Data science and machine learning algorithms are transforming the big data community. The growth in big data is well known across all industries and business functions, in particular telecoms.
However, one of the biggest complaints from operators is that they are drowning in customer data and struggling to effectively use the insights to improve day-to-day operations and their supply chains.
Unlike general FMCG supply chains, operators’ supply chains have rich datasets about customer preferences and behaviours. This means there is a huge opportunity for operators to use data science to gather, aggregate, store and analyse these trillions of bytes of customer likes and dislikes.
According to IDC, improved customer experience and customer service are ranked as top business priorities in Australia. Telecoms operators are under mounting pressure to improve their data analytics expertise and processes and actually use these insights to deliver a wireless supply chain that enhances loyalty and provides the truly tailored brand experience customers are demanding.
Operators are a step ahead of other retailers or FMCG companies because they have such close relationships with their customers. They have access to some very powerful insights that can be used to offer a personalised customer service approach. Improved technology makes it possible to collect, retain and analyse data that otherwise would have been discarded. In addition, new advancements in data science allow professionals to use more sophisticated techniques to integrate big data to a level unseen before.
“The Harvard Business Review claimed that data scientist will be the sexiest job of the 21st century.”
Statistical analysis and data mining last year was the number one skill that got people hired, according to LinkedIn. The Harvard Business Review claimed that data scientist will be the sexiest job of the 21st century. A new approach to business management and data analysis is being seen across businesses, with a growing emphasis on hypotheses, algorithms, cause and effect, and experiments to extract knowledge from large data volumes.
These data mining techniques are used by a new breed of data scientist to interpret rich data, investigate problems and provide exact solutions. Data science is used by many retailers, for example, to pinpoint what customers want, how they buy and what they might be interested in buying in the future. These skills are in such high demand that data analysts with the business know-how are earning almost three times Australia’s average salary. To capitalise on this opportunity, most Australian universities are now offering graduate degrees in data science and analytics.
Another hot topic is automation and using algorithms to classify, predict and optimise customer interactions. For example, if an algorithm is being used at a retailer to predict which stores are going to run out of a popular item, it could automatically send a signal to the supply chain to push more stock to those locations. Doing this manually for all stores and products would be very time-consuming. This is fuelling much debate about the best way of combining automation and human judgement to create optimum results. For operators automating small-scale decisions based on structured data can be a way of improving costs, quality and timeliness – and delivering a great customer experience.
Communicating all of this information clearly and efficiently is another discipline where we’re seeing a big growth. This involves building compelling visual representations of complex data using tables, charts and diagrams to make big data more accessible to non-mathematicians and non-scientists. Well-crafted data visualisation helps uncover trends, develop insights and explore scenarios. Underpinning all of this is story-telling – working across teams to uncover the story behind all of this data and communicating it in an engaging and simple way.
From a business perspective data science is an integral part of analytics that encompasses data mining and business intelligence. But what do these emerging trends and science mean for the wireless supply chain?
All of this data and scientific analysis is geared to achieving the Holy Grail: a seamlessly positive customer experience. Allowing customers to interact across multiple channels is an expectation that must be met by retailers. Better knowledge of competitor pricing, demand trends and customer buying preferences (online and in-store) can initiate sales and promotions that help avoid losing business and retain customers. This also impacts customer service and enables operators to provide a more tailored interaction, while improving supply-chain efficiency and creating a ‘smart supply chain’. Using analytics tools, such as cloud-based platforms, enables a real-time optimised experience, which is crucial to achieving this.
With such close relationships with customers, operators can use rich transactional data to accurately predict supply and demand and ensure optimum supply-chain efficiency. Supply chains need to move away from being product forecast driven to become customer demand driven. For example, when a new smartphone is launched, businesses can use data on who is using a similar product, what stage they are at in their contract, how much data they use, what accessories they’ve bought, and other preferences to predict peaks of demand and ensure adequate supply. So, ordering the right product, sending it to the right place, at the right time and with the right price point will help improve speed, accuracy and scalability of order fulfilment.
The essence of demand shaping is knowing about your most profitable customers and products, and protecting and promoting them. Consumers will not be disappointed by an “out of stock” notice and retailers won’t have excess product piled up in their store rooms, so it’s win-win for everyone.
Data scientists can also help with product design and ensure a portfolio meets customer requirements. For example, a retail phone store in a small country town is likely to stock different products to one in a central business district location because customer demand will differ – features like enhanced network connectivity and being ruggedised may be more important in a rural or remote setting than colour variety. Knowing the trends and predicting behaviours based on insights means the right phones, accessories, bundles and services are all available across all channels.
It’s also important to use data science algorithms in customer segmentation and clustering to determine which customers are the lowest cost to serve and which are likely to buy the highest profit products. Creating a balanced menu of bundled offers based on data to address customer needs across brand, price, function, accessories and value-added services (eg insurance and upgrade programmes) is key to success.
Most operators have their own analytics departments, working across customer relationship management (CRM), marketing, business intelligence and reporting. Ensure they are feeding into the right people on the supply chain, so the company is working together end-to-end. Use data and modelling on the customer and how they are using products to tailor offerings. By working with the relevant teams, insight can be used to drive the supply chain and ensure customers are at the heart of everything the business does. The current crop of big data and analytics tools provides a way to integrate data from sales, marketing, customer actions, product reviews, competitor information, warranty data and supplier status in real-time to make demand-driven supply chains a reality.
Big data is transforming all industries and business functions, and data science is the next wave of innovation set to hit the telecoms industry. Unlike the general FMCG supply chain, operators’ supply chains have rich datasets about customer preferences and behaviours. Tapping into this is a big opportunity, but operators need to ensure they are using the right technologies and platforms to manage and drive it into the supply chain to improve the customer experience.
Gregory Hill is head of business analytics at Brightstar Australia and a member of the Industry Advisory Board at Melbourne Business School’s Centre for Business Analytics. Brightstar, a SoftBank Group subsidiary, claims to be the world’s largest specialised wireless distributor.
The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.