Emerging into the light
Spelunking might be the act of exploring a cave, but a company with a similar name may be a torch for BI users, as Nick Booth discovers
If the performance of business intelligence (BI) systems over the past decade were analysed, using its own brutally frank tools, what sort of revelations could we expect? Would these BI insights be embarrassing for the CIO who bought them and the channel partner whose advice they trusted?
Almost certainly yes, if the feedback from analysts at Gartner, Ovum and The451 is accurate. Until very recently, a drill down into the data on the performance of BI made grim reading for anyone who had recommended it.
The figures for returns on investment would make a risible chart. The stock of the CIO would be at an all-time low, while the disappointment of the board could be represented graphically only by one of those upturned hockey-stick graphs that represent rapid growth.
As recently as January 2013 Gartner research director Svetlana Sicular said that the BI market had fallen into a "trough of disillusionment".
According to the Gartner analyst, BI was launched in a fanfare of hype and, once the excitement died down, a period of depressing backlash followed. Previous technologies that have gone into similar nosedives, such as the applications services market, Token Ring and ATM , never recovered.
Which mean they dragged the resellers - who had devoted themselves to this technology - down with them.
The good news is that BI is emerging from its problems and is about to embark on a period of advancement and maturity. A stretch of rapid expansion is expected and, if history is anything to go by, that will lead to a boom which is tempered by chronic skill shortages. Those skill shortages can only be good for the channel, because it offers them the chance to fill the gap with managed services.
The surge in growth of BI was created when two of its biggest limitations were removed. The raw power for processing data has come on in leaps and bounds, thanks to the logical data warehouses created by Hadoop's marshalling of cheaper processing power and Hana's in-memory processing of Petabytes of data.
At the same time it is the new-user friendliness of BI that will create a massive boom. The launch in September 2013 of Splunk 6 - which enables anyone who can use an Excel pivot table to launch complicated queries of big data - was hailed by some analysts as the BI sector's own "Windows moment". An easy-to-use interface could popularise this technology to mass market status.
Splunk's biggest achievement is that it can sit on top of other systems and bring all the information together more easily, according to Forrester analyst James Staten (pictured, left).
"No single business can create an accurate picture of its customers or its business without collaborating on the data," says Staten. Splunk has created the conditions for what Staten calls adaptive intelligence.
Where once BI was the sole domain of the data scientist, these days the people interrogating the customer information, the machine-to-machine conversations and the security data are likely to be experts in their field. A marketing executive, for example, will be able to run their own queries about customer behaviour.
So will a security manager be able to take a much more proactive, hands-on role in managing intruders, because they will now have a much easier way of setting up queries?
"BI and analytics remain a top-priority technology for CIOs," says Ovum analyst Fredrik Tunval. "They will help companies boost revenue, improve customer service and control costs, because employees can make better decisions, faster."
Or at least they would, if only someone could set up the systems and get the users started on using them. In 2014, BI skills will be a rare commodity.
The platforms are more affordable and usable, but companies may be forced to train, poach and even outsource the job of using them because the power, affordability and usability of BI systems will suddenly awaken this sleeping giant - according to Tunvall.
So there is an opportunity for channel partners to set up the systems for their clients. On the other hand, seeing how companies are going to struggle to find staff, they could go one further and run them as a managed service. Part of the fun will be revealing to clients information they didn't know about their own organisation.
BI to assess the cloud
Outsourcing giant Wipro, for example, uses business intelligence tools to analyse how cloud applications really work. This is a new growth area, because most people don't realise that the service-level agreement (SLA) information they get from the cloud service provider doesn't always tell them the full story.
BI tools could reveal much about the real levels of service in cloud computing, says Shashin Rastogi, Wipro's business service management practice head.
Amazon, for example, doesn't give you the level of detail you need about how your applications are performing when you use Amazon Web Service (AWS). It does give detailed information, but it's the wrong sort. While a customer needs to know how well their app works, they often get SLA data about how well the CPU performed, how much storage was used and the bandwidth figures.
That information describes how platform-as-a-service (Paas) offerings are working, but what the customer is actually buying is SaaS.
Customers really need a service provider who can tell them what they need to know - which relates to how the devices on the network are functioning, the operating systems and the interconnection between various pieces of coding.
As a big data expert, that is the service Wipro offers. It is also the service many resellers could be offering clients, as a trusted counsel on all matters concerning cloud and service.
Like Wipro, for example, a reseller could be collating machine data and using analytics platforms such as Splunk to make sense of all the other elements on the network.
Together, they create a picture of how well every tiny interaction performed - such as what the web server said to the database server, and how long it took them - and from each of those pixels of information they can build a high-resolution picture of how the applications performed.
"It's often a very different picture from the story told by AWS's data. That's because you are comparing PaaS with SaaS," says Ratogi. "That is not Amazon's fault but the client's, because it didn't do the due diligence."
Splunk 6 is easy enough to use to entice more people into the BI market but, like Windows, will still create massive support, training and guidance opportunities for the channel.
Some early adopters of Splunk have made a good living growing with it. "It was like being the first one at work to know how to use Google," says Kevin Tunsley, operations director for service provider Eqalis (pictured, right) "We've grown on the back of Splunk."
The irony is that, having suffered a backlash, BI is now becoming popular again. All it needs is a few exemplary case studies, and people will start to realise its potential. "Once people in a company see what it can do, someone else in a different department will want you to carry out a project for them," says Tunsley.