it’s not obscure, it’s there, there’s IP that’s just not in a format and it’s so obvious.

So I think my view is that again one needs to understand where your assets are in the organisation, because we roll out a whole lot of enterprise content management applications in Oracle, where the IP is sitting in the organisation and people can’t even access that, so never mind ‘big data’ where you’re not even sure where the IP might be, I think that technologists do get excited, and I think especially in some large organisations like utilities, telcos, banks, they offer value because there’s so much data and so many transactions, the return is multiplied so much. 

I think certainly as an ROI, there is sometimes a lot easier places to start with just collecting the internal assets and putting them in a repository and getting a brain across that to identify that.

CRN: Laura your company DataFlex has put a lot of computer power into government agencies over the year. We know from research that governments have been fairly receptive to the cloud concept, but are you finding in the government sector that they are stumbling with regard to big data as well?

Laura: Well I’m rather enjoying this concept of stumbling, because I think we are an enabler of stumbling. As an IT reseller, we are the conduit between the emerging technologies and the business requirements of these organisations.

Certainly, talking about Federal Government, all the things we’ve discussed today, even though we’ve spent years putting in place infrastructure storage, storage has become cheaper, they’ve got the data, now it’s the paradigm between us as an IT reseller responding to the requirements of a client spanning hardware and software and becoming a thought leader.

Helping them and assisting them to address what they’re trying to achieve with these major data silos in data and getting benefits and worth out of the data they already have on premise. 

So it’s interesting the discussion around the so-called dirty words like ‘cloud’ and ‘big data’.

Peter you mentioned ‘tablets’ as among the ‘dirty words’ for you. One of ours is probably ‘mobility’.

It’s an opportunity for an organisation like DataFlex, because we want to help organisations, especially in the Federal Government space, to go from what exists now, to what they could potentially be doing in the future, and I like this term ‘stumbling’, because that is it, we are the conduits from one phase to another.