Big Data has been making headlines for a number of years now. There is a massive amount of information that is generated by our online actions, transactions and M2M communications. The question now is how organisations can leverage all of this information to boost business. Big Data analytics projects are the next step, but while some analysis may benefit one business, the same analysis could be worthless for another company. In the same vein, analytics can only paint a part of the picture and relying solely on Big Data analysis alone can be a risky move.
Big Data analytics projects are being implemented across a range of industries. Retail outlets are using Big Data to customise shopping experience by highlighting which consumers are likely to purchase which products. Companies are able to target those customers with personalised offers. In the banking industry, banks can use analysis to understand which customers are likely to default and pre-emptively put strategies in place. In production, industrial machines can be outfitted with sensors to use predictive analytics to determine which machines require maintenance and replacement before they fail.
There is a great deal of information to be gleaned from a Big Data analysis. Analysis can provide strategic value, with insights that can help business make faster, more efficient and pre-emptive decisions. “These insights can help enterprises gain a significant competitive edge, create market differentiation and aid customer satisfaction,” says Sadi Aweinat, Chief Technology Officer and Global Services Lead, Gulf and Pakistan, EMC.
Big Data analysis can also go a long way to improve a company’s overall efficiency. Insights garnered from a Big Data project can improve existing structures and processes within the organisation. With a clearer view of existing issues, business can address nagging problems, identify successful areas and focus on opportunities for growth and improvement. Leveraging Big Data insights to implement automation can minimise errors and redundancy and allow business to focus on what is most important.
As useful as a Big Data analysis project can be, businesses need to avoid the temptation of relying solely on numbers to have a clear view of their operations. What is most important in Big Data analysis is what the numbers actually mean – an analysis that only a human can determine. “Today, human intervention is still required in the final decision-making process on what to do next with the analysis. Companies tend to have benchmarks for performance, so it is usually pretty easy to catch skewed data or skewed analytics,” says Paul Devlin, Head of SAP Platforms, SAP.
Indeed, there are right and wrong approaches to a Big Data project. “It is tempting at the start of the Big Data analytics journey for companies to take on too much up-front,” says Raj Wanniappa, General Manager, Big Data, Dimension Data Middle East and Africa. “A common term used is ‘store all data’ for analysis later.” While this approach has been successful with huge enterprises like eBay and Nike Fuel Band, it is not financially viable for the average company to adopt similar practices. “The recommended approach would be to start with a platform – a set of analytics technologies and methodologies – and a limited amount of data sets and then expand from there once the business benefits have been confirmed.”
To begin a Big Data analysis, there are a few steps that need to be taken that are universal to all businesses, large or small. “First, business must define key metrics,” says Boby Joseph, Data Practice Head, Data Science Technologies. “Then they must gather the appropriate data, and iterate through the analytics quickly to find reliable, repeatable, and relatable patterns.” Business sectors need to be included throughout the process, and IT and business need to remain in constant communication. Also, analysts must be willing to adapt models as needed during the process.
Even done right, there is such a thing as ‘overindulging’ in analytics. “You can analyse anything and have irrelevant metrics,” says Megha Kumar, Research Manager, Software, IDC MEA, “The executive board and LoBs should look at what is relevant and create the analytics and dashboards accordingly. Also procuring a Big Data engine should be done with an end goal in mind or basic analysis is more than sufficient for organisations of a certain scale.”
Still, Big Data analysis projects certainly should be implemented. As to how the analysis is interpreted, that is up to the stakeholders of the business. “This becomes the personal choice of the business and business stakeholders,” says Andrew Calthorpe, CEO, Condo Protego. “Ultimately information is key, and most organisations would rather have excessive information and then decide on importance than have too little and risk missing out on important insights.”
Karthik Krishnamurthy, Vice President, Enterprise Information Management, Cognizant, agrees that overdoing analysis can be an issue . “It is quite possible to fall in the trap of simply overdoing it when Big Data and analytics initiatives are not clearly aligned to well-defined focus areas and business outcomes,” he says.
Analysis projects can also fall short when initiatives go to extreme lengths to prove technology capabilities that do not really matter to the business. Investments made for the sake of creativity without a defined purpose also result in a similar situation.
Indeed, it goes without saying that Big Data analytics can be very useful to businesses of all sizes. However, a little can go a long way in these situations. When taking on an analysis project, it is essential to ensure that all the proper metrics are decided upon beforehand and, perhaps more importantly, that limits are put in place as far as budget and time limitations. When the analysis is complete, ensure that all stakeholders are involved in the interpretation of the data.
At the end of the day, a properly implemented Big Data analysis can give valuable insight into a business’ processes. A sound analysis can identify strengths and weaknesses within a company and allow businesses to focus on the improvement of key areas and processes. With a clearer understanding of the status of the company, businesses stand to gain both in the short and long term.