2nd September 2017
This article was originally published in the LinkedIn. Read the full version here.
The prospect of using Data Analytics in determining decisions in organizations is intriguing and almost inevitable as it offers unimaginable applications across industries. But only a mere 5% of these organizations are capturing the benefits as the Big Data trend advances…
The concept smart city has been defined as: “Communities that are building an infrastructure to continuously improve the collection, aggregation, and use of data to improve the life of their residents – by harnessing the growing data revolution, low-cost sensors, and research collaborations, and doing so securely to protect safety and privacy.”
A smart city should be a habitat that places technology at the core of its development strategy by harnessing the power of big data and digital technology. Information of the current conditions of a city across all areas such as weather & traffic, communicates information to other devices, control centers or servers, energy, telecommunications, transportation, water and waste water, health & human services, public safety, smart payments & finance etc.
With the internet connectivity as an underlying enabler, and utilizing the information technology software, hardware and systems, a smart city will be capable of providing an integrated solution that links up different aspect in people’s life and enhance the overall quality of life. There is basically no limitation on the extent that Internet of Things (IoT) can link with, amongst others includes transportation, logistics, safety, healthcare, education, resources consumption, and Government administration.
Besides providing a “smarter” living solutions, the application of IoT also leads to the formation of Big Data. Big Data is the results of large amount of data being generated from multiple sources. It has variety of potential uses to address problems directly from the sources. Big Data also allows analytics for deeper insights through data analytics, data intelligence and data mining.
This will improve the accessibility of information of a city for constructive usage such as city planning, safety enhancement, demand & supply forecast.
A successful smart city is able to acquire input of information via sensors, social media, apps and various digital platforms. These massive amount of information will be collated and stored in a data house for further analysis.
Big data analysis, or the mining of large data sets to identify trends and patterns in consumption, consumer behaviors, and resource availability can provide insightful information that could help in making major decisions in development, town planning, commercial related issues.
The big data collected on citizens, enterprise and government can be put into predictive analysis to create actionable information. For example, it can be used to improve customer experience satisfaction based on deeper customer understanding, predicting the issues based on extensive analysis before it happen and creating revenue projection based on identified sales opportunities.
Data centers are also as critical to monetize the big data as it serves the purpose of storing the information. A data center stores computer systems, and its associated components commonly including power supplies, data communication connections, environmental controls and security devices.
Big data and analytics for Businesses
Data is now the most important asset of a company. Many companies believe that proper use of data will propel organizations to exert higher capabilities and values. It is absolutely possible to generate great value and huge growth from data sets in every industry, from banking, finance, to healthcare, logistic, city planning and more others.
Better data leads to better decision making and an improved way to strategize for organizations regardless of their size, geography, market share and customer segmentation. It is critical to analyse data by unravelling the various patterns and revealing hidden connections within the vast sea of data.
Data analytics is a tool to analyse massive and varied data sets into some useful information for decision making. Every industry uses data analytics to help them make wise decision and ultimately to achieve their targets, typically cost minimisation and revenue maximisation which leads to maximum profitability.
There is a four-step process in data analytics and these covers Problem Statement, Data Collection, Data Analysis and Decision Making which uses the raw information collected from smart devices through various channels and convert them into actionable information that will assist in business decision making process.
Case Study: Public Transportation Business
For example in a public transportation business, the operator uses data available to discover the best routes to operate, the best suited time to run the vehicles, the most efficient asset management, leading to cost efficiency and profit maximisation. Hence, it is important to understand what database does the company maintain and whether the data is organised in structured or unstructured manner.
Let’s say for example the transportation company is experiencing poor routes performance. First step would be to state the problem clearly for the management team to further investigate the root cause of loss making routes and then, take necessary actions to turn the situation around. The second step is then to look into the company’s database and understand thoroughly what information have been captured relating to the routes operation and other management information.
The company has a real-time e-ticketing database which records the routes operated by each transport mode at its specific timing, the total distance travelled by each vehicle, the ticket sales online, number of passengers for every trip, the number of trips arranged for each route and many more. By understanding the data on hand, we could start identifying the key parameters to be used for analysis and listing all other information that are not available.
As we know, the profitability equation is made up of two components, and that is REVENUE and COST. The database captures most of the revenue component but cost component is not in the picture. The missing data needs to be collected intelligently from other channels. In this case, the finance department and operational team would be a good starting point.
Several interview sessions can be carried out with operational team to understand their operational model. The operational team members are the fieldworkers that can best provide accurate on the ground information on the operational data. However, the data provided should be validated by the finance team to access the reasonableness of the operational team understanding and actual expenses incurred in the books.
After collecting all the data required, next comes the most important step – Data Analysis. It is a critical step to put all the data together, restructuring the data and extracting the meaningful information that will aid in making the decision(s). To access the profitability of the routes, we need to list down all the revenue streams and costs components of the routes.
It is important to cross check the reasonableness of the revenue and cost components to the finance data so that the analysis does not miss out any major revenue and costs components. A deeper analysis can be carried out to further understand why certain routes are profit making and some are otherwise. There are many factors that may affect the routes performance, and these could be operational costs that are too high, sales volume that is too low, or simply the ticket could be well under-priced. Before any decisions are made, understanding the root cause for each profitable and loss making route is important in order to take the necessary measure(s) to improve the situation.
Should the operational costs be too high, it might be a signal for inefficiency in the modal operations or the management of the deployment strategy as a whole. This will lead to another round of data analysis, now emphasising on “Profitability per vehicle”, which will then require the data to be restructured to obtain the revenue generated and costs incurred for each vehicle. From here, we will be able to identify whether the high costs incurred is due to the old age of the vehicles which inevitably incur higher fuel consumption or it could be where poor management in serving the routes.
Low ticket sales volume could be caused by low demand for the route designed or that vehicle trips are scheduled during the timing where it is low in people traffic. Sometimes tight competition among operators which are operating the same route can aggravate low ticket sales too. Surveys can be carried out to gather opinions from the passengers and strangers to understand the favourable timing of routes, which route is in demand and preference for vehicle choice. Approaching the ticket counter staffs would be a wise choice to understand the underlying ticket sales factors such as staff’s attitude, ticket counter attractiveness and marketing strategy at the counter.
In most circumstances of transformation, data analysis is not sufficient in identifying the real root cause for change. It always comes with a series of investigations from different stakeholders i.e. the management, the fieldwork personnel, and most importantly from the customers. Subsequent to identifying the root cause, the management team can now make the decisions required to improve the situation. It can range from improving the deployment management strategy, changing the trip schedule timing, adding adequate vehicles on popular and high demanding route, closing down low demand and loss making routes or finding new potential routes.
To realize the value of data, an organization needs strategic solutions surrounding the aspects of data analysis. In our strategy development approach, we also fortify our data analysis with a unique hypothesis based modules that identify clearly the problem(s) that need to be resolved. What we have found to be crucial is to answer the core question on the table, ensure that the data analysis is accurate embracing the entire range of competitive drivers. The dataset must then be supported by a set of assertions for if all are true, then the hypothesis is correct and should be mutually exclusive and collectively exhaustive. Monitoring of the strategic implementation plan via action brief is the final step so that the outcomes are quantified and tested.
