In today’s digital age, we generate tremendous amounts of data. They come from our smartphones, our interaction and sharing on social media, product codes and sensors embedded in all kinds of devices. IDC predicts data to reach 40 zettabytes (that’s one with 21 zeroes after) by 2020.
US retail giant Walmart processes 2.5 petabytes (one with 15 zeroes) in transactions every hour. It can determine who bought what products using which bank’s credit card at what time. Such information and more, helps the corporation form its marketing strategies, offering promotions for certain items at certain times of the day, knowing full well that its customers are more likely to do so then.
Alibaba stores 100 petabytes of processed data and is capable of processing large numbers of transactions. On Singles’ Day in 2014, it processed 278 million orders. Its system can handle 5 million transactions per second. Google and Facebook can process one million photo images per second.
Great Predictive Value
Facebook and Google offer free services and apps in exchange for the data they collect which they can subsequently monetise. They can target online advertisements to a very specific demographic to get the best results.
A lot of data generated today are semi-structured or “unstructured” – not neatly filled into forms or spreadsheets. Advanced tools must be used to identify and classify incoming data and extract useful information. For example, Google can use a flight confirmation email and extract details of that flight to be inserted into the user’s calendar. It can also channel advertisements related to the user’s travel itinerary.
It was noted that Google can predict the outbreak of potential epidemic diseases earlier than governmental organisations. For example, Google Flu Trends observes Google searches and identifies 45 search items to predict flu spread
Facebook can predict a person’s personality traits with alarming accuracy. With only 10 “Likes,” Facebook knows a person better than his work colleague; with 70 Likes, better than his roommate; 150, his parents or sibling; 300, his spouse!
Big Data is characterised by large volumes of data from a variety of sources in multiple forms, created at incredible speeds. Big Data Analytics (BDA) has great predictive value.
Banks use BDA to detect potential misuse of credit and debit cards; to understand its customers and tailor-make products for them. Insurance companies may predict potential of a person’s sickness by checking on his use of credit cards to purchase alcohol.
The larger the dataset, the more powerful and useful BDA becomes for mining information. Traditional analytics tools cannot cope with Big Data. They cannot cope with unstructured data and will also buckle under the sheer amount of data.
It is commonly assumed that BDA is complicated and expensive, or it can only be deployed in a large corporation with a dedicated team of analysts. Certainly, the complexities of setting up such a system can be daunting and expensive. Yet, small and medium-sized business (SME) can certainly benefit from BDA as they also generate a tremendous amount of useful data.
They can introduce new products based on the information extracted. They can enter new markets and target their marketing campaign to specific demographics.
In supply chain management, BDA can ensure just the right amount of raw materials are purchased to produce just the right amount of inventory based on past records; in logistics, to optimise storage space and delivery; as well as to gain customer and business insight. They can determine which products sell best, who buys them and when – remember Walmart?
Sensors in machinery can collect data, which can be used to tweak performance and even predict when the equipment with fail, so that preventive maintenance work can be scheduled well in advance.
SMEs do not need to invest heavily in the infrastructure for BDA. They can outsource their data analytics to companies that do have the expertise. For example, Logistics Worldwide Express runs an integrated system that manages the entire logistics process. The company engaged its sister company, Unixus Solutions to develop the application.
Unixus in turn developed the system using Microsoft Azure cloud technologies. Thus, Logistics Worldwide Express only requires a PC or smartphone with an Internet connection to access the system. It did not have to commit to expensive hardware, maintenance and setup.
IBM’s Watson Analytics can be used to analyse a variety of data, from marketing to sales, finance, human resources and other operations. It can help answer questions such as what drives sales, which deals are likely to close, how to make employees happy, etc.
InsightSquared can connect to other services such as Salesforce, QuickBooks, ShoreTel Sky, Google Analytics, Zendesk and customer relationship management software, to perform pipeline forecasting, lead generation and tracking, profitability analysis and activity monitoring.
Canopy Labs analyses customer behaviour, sales trends and uses predictive behavioural models to make product recommendations. It can also show a customer’s standing – his lifetime value, loyalty and engagement level – and determine which customers are profitable and worth reaching out to.
The benefits of BDA are compelling. However, “culture is holding back the adoption of digital technologies,” lamented Baseer Siddiqui, senior research manager, IDC ASEAN.
In its journey towards becoming a high-income nation achieving greater economic progress, Malaysia needs to do more to unlock the full potential of its digital economy.
IDC predicts that by 2022, over 21% of Malaysia’s GDP will be digitalised. “Growth in every industry will be driven by digitally enhanced offerings, operations and relationships,” Baseer said. This means, SMEs that embrace digital technologies are likely to thrive compared to those that do not.
The adoption of emerging technologies such as artificial intelligence, robotics and internet of things has enabled businesses across industries in Malaysia to start realising the desired outcomes such as improved customer experience, better workforce efficiency and new business models.
However, the number of success stories in Malaysia are still small compared to other matured digital economies. According to Baseer to achieve a greater digital success at bigger scale, every stakeholder, including the government and IT vendors have to work in tandem with a comprehensive and actionable roadmap in to help end- user enterprises thrive and grow the country’s digital economy.
He also warned that businesses must continue to innovate themselves, as even “disruptors” can get disrupted if they become complacent. Citing how mobile phone leader Nokia was disrupted by smartphones from Apple, which was disrupted by Samsung, which in turn was disrupted by the likes of Xiaomi and Oppo, he emphasised that “companies must be willing to disrupt themselves” to stay relevant.
Technology has matured tremendously in the past five years. There are also numerous success stories and case studies on digitalisation today compared to 2014. There is no reason why SMEs should not adopt digital transformation.
However, he also advocated caution. Not every technology needs to be adopted. SMEs should carefully determine what technologies would be helpful to them and take it one step at a time.
Microsoft Malaysia’s director for commercial partners and SMC, Yeo Swee Key echoed the same sentiments. IT investment depends on numerous factors but “for most SMEs, purchasing actual datacentres and hardware just to get on the cloud may be detrimental to the longevity of the business,” he said. “Finding a trusted solutions provider would be a better first step. They have the hardware for you to outsource some of your work and thy have professionals who can offer advice and help.”
Baseer added, “IT companies should also be proactive and approach their clients with solutions, instead of waiting for the clients to come forward and tell them what they want. They should look at them as partners rather than clients.”