Starting with Traditional Tools for Real-time Applications

Get the WebProNews Newsletter:

[ Business]

Have you heard about:

  • The real-time enterprise?
  • The agile enterprise?
  • The zero latency enterprise?

I’m willing to bet you’ve run into at least one of these terms lately in your reading. Research analysts, industry pundits, and software vendors have latched onto these phrases as ways to promote a simple concept with a profound impact on business results:

Enterprises that are able to most quickly and effectively act on the information they receive achieve competitive success.

There is a lot of power in that sentence. And quite a few challenges, from both a business perspective and a technological one. The idea of acting “quickly” means that automated systems are necessary to assist in correlating incoming information with a set of predefined policies and decision strategies so that a response can be made when needed. Acting “effectively” means that companies must be able to evaluate the success of their policies and strategies and continually refine them to improve business performance. And, the concept that “enterprises” must accomplish this indicates that actions must be consistent across all operations and not be restricted to individual employees, systems, locations, or communication channels.

What’s Real about Real Time?
You are most likely to find “real-time response” mentioned in connection with CRM. Quick and effective decisions are critical when relating to your customer base. In a research note entitled Real-Time Analytics Techniques to Support CRM, Gartner Research(1) identified six key components of real-time analytics:

  • Real-Time Modeling
  • Dynamic Scoring
  • Real-Time Scoring
  • Real-Time Business Prioritization
  • Real-Time Decisioning
  • Real-Time Offering

Modeling involves an analysis of past transaction history and business results in order to find patterns that can be applied to future activity. For instance, we may have extended an offer to the public to sign up for a new type of service. An analysis of results may show us that young, single females had the highest response rate; that young, single males did not show much interest; and that middle-aged, married persons bought higher fee/higher profit services. This allows us to build a model for making future offers that focus our efforts on the most cost-effective segments of the population.

Although modeling has been used in this way for many years, it is only recently that hardware and software performance has increased to the point where real-time modeling is possible. This allows companies to make fine adjustments to models on a continuous basis to account for analysis of the latest data, to incorporate new business guidelines, and to compare strategies based on different criteria (for instance, is our goal to increase the total number of customers or the average profit per customer?).

Scoring is another traditional business tool that has seen great improvements in power and flexibility with the latest automated systems. An outgrowth of modeling, scoring systems can use the same historical data analysis combined with human expertise to construct decision trees that execute in-process with a transaction system to determine an appropriate action at runtime. Advanced scoring systems can report significant factors in the customer’s score calculation that influence the action decision. Dynamic scoring systems can integrate data from the current transaction with other historical data to modify the recommended action to account for external factors.

Once a basic strategy is put in place, adjustments must be made on an ongoing basis to incorporate business goals, requirements, and constraints. Businesses often have conflicting goals at the corporate and department level, while resource limitations and executive policy all play a part in determining actions that should be chosen. Sometimes referred to as “strategy optimization,” tools available in this area should have the ability to put limits on variables included in the strategy determination (outbound phone calls cannot exceed 1000 per day), minimum requirements (program must result in a minimum of 50 new customers per day), and competing goals (maximize number of orders taken, but minimize returns and bad debt).

Advanced capabilities for optimization software include the ability to simulate and compare the results of strategies assuming different projected response rates and uncertain business conditions. For instance, if we assume that customer responses may vary between 2 percent and 12 percent on a new promotional offer, what is the optimal strategy to pursue that has the best overall composite performance? As more data comes in, uncertainty ranges can be reduced and strategies refined.

Decision Logic
Real-time decision execution incorporates predefined models, scoring trees, transaction data, corporate policies, and business expertise to arrive at a consistent methodology for picking actions to take in any given scenario. Often implemented with business rules technology, decision-making systems allow explicit management and control of the logic used to perform specific business functions. A real-time decisioning system should separate the “what to do” logic from the “how to do it” mechanics in a business task. The rules for making a decision can then be applied consistently across multiple physical systems or communication channels (for example, a customer will be presented with the same offer whether she visits a Web site, gets a promotional insert in her monthly bill, or calls the customer service center).

The use of “rule engine” software and dedicated rules management systems can offer enterprises the ability to make changes to their business logic at any time, as dictated by changing business conditions, development of new models, or introduction of new business policies. These specialized software packages remove the need to recompile applications and interrupt production operations in order to make a change to decision logic. All decision making is performed outside of the functional application code that controls interfaces, reporting, input/output, and the like. The separation of functionality makes rule-based processing well suited to service oriented architectures, including the use of inter- or intra-company web services.

Once the applicable decision-making steps have been processed, the organization is ready to apply the most appropriate offer in real-time to the customer interaction. One of the important benefits to organizations is that the decision-making is separated from the mechanics of how the offer is presented so that the customer has a consistent experience and can view the organization as a single entity, rather than as a collection of distinct business functions, each with its own view of the customer.

A Real-Time Customer Reaction
Companies looking to increase their agility and expertise in responding to customers in real-time should not be frightened by the number of components available to help with different tasks. Even though each of the technologies mentioned above can cooperate and contribute to an overall real-time enterprise decisioning process, each can offer important benefits on its own.

Companies should start by identifying business functions that could give them the highest return if improved, and then concentrate on building results in that one area. As additional tasks and technologies are added to the real-time enterprise stack, the organization is likely to recognize incremental benefits from the integration of functions and data used by the different processes. This makes it important to include technology selection criteria that consider the expandability and interoperation of models, scoring, optimization, and rules so that future development work can benefit from their integration.

By applying real-time decision-making technology to CRM business processes, organizations can reduce latency between receiving information and taking effective actions based on it. They benefit from a continuous refinement of business decisions and strategies to stay current with business conditions and new data, and they ensure that the entire organization is working in a consistent manner to achieve prioritized business goals.

(1)Real-time Analytics Techniques to Support CRM, G. Herschel, Gartner, Inc. Research Note published on April 24, 2002. Decision Framework, DF-15-9474

*Originally published at CRMguru.com

Click here to sign up for free B2B / Tech. newsletters from iEntry!

With a background in software development and marketing, Ken has been producing and delivering business webinars since 1999. His background in public speaking, radio, stage acting, and training has given him a unique perspective on what it takes to create a compelling and effective presentation. Currently Ken offers consulting services through his company Webinar Success (www.wsuccess.com).

Starting with Traditional Tools for Real-time Applications
Comments Off on Starting with Traditional Tools for Real-time Applications
About Ken Molay
With a background in software development and marketing, Ken has been producing and delivering business webinars since 1999. His background in public speaking, radio, stage acting, and training has given him a unique perspective on what it takes to create a compelling and effective presentation. Currently Ken offers consulting services through his company Webinar Success (www.wsuccess.com). WebProNews Writer
Top Rated White Papers and Resources

Comments are closed.

  • Join for Access to Our Exclusive Web Tools
  • Sidebar Top
  • Sidebar Middle
  • Sign Up For The Free Newsletter
  • Sidebar Bottom