Predictive Analytic Tools to Gain

¶ … Predictive Analytic Tools to Gain Market Niche

Predictive analytics defined

The first step in reviewing the literature is that of assessing the meaning of predictive analytics. The academic community presents the reader with a multitude of definitions and explanations of predictive analytics, but at a most simplistic level, this refers to the analysis of the past and the present with the intent of predicting the future. Nathan Conz for instance argues that predictive analytics tools "allow organizations to leverage their massive depositories of data - historical and otherwise - to make better decisions going forward" (Conz, 2008).

Don Murphy (2002) on the other hand approaches the definition of predictive analysis as a comparison with descriptive analytics. Descriptive analytics refers to the ability to observe and understand all elements and actions present with a setting. Predictive analysis refers to the usage of specific tools in order to transform the observations gathered into an ability to predict the future behavior of the assessed setting.

John Goff (2004) also approaches the issue of consumer analytics and considers the topic one of the hottest developments in the recent years with respect to customer relationship management. The definition he offers of consumer analytics is simple and it states that the concept refers to "tools that dissect consumer-buying patterns, suss out preferences, and predict future behavior" (Goff, 2004).

John Zyskowski (2010) takes a more specific approach to predictive analytics and assesses it in light of technological advancements. Predictive analytics is presented as one of the top strategic tools and technologies of 2010 and Zyskowski argues that predictive analytics represents a new spin on data mining. This view is supported by others as well, who state that predictive analytics is in fact a combination of previous technologies and tools, which are now implemented at superior levels and they as such generate superior results. But in spite of the novelty or lack of novelty behind predictive analytics, fact remains that their implementation is expected to generate additional value and benefits within the communities.

Another belief of Zyskowski is that the new procedure is gaining popularity alike in the business community as well as within the technological community. Additionally, the author mentions that, due to the benefits it generates, predictive analytics would become an increasing presence within the federal field. At this level, predictive analytics could be used to improve the operations of fraud detection, waste management and apprehension of criminals.

Regardless of the source of information and of the form of presentation, the understanding of predictive analytics is the same in all instances -- the usage of customer information to predict their behavior. And the popularity of the practice has increased dramatically over the past years due to elements such as technological advancements, customer demands or higher levels of competition across industries. Ultimately, this popularity of predictive analytics is due to the ever changing nature of the environment and the consumers, and the constant need of economic agents to understand their clients. From this angle then, customer analytics emerges as an imperfect and ever changing domain (Goff, 2004).

Predictive analytic in the modern day business community

Thomas H. Davenport, Jeanne G. Harris and Robert Morison (2010) argue that predictive analytics has three distinct applications within the organizational context. First of all, it can be used to better understand the company customers; secondly, it can be used to improve the performances of the organizational operations and third, it can be used to support and improve the decision making processes. Today, more and more firms are beginning to implement predictive analytics as a means of increasing their chances of reaching the organizational goals.

But aside from the importance of analytics within the organizational context, the three authors mention that a broad approach and implementation is not sufficient. In this order of ideas, they argue that in order for the economic agent to become more successful and more powerful from the competitive standpoint, it has to select a specific domain and implement predictive analytics to a high extent at the level of this domain.

Davenport, Harris and Morison (2010) reveal several instances in which predictive analytics can be used and capitalized on within the business community. In this order of ideas, they argue that:

Within the financial services sector, predictive analytics can improve operations of credit scoring, detection of fraud, pricing strategies, or customer profitability

Within the retail industry, predictive analytics can enhance promotional efforts, shelf management operations, forecasting of products and/or service demand, inventory management and so on Within the manufacturing industry, predictive analytics can improve supply chain operations, inventory management, demand estimation, customization of products, the development of the items and so on Within transportation, predictive analytics can support operations of scheduling, routing or yield management

Within the healthcare industry, predictive analytics can improve operations related to drug interaction, management of diseases or early on diagnosis. As a parenthesis, a more in depth analysis of predictive analytics within the healthcare industry is conducted by Judith Lamont. Her findings indicate that the usage of predictive analytics within the sector is more and more popular and this is because of the benefits it generates, such as: advantages to the patient (improved medical services, earlier diagnosis, smarter health care and so on), reduced fraud in the medical system, creation and enforcement of more adequate rules, superior processing or centralized mapping (Lamont, 2010).

Within the hospitality industry, predictive analytics can support the development and implementation of pricing strategies or customer retention strategies

Within the energy industry, predictive analytics can support operations of trade, supply and demand estimations or legislative compliance

Within the communications industry, predictive analytics can improve pricing strategies, customer retention strategies, demand estimations, capacity analysis and planning, network optimization and so on Within the services industry, predictive analytics can improve operations related to customer service staffing, or the profitability of the organization

Within the federal sector, predictive analytics can improve operations of fraud detection, crime prevention or budget balancing

Within the virtual business community, predictive analytics can be used to improve the operations related to web metrics, website design or customer recommendations

Finally, within each and every one of the fields mentioned above, as well as within any other field, predictive analyses can successfully support increases in performance management (Davenport, Harris and Morison).

At the level of customer satisfaction, this is now the number one priority of economic agents. From a point in time at which customers would simply purchase whatever the companies would produce, the modern day clients have evolved to a stage at which they demand the producers what items to manufacture and sell. The progression of the customer role within the market has been marked by a series of economic, political or social modifications, but an important element is constituted by the globalization phenomenon. Through the opening of boundaries, firms increased their market penetration and this materialized in significantly higher levels of competition. Subsequently, a situation aroused in which organizational emphasis on customer satisfaction became pivotal. And this customer orientation was obvious even in the most unlikely industry sectors, such as the insurance industry.

In order to cope with the higher levels of modern day competition and incremental customer demands, economic entities developed and implemented a series of strategic courses of action. In support of the creation of the most adequate customer strategies, firms commenced to implement predictive analytics. Within the insurance industry for instance, predictive analytics was no longer implemented solely as a statistical tool by which insurance rates and damage compensations were estimated, but as a means of creating more value to the customer (Conz, 2008).

But the growing importance of predictive analytics tools within the organizational contexts has not only been due to increases in competition due to globalization, but also due to the advent of technology. In this order of ideas, customers were better able to gather product and service information and became as such more demanding. On the other hand however, the advent of technology also allowed companies to enhance the quality of the consumer experience, and this was possible -- among other things -- through the integration of predictive analytics. The direct result was that of analytic software, which emerged and became more and more popular within the totality of the business community. Software analytics is now present in manufacturing industries, in service sectors, in food market segments, consumer goods, financial markets and so on. They address aspects of inventory management, demand forecasting, estimation of labor force needs, pricing strategies and several other fields in which decision making is based on critical data. Peter Alpern (2010) argues that this situation was created as a result of evolution within the business community, or the result of increased business intelligence due to a higher access to information.

A mostly specific application of predictive analytics within the organizational context is observable at the level of the marketing efforts. The marketing operations draw tremendously on lists, databases and customer information. And predictive analytics has emerged to help marketing staffs better organize, assess and learn form customer information. As Christopher Hosford (2009) best put it, "to marketers, lists are still important but the uses to which they are put seem to be in transition."

With the aid of predictive analytics, marketing activities make a transition from the simple identification and attraction of customers, to more complex endeavors of customer loyalty and retention, generation of sustainable revenues as well as other elements of marketing ROI (return on investment). This trend has had a direct impact on the secondary industry sector of list construction in the meaning of an increasing popularity and demand for these specific services. This in turn led to an increasing offer of listing services and the increase of competition within the field.

But given the complexities and requirements of the modern day business environment and social community, a necessity arises in the continuous assessment and upgrading of the databases and the information they contain. Examples of how the databases could be maintained and upgraded include the sustained gathering of new information, the more in depth segmentation of the consumers or the identification of current and emergent responsibilities (Hosford, 2009).

Predictive analytics and safety

Craig Lawson (2010) assesses the issue of predictive analytics from a rather intriguing standpoint -- the well-being of the equipment. He argues that through the use of predictive analytics, organizational leaders can identify errors and other problems with equipments early on. This would not only increase staff, operation and consumer safety, but would also reveal an ability to reduce maintenance costs by identifying and repairing damages early on, rather than waiting until the equipment problem is major and requires major investments to fix.

In order for predictive analytics to increase equipment safety and as such generate the adjacent benefits, they need to be supported by technologies -- both software as well as hardware. One such application is the SmartSignal Predictive Analytics software applications which is able to identify rod drop problems when other traditional applications cannot detect the problems that early on. When companies do not wish to invest tremendous sums of money into the development of their own predictive analytics tools and methodologies, they are presented with the opportunity of hiring third party firms which offer the respective services. Through extrapolation, this could constitute a specific case of penetration into a specific market of equipment safety.

Lawson's study is conducted within the pipeline industry, but the findings are relevant and can be extrapolated to include the entire manufacturing industry where increased usage of manufacturing equipments is made. Here, organizational leaders can predict problems and resolve them early on, without delays in operational processes or even without the incurrence of casualties. "This innovative technology allows you to migrate your maintenance strategy to a predictive, proactive strategy, where the morning water-cooler talk is about how to prevent an identified problem from happening, not how to fix one that already has occurred" (Lawson, 2010). Also as an extrapolation, predictive analytics can be used to penetrate the niche market of safety tools and equipments.

Predictive analytics to create customer value

A first application of predictive analytics is that it allows firms to better understand the customers. Traditionally, it was believed that customers could be stimulated solely through price reductions. The collection and processing of customer information with the use of predictive analytics however has revealed that not all customers are price sensitive, but that a large portion of them are driven by other elements.

Don Murphy (2002) argues that at the level of the marketing department, predictive analytics are presented as customer relationship management (CRM) analytics. Through them, companies collect customer data and then synthesize this data. The result is an increased ability to understand and predict customer behavior, which in turn can materialize in organizational success. Murphy explains: "With the right tools and a little intuition, organizations can increase profitability by leveraging customer information to anticipate customers' needs and influence their behavior. By consolidating customer information from within and beyond an enterprise, organizations establish the groundwork for modeling and predicting customer behavior. Quality information gathered from numerous customer-facing and third-party systems can help companies define and select customer populations for building predictive models. Companies can the use these models to segment and target customers for marketing campaigns, service programs and customer loyalty and churn analyses."

A relevant example is offered by the hospitality industry, where companies offer services rather than material products. This virtually means that the success of the respective entity -- hotel, entertainment facility and so on -- depends directly on its very ability to understand and serve its customer. These two goals are best achieved through intense communications with the customers and through the implementation of a wide array of strategies that generate and support customer loyalty.

In order to use predictive analytics to generate more customer value, Davenport, Harris and Morison (2010) have identified the necessity for the implementation of a six stage model. This model resembles a ladder of marketing predictive analytics and it is achieved as follows:

The construction and maintenance of an adequate customer database which includes information on all organizational customers

The segmentation of the market into target audiences based on a series of demographic and otherwise criteria. At this stage, it is important to understand the differences between various consumer categories in order to present them with different offers based on the elements which satisfy their specific needs.

The development of organizational strategies which create points of difference. At this stage, it is pivotal to respond specifically to the actions, decisions, needs and so on of the customers.

The management of the marketing campaign in an adequate means to promote the company, the product and to attract and retain the customers. At this level, it is necessary to keep a faithful track of all previous customer decisions in order to be able to personalize the product / service offering.

The creation of predictive models based on the observation of the past events. Throughout this stage, the company strives to foresee the future actions of the customer based on the observation of his past actions and decisions.

Finally, at the sixth and last stage, the company is expected to make a real time offer in the meaning that it is expected to create and offer a customized product / service for the customer (Davenport, Harris and Morison, 2010).

Don Murphy (2002) takes his analysis one step further and reveals the three specific steps companies need to undertake in order to predict the behavior of customers. In this order of ideas, at a primary level, it is required that the economic agents profile their customers. In order to create the profile, company members assess the behavior of their customers in specific circumstances -- especially those of interest to them. They use a multitude of sources of information, such as internal customer data previously collected, purchase information, call center information, performance indicators or other third party data.

At a secondary level, the company is modeling the customer data. "By using data mining on the profile information, analysts can uncover the most relevant characteristics of the customer segment being analyzed" (Murphy, 2002). Finally, at the third level, the economic agent engages in scoring operations. "Analysts use predictive analytics to score existing customers by comparing them to the model. Those most closely matching the characteristics included in the model are most likely to exhibit the targeted behavior" (Murphy, 2002).

While not referring specifically to predictive analytics, Alexei Sarnevitz (2008) mentions that a successful relationship with the customer is only achievable if both of two necessary conditions are met. The first of these conditions is that of segmentation, which has already been discussed and will not be focused on at this stage. The second condition is that of automation, which, in Sarnevitz's view, is the key to organizational success. And by the key to organizational success, he refers to the high degree of knowing the customer up to a point at which a long-term bond is created between the parties. "The ultimate goal is more than having the right product in stock at the right price; it's about tailoring the entire shopping experience to create an emotional bond with the customer. In effect, this means turning today's multi-channel retail enterprise -- in a consumer's eyes -- from "the store" to "my store." [And this goal is achieved through automation]. Automation is the key to enabling the delivery of just the right offer to each consumer, deploying exactly the right assortment to each store and optimizing the regular price for millions of SKUs. Examples include the latest marketing automation solutions that can target each consumer with relevant personalized messages or revenue optimization applications that determine optimal pricing for each item in every store" (Sarnevitz, 2008).

Predictive analytics to create customer value -- the case of the insurance industry

When predictive analytics was first implemented within the insurance industry as a means of creating customer value, its applications were limited and the immediate results were at least questionable in terms of efforts and outcomes. Today, predictive analytics remains important within the insurance industry due to its actuarial and underwriting applicability, but an obvious move is being observed towards its usage in the direction of customer value creation. In other words, more and more firms within the insurance industry are using predictive analytics to increase the levels of customer satisfaction and to as such consolidate their competitive positions. "If you look at most insurance companies, I think the areas that are most ripe for more science-driven thinking are distribution strategy, claims management, product and pricing, marketing, branding and consumer research, and customer experience," Buluswar [vice president, insight and innovation, at Los Angeles-based Farmers Insurance] explains. "It's become the competitive position," adds Jon Walheim, a senior executive in Accenture's insurance practice that is responsible for marketing, sales and service work in North America" (Conz, 2008).

Predictive analytics within the insurance industry were initially implemented by the Information Technology departments as a means of better managing data and increasing operational efficiency. As IT was implementing predictive analytics, the marketing departments were searching for an IT solution to improve its own operations. The two departments eventually formed a united from in the implementation of predictive analytics tools. As a result, predictive analytics became a strategic tool with applicability in several other organizational departments.

As it has been previously mentioned, predictive analyses create customer value by allowing the firm to better understand (and as such serve) its client base. In the case of the insurance industry, this is achieved through a superior comprehension of the following:

A more thorough analysis of the customer -- firm relationship

The technological infrastructure used by the firm and its levels of innovation to serve customer needs

The life cycle of the relationship with the customer

The strategies developed and implemented in the sense of customer loyalty and retention

The collection and analysis of customer data (including historic customer data) to be integrated in the improvement of the relationship with the customers

The promotion of the insurance products and services within the virtual community and the increasing levels of online sales (Conz, 2008).

Predictive analytics to gain niche market

The case of the insurance company is once again extremely relevant to assess the topic at hand. And the specific example in this case is offered by insurance company Unitrin Direct, which initially implemented predictive analytics as a strategic organizational tool. It eventually came to a point in which the predictive mechanisms were utilized in the efficient allocation of resources. At the level of the customer relationship, this materialized in a more efficient allocation of the financial resources to better reach the consumers and to better draw them towards the company's products and services. With the aid of better allocated resources, the company would be able to better penetrate niche markets.

A second -- and more revealing example -- is offered by the Brussels-based company Corona Direct. One of their products is that of funeral coverage, which is a difficult to sell, expensive to promote, not happy and niche product. This situation led to the necessity for a specific strategy to be created. In order to succeed in this niche market, Corona Direct developed and implemented a strategic approach based on predictive analytics. The main question was that of how could the company go about in increasing the response rate to the advertising mails without increasing its advertising budget. In order to answer the question, the company selected a set of five or six criteria to be analyzed, but the problem was that relevant answers could only be retrieved after the analysis of at least one hundred criteria. Given this necessity, Corona Direct partnered with predictive analytics firm SPSS.