Mar 28, 2015

What Aussie businesses are getting wrong about big data

Big Data has gained widespread popularity amongst Australia’s business community in the past two years. Despite this, Australian organisations still aren’t taking full advantage of the benefits that advanced analytics brings. A recent study into the top IT spenders in the country by research firm ITNewcom found that, while more than 50 per cent of organisations had adopted big data solutions, about 40 per cent were not even considering it.
Analyst firm Gartner reported recently that the end-user clamour for access to business data has manifested in ‘self-service’ business intelligence (BI) initiatives that have circumvented IT and as a result, are disposed to analytic sprawl. Research VP at Gartner, Doug Laney, said: “As a result of the limited governance of self-service BI implementations, we see few examples of those that are materially successful — other than in satisfying end-user urges for data access.”
So what are Australian organisations doing wrong?
Well, firstly, it’s not all bad news. Not all organisations in Australia have failed in their information management efforts. Indeed, we’ve worked with many over the past couple of years which have achieved some very good results, including some of the big banks, retailers and insurance companies. Also online businesses are reaping the benefits of their analytics and big data projects, marketing to the individual, rather than the many, with retailers suggesting what you might like to buy following a recent purchase you have made. Bricks and mortar businesses have been a bit slower to do this because they’re less data driven.
It’s in the large organisations and government departments where it gets complicated. Often working in silos, it can be very difficult to get one view across the organisation and whilst in the past the IT function might have been responsible for pulling data together for an analytics project, today information is coming from all different areas, HR, marketing, operations etc. For example social data isn’t handled by IT teams, marketing leads it. Similarly, operations teams have access to sensor networks, where train operators have sensors on their carriages to know where they are and distributors use RFID tags to track where their goods are in the supply chain. Both bypass IT.
The best way for large organisations to embrace big data or any information management project is to really think about the information you need and how that information will help you achieve the business outcomes you want. Below are a few tips about how to start:

1. Look at your business

Look at how much your business is driven, or should be driven, by information. For many industries, such as insurance, information is critical. For others, particularly those operating in a highly competitive environment, a small change in a product or service based on customer behaviour can have an enormous impact.
Looking honestly at your business at this stage frames the level of investment and the level of buy-in required at executive level.

2. Look at the answers you’re seeking

The most important element of any analytics program is defining the business objective. All analytics activity should be based on the fundamental question of what your business wants and how data can support that.
Prior to advanced analytics, we tended to base our business decisions on educated guesses – now we can close the feedback loop by looking at the impact of specific actions on customers, employees and assets.

3. Look at your stakeholders

Look at the internal stakeholder environment. There is always going to be resistance for any new approaches to how you do business. Analytics is a complex and relatively new undertaking, so it’s important to educate everyone in the business about the value of analytics.
The primary challenge of any analytics project is taking complexity out to provide a clear path to benefit realisation and gain wider business buy-in. Every successful analytics project relies on the green light from multiple areas and a comprehensive business change management program.
Don’t be afraid to enlist help from those who can complement your internal skill set by seeking external assistance where required – this can help ensure the project is being communicated and delivered effectively. Stakeholders will start to trust you early in the cycle if you give them insight they can trust and use.

4. Look at your data

Despite the common misconception coupled with the hype around big data, organisations don’t necessarily need vast tracts of data to benefit from advanced analytics. Outcomes and findings will not tend to change with a larger sample size, assuming you already have a significant sample size.
It’s often surprising just how much data is available to you; combining the finance system data with your sales, marketing, and operational data, sets a solid base upon which you can potentially add selected outside sources such as Australian Bureau of Statistics (ABS) demographics, weather data, etc.
Customer interaction data sources can complement your traditional customer data. Call (voice), video, web and text analytics tools offer new signals. These can monitor, decode and analyse how customers talk when they call your organisation, how they move around your physical store front or office, how they navigate your website and what they say when they write to you.
Leveraging these data sources represents an insight revolution for predicting human behaviour which is, after all, the most difficult thing to predict. This information facilitates a deeper understanding of how people behave at the most intricate level.

5. Look at the technology

Only once the previous four steps have been ticked do we recommend looking at the technology platforms available and the best options to suit the requirements of your project. Every IT project should run in line with a change management project to embed the business process changes in the organisation.

No comments: