Modern day data science and analytics serve as one of the best decision-support tools for top management by creating predictive models that help management focus on ‘precise targets’. Precision is the key to competitive edge. People are diverse, and their needs are best catered specifically. Marketers need to customize their messages and promotions for target audiences; management wants to pick the best candidates for jobs; public authorities need to prioritize services to the most needed. With limited resources, precise targeting is the most cost-effective and profitable way of doing business. This is where data and analytics play a crucial role. Several world best companies have acquired their current positions through strategic business analytics.
Enterprises are increasingly harnessing the power of data and information technology to enhance their competitiveness. Successful ones make day-to-day decisions based on powerful analytics which also, gradually transform their organizations into ones of fact-based culture. Less advanced organizations develop analytical capability in certain functions, while novices are making use of small-scale metrics. However, as data storage and application software are becoming cheaper, organizations large and small can all have better analytics and enjoy the same agility – if they have smart management and right talents.
The major impediment for a novice to become matured in this area is the false belief that it requires a hefty investment in technology and that it is the job of their IT personnel to carry out the task. In fact, most metrics can be crafted with existing resources on condition that business managers or agency heads have an explicit demand for fact-based support in decision making. Advanced analytics do not require much investment in hardware, but do require the right kind of people and work processes. That is why Andrew McAfee and Erik Brynjolfsson said in Harvard Business Review (October 2012) that Big Data will revolutionize management.
Business Intelligence and Analytical Maturity
The use of technology and methodology to transform raw data into meaningful and useful information for business purposes is not new. Gartner, a leading American information technology research and advisory firm, coined the term “Business Intelligence” (BI) in 1996, and later defined it as ‘an umbrella term that includes the applications, infrastructure and tools, and best practices that enable access to and analysis of information to improve and optimize decisions and performance’. To move towards this direction, any enterprise can set a maturity framework as a roadmap and set a level of maturity as its business milestone. Advanced firms usually move to predictive analytics based on broad-based data to model scenarios and support strategic decisions while less sophisticated users may rely on a few metrics to evaluate past outputs and behaviors. Ultimately, as Thomas Davenport and Jeanne Harris[i] suggest, firms can efficiently compete on analytics with enterprise-wide infrastructure and having analytics embedded into their core business processes.
Achievement in BI and analytics requires less of advanced technology than right people plus conducive environment in organizations. Technological firm SAS lists 9 items, only one of which concerns technology, that determine a firm’s analytical maturity. The remaining items are strategy, management support, culture, data management, analytical skills, execution, and reporting/monitoring.
SAS uses the 9 elements to evaluate firm’s level of maturity using the five-level model proposed by Davenport and Harris (1997). Starting with the “analytical impaired” level, Davenport and Harris cite the breadth of data usage, coverage of infrastructure and management involvement as key determinants of the maturity progress towards the highest level of “analytical competitor”. These five levels are:
- Analytical impaired – has some data, has buy-in, fact-based decision making.
- Localized analytics – has management support/ pilot analytics implementation
- Analytical aspirations – has executive sponsorship, isolated analytics
- Analytical company – enterprise-wide capability with centralized infrastructure
- Analytical competitor – analytics are central theme, embedded into the core business processes
In 2013 data science experts, Provost and Fawcett, simply ranks data maturity into 3 levels of Ad hoc, Medium, and High-End, in terms of an organization’s capability in using data analytic thinking. Those who are in the High end have well-trained data scientists and participative stakeholders and have a capability to improve their data science process, not only to find solutions.
Capgemini, another technology consultant, defines maturity in its own term which comprises 5 elements, namely, vision & strategy, enablers, competency, deployment and governance. These 5 dimensions constitute 4 levels of maturity:
Level 1 Impromptu: sporadic and isolated analytic capability
Level 2 Solo: amateur or professional
Level 3 Ensemble: analytic initiatives across business functions
Level 4: Symphony: enterprise-wide initiatives.
However one may classify and label them, indispensable ingredients for advanced analytics can be summarized into 1) the data, 2) the data scientist(s), and 3) the executor(s). Data nowadays cover numbers, texts, images and moving pictures in digital formats. Data also requires appropriate applications and storage capacity. Data scientists are the ones who synthesize those diverse data into meaningful information. An ideal data scientist is the one who has a basic understanding of the core business, plus some knowledge of statistics and programming. In an amateur enterprise, this can applies to a team of one business talent who are somewhat familiar with data science and quantitative analysis, plus a programmer who is familiar with the business process. Executors can be the management, managers, or a team of pioneers in the organization who know how to make use of the information in an effective way.
Moving from square one to a more advanced stage may begin from either the data, the data scientist(s), or the executor(s). Availability of data may prompt somebody to do something with the data; or an enthusiast may try to develop a fact-based decision support system and squeezes out data from otherwise piles of unused materials; or a visionary leader may see the potential of analytics in value creation and leads the organization to adopt the approach. One initiative can lead to another. An enthusiast may drive the availability of quality data; well-craft reports based on these data may spur growth in data-driven projects; and management supports may enhance the data scientist’s chance of success. One movement towards the right direction may start a series of further value-adding actions down the road.
The Maturity Process
To sustain the momentum and progress towards maturity in business analytics, an enterprise needs the whole 3 elements to mature together. Data gathering, storage and processing must become more systematic and more conclusive. Complexity of data requires a more sophisticated synthesizer and a professional data scientist might be inevitable. The momentum will be sustained only when insightful metrics and analytics is embedded into the normal process of decision making. During budding stages, symbolic management support is adequate to get the ball rolling. But during the advancing period, it is the real usage that vitalizes the program. Quality analytics must yield a tangible competitive advantage or be terminated. The question is whether they will be allowed to do so or hindered by inflexibility in the organization.
Quality metrics will shine light on what and where in the organization to improve; and quality predictive analytics will identify what can be optimized by which way and how much better off the organization will be by doing so. Gradually, this insightful information will marginalize some existing jobs/roles while highlight some others. If well-crafted, these metrics will also replace some of management’s intuition and rationales where fact-based business analytics yield more reliable and insightful answers.
When such time arrives, the role of management decisions will be ‘lifted’ to focus more on larger, enterprise-wide strategic issues. Positive as it may sound, instead of feeling relieved, managers and executives may find the change too challenging. They, too, will need help and roadmaps in transforming their roles.
When all factors: the data, the data scientist(s), and the executors, are matured together, the enterprise will gain an unprecedented efficiency and competitiveness. Management will be relieved from routines to focus on more and more strategic decisions.
The change mentioned above, however, will not occur accidentally.
[i] In Competing on Analytics: The New Science of Winning, 1997.