Getting the most value possible from the data you have gathered is something every business should aim to do, but the high-end tools traditionally associated with enterprise-level business intelligence (BI) and analytics have always been out of the reach of smaller companies -too expensive, too specialised or too complicated to be worth applying.
However things are changing in this regard and the same tools are increasingly to be found at more affordable levels, requiring less specialist training and sometimes even bundled with SME-targeted business systems.
Meanwhile, Gartner says it expects global business intelligence software revenue to reach €10.3 billion this year, a 7% increase on 2012. So with all these tools at the disposal of companies who previously probably could not afford them, just how should they be used and what is the bare minimum any business should be doing?
Immediate answers
"People who work in the technology sector are used to being able to get immediate answers from their data in their personal lives through search engines and so on, and also getting answers on the types of devices that are most convenient for them," said Microsoft’s Ronan Geraghty.
"Many of these people are decision makers and when they walk into their work environment, the idea is that they need to be provided with tools that work on familiar devices, using familiar interfaces that give them the answers they need to support the decisions they need to make. That’s the fundamental premise of analytics – to enable decision makers to be nimble in their decision making and to make sure the relevant data and answers are available when and where those decisions are being made."
According to Mike Baker, business intelligence lead for IBM Ireland, while the tools are becoming more accessible, there is a lot more to a successful BI project than just having the right software.
"Business intelligence tools are becoming ubiquitous-I don’t think you can buy a financial system or a basic ERP or ordering system without what the vendor will call BI or analytics built in. That is creating an awareness that is making people think that they need to get a strategic view of how this technology is working," he said.
"However most of these bundled tools are very focused on delivering BI and analytics from that specific data set – they’re not really set up to give the user a total overview approach. But you need that to really get value for this kind of system."
Baker believes that there are two key components of any basic business intelligence or analytics project.
Key components
"The first is people, because getting good analytics people is hard. They’re not 10 a penny and it’s hard to find people who can really deliver value, bring it to life and embed it in the day-to-day operations of the company. You need people who understand what you do in your business, how your data was collected, what it means and how it can be plugged back into your on-going needs," he said.
"The second thing is you have to understand what you want. The tools are out there and the skills are out there, but ultimately they’re not worth much if there isn’t a strategic understanding from the end users of what they’re hoping to achieve. Otherwise you end up in the land of operational reports that get scanned and basically ignored because they’re not delivering the information needed to be relevant to the decision making needs of the business."
Without these two issues squared away, Baker suggests that it doesn’t matter what tools you have or what data those tools are working with, you’re unlikely to come up with results that will drive competitive advantage, at least not in the long run.
"The tools are out there, but regardless of where they come from, you need those other two pieces of the puzzle. There’s nothing new in that – people have been talking about people, product and process for as long as I can remember-but if you don’t have this overview, this focus on the endgame and what you’re trying to achieve, then it’s a waste of time," Baker said. "Without this, what you’ll find is that you’ll gain some short term wins, but then that will dwindle off," he said.
Problem ID
Edel Lynch, analytics centre lead for Accenture, agrees. "The thing to do is to start by asking what the problem you’re trying to solve-what is the specific business issue that you are seeking to address? How would better leveraging of your data lead to better results against that issue? That can be a very far reaching question," she said.
"For many companies, a good place to start asking this question is in the area of fraud and compliance. How do you better tackle fraud, how do you better target your audit and investigation resources to make sure you’re auditing the cases where there is a high probability of fraud? Then look at the CRM side of things-why are customers leaving? Where should you focus your customer retention efforts? Without that kind of specific focus, there is a risk that the money you spend on business intelligence and analytics projects doesn’t get you the right outcomes. It’s this last mile that brings the insights to the decision makers."
According to Lynch, research recently carried out by Accenture in the UK and US among 600 executives showed that while adoption rates for business intelligence and analytics are increasing, surprisingly few of those companies that have spent money on the technology are happy with the value they are getting.
Satisfaction
"Those results resonate with what we’re seeing in the market here in Ireland. There is a real sense that companies are starting to see the value in analytics and want to adopt it, but it’s not a clear cut picture. Around one third of companies surveyed said they were using it aggressively, and that number is around three times more than we saw back in 2009 when we did similar research, so it’s a strong trend," she said.
"What we’re seeing in general is that there is a lot of regard for analytics at senior levels, with two thirds of those companies surveyed appointing chief data officers roles in the last few years. It’s being taken seriously at board level, but when we asked about using analytics in decision making, the trend wasn’t so positive."
Only about 39% of companies said that the data generated by their business intelligence systems was relevant to business strategy, while just 22% said they were very satisfied with the outcomes the received from their analytics investment.
"That really shows that while analytics has strong value associated with it, it’s still very much a new kid on the block and companies are still struggling to get return on investment. The proliferation of some of the cheaper tools shows that companies can’t just buy the tools, plug them in and expect to get value," said Lynch.
"It doesn’t work that way – you have to embed analytics into your DNA and culture to get the most out of it. It’s that last bit that is the key challenge for companies."
The start
Another challenge for firms thinking of dipping their toe in the water in this area is knowing where to start. But according to Microsoft’s Ronan Geraghty, the average company is already sitting on large amounts of data that they could be making more of.
"Any organisation accumulates data as part of its day to day activities. It can be data about sales, about relationships with customers, about external or publicly-available data-anything really. An easy way for people to get started is with something like Excel 2013, which has features that allow you to drill down and mash up different sources of data to create your own analytical applications and reports within that package," he said.
"For example, you could point Excel at an internal data base and also suck in data from a third party data source, maybe market information drawn online from the Azure marketplace. Then you could match your data against this third party data to see how you’re doing from a business perspective against the market."
This is self-service business intelligence, something quite different from the traditional approach to analytics. Traditionally, this kind of work is carried out by a team tasked with producing reports, which are in turn given to decision makers in the business. But increasingly, this kind of data is being put into the hands of the decision makers directly.
"You can use something like Powerlead to put together a very intuitive dashboard which you can drill down into to map out how your sales are doing across a particular period, using lots of different visualisation techniques within that to see how a particular trend is unfolding over time," said Geraghty.
Democratisation
"A bubble chart, for example, can give you a dynamic graph where the ‘x’ axis might be time, and the ‘y’ axis might be cumulative sales of, say, cars. As you work your way through a calendar year, you could see how car sales go up or down, depending on the time of year. That information in turn can be used to help you decide when marketing activities need to be ramped up to make sure your products and brand are foremost in people’s thoughts at particular times when you know they’re more likely to spend."
Microsoft’s position on the democratisation of business intelligence tools is that it wants to make the adoption of analytics as straightforward as possible.
"We want to break down the barriers to adoption. Depending on the size of the organisation, you may have a team in the IT department that’s responsible for deploying and managing a BI system, but we want to make it as easy as possible for that solution to deliver value as quickly as possible," said Geraghty.
One of the reasons why business intelligence projects tend not to deliver the value expected of them is very simple-the data the projects are based on is not up to scratch. At least that’s the opinion of Ergo’s chief technology officer, Mark Bate.
"We’re seeing an increase in companies realising that for the first time, the toolsets they need are within reach. They don’t cost as much as they used to, and they don’t require months of roll out training and complicated support capability," he said.
Quality reports
"But where they’re going wrong is that they don’t really understand that the quality of the reports generated by these tools is only as good as the data used to create them. It’s quite common that companies just assume that the quality of the data they are basing their BI on is good when it’s not. Depending on the age of the company and the evolution of its technology, you’ll often find that the data being used is formatted in multiple different ways, is duplicated needlessly, is incomplete or is otherwise compromised."
Bate suggests that you can take data and put it into a data warehouse, create a cube and make very impressive looking dashboards and visualisations for it, but if it is poor quality data then the picture produced is not accurate.
"This really becomes an issue when data is acquired that starts out quite separate and ends up pooled together-something that is increasingly happening as mergers and acquisitions in the tech sector sees many small companies being bought up by larger competitors," he said.
"This creates an integration challenge as the overall owner seeks to apply business intelligence or analytics tools to it. We’ve come across cases of companies attempting to integrate data from five or ten different sources to create one data warehouse. That’s not an easy thing to do."
Ltd
