CIO Strategies: Five best practice tips for implementing business intelligence

Getting business intelligence right is about much more than just building a data warehouse.
Written by Mark Samuels, Contributor

Martin Draper, technology director at luxury retailer Liberty, says his big priority for 2017 is business intelligence. The firm gathers information from various sources and Draper wants to create a data-driven strategy that helps executives at the firm make smarter decisions quickly and effectively.

He joined the world-famous retailer in March last year, having completed a 13-year IT leadership stint at Harrods. Throughout his career, Draper has been eager to drive transformation through information. He spent a decade running his own business intelligence (BI) consultancy before becoming a retail CIO, helping organisations in all sectors turn data into insight.

Here, Draper draws on his experience and provides five best practice tips for running a successful BI imitative.

1. View your initiative as a pervasive cultural approach

Analysts are keen to talk up the continued investment in BI and analytics. Gartner says 48 percent of companies invested in big data in 2016, up three percent from 2015. IDC, meanwhile, expects global spending on big data and business analytics to grow from $130.1 billion in 2016 to more than $203 billion in 2020.

Yet Draper says technology is just one element of a strong BI approach. He says success in BI is about much more than the sum of tools and process. Too many CIOs still fail to recognise that the cultural element is essential. "I've made a pretty good living out of technologists building data platforms and then seeing that the realisation of the benefit hasn't come through," he says, referring to his prior experience as a BI consultant.

"You've got to have the right technology and a clear data architecture roadmap, but you've also got to have an organisation that can consume information and develop insight. That's absolutely the number one tip: if you don't view your BI project as a pervasive cultural approach to business, you might as well not bother starting."

2. Gain top-level backing for your long-term goals

Draper says CIOs must recognise that an analytics initiative is about much more than achieving short-term goals. "Business intelligence is forever," he says. "If you're going to do it correctly, you have to understand all aspects of your organisation and plan a roadmap for implementation for as far as you can see."

CIOs, says Draper, should view their BI project as a programme, rather than a one-off hit. "It can take years to create a BI programme in some cases," he says. The protracted -- and, at times, costly -- nature of an analytics initiative means top-level support is essential.

"Make sure the number one person is your organisation is the key sponsor for your BI programme," says Draper. "The idea that you're going to use the data to tell the truth and to make decisions means support from the chief executive is absolutely key."

3. Use specialist capability to develop early wins for the business

Whilst Draper believes BI should be viewed as a "forever programme", he also recognises that quick wins can help demonstrate the benefits of your ideas to doubters, particularly in the early days of an initiative. These early gains can help dictate future direction.

"Don't predict the weather before you get on the boat. The beauty of today's capability and technology is that experienced individuals, either internally or externally, can help your business to deliver some benefits from BI pretty quickly," says Draper.

"You must always think in terms of long-term aims but you can also get short-term success quickly, if you create an effective balance between technology, IT capability, commercial talent and data architecture. Recognise that the world is a set of business processes -- our job in technology is to automate as many of those processes as we can and to leave the creative elements to the human experts."

4. Hook back into organisational demand on an ongoing basis

Draper says CIOs must define what success in terms of BI looks like. For some people, success is about business reporting. For others, it might involve self-service, advanced analytic capabilities, or a single version of the truth.

"The reality is that, depending on the audience, it can be all of those elements or just one," says Draper. "Realising that BI is an enabler to better decision making, and defining what it needs to deliver to enable that insight, is absolutely critical."

Executive expectations around the potential power of analytics remain high. Draper says business requirements can flex and change, so CIOs must ensure their BI programmes match technical capability with commercial awareness. "You will have to hook back into the business on a continual basis -- and forever," says Draper.

"BI is not just about building a SQL Server data warehouse. Great BI is all about creating benefits at some definable point down the line. Otherwise, it's just another platform -- and your business will not see the value from its investment."

5. Aim for a nuanced understanding of the information you provide

Draper warns executives not to focus solely on the visual representation of data. True insight, he says, comes from a more detailed level of understanding.

"Unless your organisation can look at the dashboard, and understand how to turn the red indicator to orange and onto green, then the gains will be small," he says. "The new focus on big data and real-time is potentially brilliant. But if your organisation can't react to the insight, then the information is pointless."

Draper says he was taught by finance chiefs early in his career that the conversation around data should always focus on objectives. "Great BI is about understanding what you and your colleagues are going to do differently tomorrow because of the insight you have today," he says.

"Unless you can decide and act, all analytics provides is a number on the screen. So, defining what success and execution looks like is key. It's very easy to produce numbers and opinions; its's very difficult to execute."

Read more about business intelligence

Editorial standards