The five pillars of building a business case for analytics
Today analytics has become the top priority for most enterprises as they strive to find avenues for revenue growth and operational efficiency. It is one of the most noted trends in the industry. However, with exponential growth of data, enterprises are struggling to make significant progress. There is an urgent need to identify opportunity costs and clearly quantify benefits which are measurable. Business challenge: 1. A general lack of understanding of analytics Various analytic applications with different definitions include business, customer, marketing, multi-channel, web, social media, mobile and supply chain analytics 2. Understanding actionable analytics Enterprises understand the value of making data driven decisions, but struggle in measuring the ROI of their analytic programs. 3. The limited availability of analytic talent There is a shortage of specialized skill sets to fit into an analytics role. Gartner predicts that there will be 4.4million jobs globally through 2016, but only 1/3 of them will be filled up 4. Data is the catalyst for all analytic initiatives Common barriers to success are: Data availability Data quality Data integration Data volumes In this paper we highlight the five pillars of making a business case for analytics. These ensure that analytics is not an afterthought, but an integral part of any business or technology investment decision. Fig: 01 Business challenge Common definition Actionability Current issues Data Talent Fig: 02 Five pillars Align with & impact strategic business objectives Consider your actions in advance Assess your data maturity Leverage crossindustry practices Build a Partner strategy The framework helps develop a common language to adopt across the enterprise. It addresses important challenges and creates a roadmap to ensure that the business value promise is met. 03
1. Align analytic initiatives to the enterprise strategic business objectives This is crucial because it helps to focus analytic efforts and articulate the analytic initiatives on the enterprise in the areas of highest impact. The best place to start is to analyze the business roadmap and use analytics as a means to not only improve performance but also look for new areas of impact: 2. Have business foresight Before starting any analytic initiative, it is important to not only consider the actions that can be enabled but how the analysis will be implemented. End Goal What actions will the analysis enable Improve Current Performance How to measure results? How will the analysis be implemented? Measurement Execution Generate potential areas of investment By evaluating these aspects, a common base is created between the business and technology stakeholders, resulting in a CIO-CEO partnership. Let s take an example from the consumer banking industry. A key digital e-enablement initiative aims to enhance responsiveness to customers and improve collaboration across functional stakeholders, while significantly reducing operational costs. Alignment of analytics to the strategic objectives include: Defining the program to some relevant questions like: Who will adopt the new features Which aspects of operations will e-enablement really help What kind of inbound call volumes will reduce Any other relevant business questions Measuring and validating the ROI of the initiative It is important to know when the initiative has succeeded and how one can apportion returns among various parallel initiatives Evolving the roadmap One should consider how to decide the future investments into the digital customer experience and when to start thinking about the operational dependencies. Improve performance of new investments What actions can be taken post-analysis A strategic business initiative is often executed (and refined) over multiple phases. For the analytics to be actionable and quantifiable, the program should evolve. Example of an evolved program is one that aims to target the right customers with the right offer, through the right channel at the right time. In such a case, the power to take action on the audience, offer, channel and timing should exist. If that is not possible, then efforts should be directed towards impacting the audience and channel. For testing out a business hypothesis, all possible business decisions should be explored. How to implement the analysis Setting up the analysis process, building a predictive model to score inline as a database and how to score data which will vary from the development data set, are the vital factors to be kept in mind. This knowledge helps in analyzing and developing the predictive models. For example, to build a model for predicting the likelihood of purchasing a particular product, one must integrate and make available the following data sources: Purchase behavior Web browsing data Demographic data Social media data This will help in preventing, investments develop a model that cannot be optimally leveraged. 04
How to measure results Determining the effectiveness or ROI of the analysis requires collection, organization and storage of the right data. Planning in advance ensures that the required data is available to measure and share results. The key metrics needed for decision making are building the right processes and tools for quick analysis. In the initial stages of maturity, it helps to maintain flexibility in reporting. Being flexible ensures that the learnings arising from changes can be applied towards building a user-friendly infrastructure. 3. Get to know the data The next step in building the analytics business case is to assess the current data landscape and create a target state. Key points to consider are: What data is available: Different analytic initiatives require different types of data. An inventory of the data categories that support the initiative is helpful. What data is needed: Due to an ever-changing business environment, one may not have all the data available. An inventory of data categories and elements is critical to avoid re-engineering processes every time, new analysis is required. Is the data in the right form: The data should be in an accessible form to be consumed readily by the initiatives. How to introduce the required data: Data comes from various channels and mediums, such as: Online Mobile Through third-party partners Through consumer applications such as Facebook, Pinterest, among others Each of these sources will need different type of pipes to build, as they have varying latencies and have gaps in the quality of data, among other aspects. Evaluating these sources facilitates developing an integration roadmap. How can the data be integrated in the right form. Data can be stored in single and Multiple repositories. Existing platforms and processes have to be considered while building a data consolidation plan. It can either be in the form of a master data management or data governance roadmap. 4. Consider cross-industry best practices Each enterprise and industry is at a different place in the analytics maturity curve. Analytic maturity can be advanced by assimilating more sophisticated methods. The statistical methods and applications leveraged are easily transferable, despite variations in data and key performance indicators. As an example, regardless of a long history of leveraging advanced analytics for claims, fraud and risk modeling, the insurance industry is now focusing on customer analytics. This is an area where retailers are more advanced. Other examples of industries advanced in their analytic methods include: Telecom: Churn modeling and network analysis Financial services: Fraud and risk modeling Retail: Customer lifetime value analysis, cross-sell / upsell modeling, market basket analysis CPG and manufacturing: Demand planning, forecasting and merchandise analytics, supply chain analytics 5. Build a partner strategy The analytics maturity of an enterprise determines what type of partnerships work best for them. This enables them to leverage partners to help: Develop an analytics road map: Build a plan that aligns with corporate goals while elevating analytic maturity. Build new analytic capabilities: Mine data, develop frameworks or develop predictive models as outlined in the analytics road map. Use cutting edge tools one lacks expertise in: By evaluating different tools to select the one that best meets requirements. Integrate new data sources: Leveraging the partner s big data and data processing expertise to link data silos or new data sources. Examples for data include structured, unstructured and the third-party demographic or research data. Summary A business case for analytics is crucial for a CIO. Embarking on the journey itself fulfills the CIO-CEO partnership gap. The first step is to define the enterprise s business objectives and make analytics an important part of all initiatives starting with the conceptualization stage. This will ensure happy shareholders. 05
About the author Melanie Murphy Senior Director Analytics & Information Management (AIM) Melanie Murphy is the Senior Director of Analytics within Mindtree s Analytics & Information Management practice. Melanie has 18 years of IT experience with the last 14 years focused in analytical and customer marketing applications. She currently leads strategic consulting engagements, guiding clients through the customer marketing landscape, including customer engagement and database marketing. She assists clients in determining ways to monetize their information assets through enhancements to data strategy, analytic processes or product offerings. 06