WO2008075936A1 - Support of decision-making in a telecommunications business - Google Patents

Abstract

Intelligent Workflow (IW) computer technology according to principles of Decision Making and Execution (DME) includes a decision bus, integrated with demand/supply chains and with command/control functions, engaging all actors who make, manage, or execute decisions. Actors step through a flow of work, making optimised decisions using provided critical knowledge. The IW executes signed-off decisions, engaging relevant actors as necessary. Decision management includes monitoring the status of and impact on enterprise performance of any decision. Information related to products/services and to customers' and suppliers' dealings with the enterprise is captured by the IW and made available to actors.

Description

TITLE SUPPORT OF DECISION-MAKING IN A TELECOMMUNICATIONS BUSINESS

FIELD

The invention relates to software for carrying out a more rigorous process of decision-making within an enterprise than that which is customarily practised. The invention is based on a closer understanding of the process of decision-making.

DEFINITIONS OF TERMS

Actors, shown in Fig 1 as customers, associates and suppliers, are persons who are continually making decisions and executing them.

Churn is the effect whereby dissatisfied customers swap between suppliers.

CPDM; a set of customer knowledge held in an analytical data set (see Table 10) that explains why a customer makes purchase decisions is continuously analysed, stored and provided to the workflow at the right time to dramatically increase revenue generation.

DME (1); Decision Making and Execution. This is a decision discipline for use particularly in business, and is focused on achieving results. It is a quantified and catalogued method to remove extraneous or non-relevant factors from the pfrocess of making a decision and then executing it such that there is a high change of achieving the desired results. This discpline underlies the invention.

DME (2); A Decision Making and Management Intelligent Workflow Engine integrates the right knowledge at the right decision point with a workflow process for all decisions across the enterprise. Each decision is tracked and the performance impact of its execution stored in the KnowledgeSphere as learning for future use, thus forming part of the corporate memory.

EUAD; End user access device, e.g. set top box, mobile phone, PDA, fixed line phone or computer.

KnowledgeSphere,' A set of codified enterprise knowledge, relating to products and customers plus DME workflow intelligence is stored in a KnowledgeSphere. This is integrated with the DME workflow to support decision making, management and execution. This approach to the storage of knowledge removes the traditional wall between decision making and execution thus dramatically increasing demand and supply chain efficiency and effectiveness. Telco; A shorthand name for a telecommunications business such as (but without limitation) a regional or national service provider.

BACKGROUND

Business management practices and software have developed through a series of distinct evolutionary stages since the early 1960's. Today's state of the art as exemplified by ERP (Enterprise Resource Planning), and as typified by software from SAP and Oracle is described as being "process centric", i.e. the examples deliver business benefits by automating operational business processes.

A separate branch of business management practices and software that started in the mid 1990's called decision support was designed to assist decision makers by analyzing the data stored in their Enterprise databases. Today's state of the art in this field is exemplified by BI (Business Intelligence) and typified by software from SAS, Business Objects, and Micro strategy. These examples provide sophisticated statistical tools and utilities to support decision making at all levels within and across the Enterprise. Such business practices and software are described as being "data centric", i.e. they deliver business benefits by automating analytical business processes.

The advent of two global change drivers, the World Wide Web and Mobility, is placing increasing pressure on enterprises to integrate their operational and analytical capabilities in order to remain competitive, continue to increase revenues and decrease costs.

PROBLEM TO BE SOLVED

Many Business Enterprises have invested heavily in software, tools, and infrastructure that support their operational and analytical capabilities such as ERP and Data Warehouses. The Return on Investment (ROI) from these investments is frequently poor and below management expectations.

Three major root causes exist,

First; A business process is an artefact of a decision, i.e. the action steps in a process do not exist on their own. A decision must be made and then executed in order to complete the process. Thus at each decision point the relevant knowledge required to support its making and execution is required for the process to be effective. Process automation that is not guided by the correct contextual knowledge fails to deliver substantive benefits to the enterprise. Second; failure to treat decision making from initiation through to execution as a continuum, i.e. placement of an artificial wall between decision making and execution, failure to integrate analytical and operational systems and processes, failure to integrate «think the business» capabilities with «run the business» capabilities.

Two negative outcomes result from this:

The significant investment in decision making is lost, and risk to the enterprise is increased when the decision is not executed according to the intent of the decision maker/s

The Enterprise does not incorporate the learning gained from the impact of each decision into its corporate knowledgebase for future use, thus valuable learning is lost

Third; Enterprises do not have a detail knowledge of what motivates their customers to make product (where a service is generically classified as a product) purchase decisions. This decision is arguably the most important decision of all, for without a customer purchase decision the enterprise will fail to generate revenue and go out of business. CPDM as herein defined uses advanced models and algorithms to understand what motivates the individual customer to make a purchase decision, in short why the customer behaves the way they do. Using this knowledge the Enterprise is able to enter into unique customer dialogue with a customers at any time plus prepare products and services for the market that are highly differentiated from the competition. Both of these result in a significant increase in revenue generation.

OBJECT

It is an object of this invention to provide an improved system and method for intelligent workflow; particularly one based on DME; when used as a system for use in a telecommunications enterprise, or at least to provide the public with a useful choice.

STATEMENT OF INVENTION

In a first broad aspect the invention provides a system based on DME (Decision Making and Execution; a decision discipline as herin described) for use when making decisions in an enterprise such as a business, wherein the system (100) includes means to provide, when in use within digital computer equipment, (a) a centralised store of knowledge (102) including knowledge of past events relevant to the or each decision to be made; (b) a logical evaluation centre (101) for evaluating said knowledge in relation to a decision, (c) connecting data links (104) connecting between parts (a) and (b) and (d) at least one connecting data link (103) between part (b) and an external environment including at least one of 113, 114, 115, 116, and 117, the software including functional means to store previously established knowledge and return selected items of knowledge pertaining to past events when making decisions,and including means to provide an effective amount of integration between included decision- forming software routines and included decision-implementation software routines; so that decisions carried out shall be made more correctly and shall then be more closely adhered to after being decided.

More preferably, the invention provides a system that is adapted for use in a telecommunications business .

Illustratively, the enterprise using the system is a telecommunications business using the system to control aspects of customer behaviour, including churn (the effect whereby dissatisfied customers swap between suppliers).

Preferably the centralised store of knowledge (102) stores the knowledge in a set of functional groups including (a) immediately relevant memory as "customer analytical data sets" (105) (b) one or more databases of product and portfolio assets (106), (c) sets of beliefs (107), and (d) procedures, in the form of decision trees (108), and includes a function of analytical support having a plurality of components that is capable when in use of providing analytical support services.

In a first related aspect the system allows actors, some of whom are shown on Fig 1 as customers (116), suppliers, and associates (115) to constantly make and execute decisions.

In an alternative aspect, intelligent workflow computer technology (IW) including software according to the system of the invention for an enterprise includes a decision bus, integrated with demand and supply chains, and with command and control functions, capable of engaging all actors who make, manage, or execute decisions; who in a first phase of decision creation step through a flow of work, making optimised decisions thanks to provided selected items of knowledge and are then in a second phase engaged as necessary while the IW executes and monitors signed-off decisions; wherein the status and any resulting impact on enterprise performance of any decision is monitored and while information related to products/services and to customers' and suppliers' dealings with the enterprise is captured by the IW, stored, and made available to actors.

Preferably the DME intelligent workflow engine provides all actors with supporting knowledge for use when making, managing and executing all decisions in an intelligent manner.

Preferably, the ssystem allows customers to make purchase decisions; allows associates to make and execute all the decisions necessary to design, develop and execute products in the market for the customers to purchase and use; and allows suppliers to make and execute all the decisions necessary to provide associates with the inputs necessary to operate their enterprise.

Preferably, suppliers' products and services will either be an input to the products/services or infrastructure required for them to operate.

In a second related aspect the the invention provides a system that includes a function of analytical support having a plurality of components and is capable when in use of providing analytical support services.

In a third related aspect the invention provides a system for use in support of the making of decisions in a business as previously described in this section, wherein the software provides an organisation capable of supporting an at least partly physical layout of modules interconnected by data pathways or data links as a set of Decision-making and Management intelligent Engine (DME) processor modules (including 203, 204, 205, 206, 207, 208, 212); the modules being interconnected for the transfer of workflows by a data pathway (202) as further described in Tables 1, 2 and 3; in addition to conventional modules (including modules 214-211 and their data pathway (201)).

Preferably a decision transfer bus transports the knowledge between actors while they make, manage and execute decisions relevant to them.

In a second broad aspect the invention provides a system for use by an enterprise or business for rationalising and/or modifying the making of decisions by customers within an environment while the customers interact with the enterprise or business, wherein the software of the system provides at least one algorithm capable of describing and recording a particular customer's motivation and behaviour during an episode of contact with the business, with reference to previous customer behaviour and thereby allowing increased understanding of the customer's motivation and behaviour, so that the business may: enter into a unique dialogue with the customer in relation to a purchase of goods or services under consideration and prepare products and/or services for sale that are clearly differentiated from those of the business's competitors so that the business can increase revenue from sales and/or services.

In a related aspect, the software of the system includes at least one analytical data set capable of describing aspects of an episode involving a customer (Table 10).

Preferably and again in relation to customers, the software of the system further includes means to describe and evaluate, in order to optimise, at least one of (a) preventable churn (fig 5 a, 5b) and (b) price plan migration (fig 3 a, 3b).

PREFERRED EMBODIMENT

The description of the invention to be provided herein is given purely by way of example and is not to be taken in any way as limiting the scope or extent of the invention. Throughout this specification, unless the text requires otherwise, the word "comprise" and variations such as "comprising" or "comprises" will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

Copyright Notice. Portions of this patent include materials that are subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document itself, but otherwise reserves all copyright rights whatsoever in such included copyrighted materials.

DRAWINGS AND TABLES

LIST OF DRAWINGS

Figure 1 : Conceptual Model identifying the major components that interact to make

DME intelligent workflow function, i.e. the KnowledgeSphere and the decision bus.

Figure 2: Block Diagram of a Telecommunications Common Service Platform, identifying DME workflows and components.

Figure 3 as Figs 3 A, 3B. (two parts): Example of a decision tree. Price Plan Migration with decision points and actions required to address the problem of price plan migration

Figure 4: Telecommunications Value Creation Map, identifies five value creation objectives and a total of forty four value creation actions

Figure 5 as Figs 5 A, 5B (two parts): Example of a decision tree according to the invention, for handling Preventable Churn with decision points and actions required to address the problem of preventable churn. LIST OF TEXT TABLES (INCLUDED IN TEXT).

Table 1. Product Execution Workflow Components and definitions

Table 2. Product Development workflow components and definitions

Table 3. Product Execution workflow components and definitions

Table 4. Workflow Services components, definitions and data sets

Table 5. Application Services components, definitions and data sets

Table 6. Registration Services components, definitions and data sets

Table 7 (as three adjoined sheets). Analytical Support Services components, definitions and data sets

Table 8 Example of a belief model. The independent variables that contribute to preventable churn.

Table 9 'Decision tree process' type of classification table, with descriptions

Table 10. Example of an analytical data set holding customer information.

Table 11. compares current practice with the proposed new practice delivered by this software

Table 12: defines or clarifies some terminology used in this area

Table 13 : The part played by Analytical Packages.

Table 14: Telecommunication Value Creation Map (associated with Fig 4)

UNDERLYING PRINCIPLES

The underlying principles of the invention involve DME as defined above. In an ideal case, DME states that there is a symmetry between desired and actual results, expressed in equation

(1):

Desired Business Results = Actual Results. .. (1)

More realistically, equation (2) describes an asymmetry between Desired Business Results and Actual Results, in which the desired business results are diluted by Confounding Factors. We define confounding factors in a collective sense as a variable, or a stated constant affecting a desired result that has not been catalogued (identified). The effect of confounding factors is emphasised in equation (2) by squaring, owing to the mutually compounding nature of confounding factors.

Desired Business Results

_"■ = Actual Business Results

Confounding Factors^

•• (2)

In applied DME schemes, there is a need for a function of analytical support, or analytical support services. With reference to Fig 1, a conceptual model including the major components (the RnowledgeSphere and the decision bus) that interact to make DME intelligent workflow function is capable when in use of providing analytical support services. In Fig 1, the centralised store of knowledge (102) stores the knowledge in a set of functional groups including (a) immediately relevant memory as "customer analytical data sets" (105), (b) one or more databases of product and portfolio assets (106), (c) sets of beliefs (107) and (d) procedures, in the form of decision trees (108), and these components implement a function of analytical support that is capable when in use of providing analytical support services to a business operation. These are explained further under the Overall Example below, and are expressed in equation (3):

Analysis^(t)2*Decision²

= Desired Business Results

Decision Execution

Here, the variables determining the Desired Business Result are a constraint due to the natural liability of a business process. The covariant 't' as applied to the variable "analysis" represents truth or actuality determinants. As the analysis tends closer to the truth, so the analysis makes a greater contribution to the Decision Framework. As a corollary, if decisions are being made that are not based on "truth" (full or partial), the decision output must be false as a statement of Desired Business Results. The value oft' lies between just above zero and 1. If zero, it represents a "guess", and not a "decision". If greater than 1, advance knowledge is evidently present. Note that equation (3) tends to restore the original property of symmetry. The invention is embodied in a system, much of which is put into practical effect as software running in a digital computer (or network thereof).

These analytical support services offer an improvement over current software due to the way they treat events as multidimensional in nature, as opposed to the current state of the art where an event has only one dimension. A multidimensional event is called an Event Tensor. The tensor variables are loosely based on Chaos mathematical definition and concepts. Although it may seem counter-intuitive, there isn't an infinite set of usage behaviors and variables. Thus through the use of Belief and Segmentation models we create a "closed" behavioral environment - and therefore, the analysis can use techniques to look for patterns inside patterns. If the analysis is predictive in nature, the event tensor acts as the predictive agent. If the analysis is backward chaining, the event tensor acts as an isolation agent - weeding out arbitrary, ambiguous or confounding data.

By way of example we shall apply these concepts below in: "Example of DME applied to a Prepaid Mobile Telephone Business", in relation to Chum Management Software.

The basis for analytical "apple to apple" comparison is founded on establishing variability and cataloguing measurements into "Event Tensors". Although many business intelligence developers use the concept of Event, the definition is simplistic and one dimensional - capturing "something happened at this time" - basically a time stamp to establish an analytical baseline.

OVERALL EXAMPLE OF THE PREFERRED EMBODIMENT

This preferred embodiment of a more generalised invention has been constructed in order to relate an enterprise such as, but not only, to a telecommunications enterprise or Telco. It includes a number of sections or groups. Please refer to Fig I₅ which is a conceptual overview of the components and how they work together, thereby comprising the invention works. 100 indicates an entire body of working parts which is perhaps analogous to the human central nervous system. By analogy with the human body's brain, neurones (collectively referred to as "grey matter" are held in a group (101) herein referred to as the "DME workflow engine" and individual neurones could be regarded as analogies of various software elements (programmes or parts thereof), the neuropil (white matter) of the brain is analogous to connecting links such as the connection bus (103, 103, 104) and memory is loosely analogous to the "KiiowledgeSphere" (102).

As summarised in the Abstract, the invention provides Intelligent Workflow (IW) computer technology for an enterprise that includes a decision bus, integrated with demand and supply chains, and with command and control functions, that engages all actors who make, manage, or execute decisions. Optimised decisions are made by actors who step through a flow of work with guidance from previously collected critical knowledge (usually from within the Knowledgesphere). The same (or a closely linked) IW executes signed-off decisions, engaging relevant actors as necessary, thereby ensuring connection between making and executing decisions. Decision management includes monitoring the status, and any impact on performance by the enterprise, of any decision. Information related to products/services and to customers' and suppliers' dealings with the enterprise is captured by the IW, held in the

Knowledgesphere, and made available to actors.

The KnowledgeSphere (102) contains four types of specific codified knowledge that are used by the DME workflow engine to add intelligence at decision points. The "KnowledgeSphere" is analogous to "consciousness" i.e. an awareness of current events, and is responsive to beliefs 107, which are like unalterable instinctive behaviour; to decision trees (108) which are belief- based but can be altered over time (analogous to an animal or person learning its surroundings, to databases of Product and Asset Portfolios (106) and to immediately relevant memory within the "Customer analytical data sets" (105). Various parts of the IW are represented by 109 Marketing, 110 Network, 111 Billing, and 112 the Manager or management. The environment in which the IW exists is shown as the cloud 113 representing the Internet or World- wide Web - a public kind of wide- area network (WAN), and the cloud 114 containing a variety of non- weblike WANs such as WiFi, GPRS, 3 G; also DSL(Digital subscriber linking), cable and satellite links. These latter WANs terminate at (for example) mobile telephone handsets, ATM machines, STB and POS machines and importantly, PCs (personal computers. Finally the model provides that the cloud 114 includes customers 116. Also, cloud 113 includes suppliers 115 although in practice, either entity may be found inside either cloud, and "associate" entities (117) independently link the customers and the suppliers (dashed line).

The portfolio (block 106 in Fig 1) represents management of products and as a portfolio of data sets comprising a group of assets.

An example of a belief (107) is given in Table 8; which may be regarded as a list of factors liable to increase preventable churn, selected as one example of an important factor of concern toTelcos. The factors, or at least some of them are inter-related by means of the flow charts of Figs 3a and 3b, taken with Table 9. Fig 4 shows five value creation objectives and a total of 44 value creation actions. Belief Models, for which an example is shown in Table 8, perform two functions. They are used during the assessment stage of decision making to isolate independent variables for analysis purposes, in order to determine root cause of a problem during the decision execution stage to define the actions required to correct a problem.

For example, as per Table 10: analysis of the data shows that the Telco has failed to provide the customers with their monthly bonus. This then is a contributing variable to why the customers will consider that the Telco has not kept its promise to them, and hence churn. In order to address this problem the Telco will need to take and action that ensures the systems and processes responsible for paying each customer their monthly bonus (or whatever other point of difference is raised) are functioning correctly.

The decision trees of Figs 3 and 5 define a series of decisions and action steps that are used to automate common business operations. Decision trees have links to relevant contextual, industry specific knowledge held in the KnowledgeSphere. Table 9 sets out the types of process for the steps in a decision tree. For example the decision tree of Figure 3 a with 3b refers to Price Plan Migration and includes a series of decision points (diamonds) and actions (rectangles) so that a process can step through the decision tree to first decide if a price plan is performing as expected and then to make changes if necessary. Depending on various outcomes of the performance analysis a number of decisions are possible, i.e. the decision may be to shut down the existing plan and migrate all the customers to a new plan, or keep some customers on the existing plan while migrating others to a range of new plans. For example in figure 5a with 5b, one would step through the decision tree to first determine which pattern is seen in the data. The pattern will determine which are the appropriate actions to be taken to prevent churn. A decreasing pattern over time indicates that customers are churning because the Telco is not keeping its promise and is confusing the customer. An increasing pattern indicates that the customer is becoming more and more frustrated over time until they decide to churn.

Customer Analytical Data Sets as set out in Table 10 further describe some of the concepts used in the software of the invention. Decision Trees as in Figs 3 and 5 define a series of decisions and action steps that are used to automate common business operations. Decision Trees have links to relevant contextual, industry specific knowledge held in the KnowledgeSphere. Table 9 sets out the types of process for the steps in a decision tree.

For example in figures 3 a and 3b, Price Plan Migration a series of decision points (diamonds) and actions (rectangles) step through the decision tree to first decide if a price plan is performing as expected and then action changes. Depending on the outcomes of the performance analysis a number of decisions are possible, i.e. the decision may be to shutdown the existing plan and migrate all the customers to a new plan, or keep some customers on the existing plan while migrating others to a range of new plans. For example in figures 5a and 5b relating to Preventable Churn, a series of decision points (diamonds) and actions (rectangles) step through the decision tree to first determine which pattern is seen in the data. That will determine which actions are best taken to prevent churn. A decreasing pattern over time indicates that customers are churning because the Telco is not keeping its promise and is confusing the customer. An increasing pattern indicates that the customer is becoming more and more frustrated over time until they decide to churn.

DME Workflow Engine. Any one decision will have one status from the following range: Decision Making, Signed Off, Execution Phase, Performance Monitoring Phase, or Learning Captured. The workflow engine controls all decisions being made, and then controls the management and execution of the decisions. It ensures that the right knowledge from the KnowledgeSphere is provided to the actor at the right decision point. The DME Workflow Engine transports decisions on the decision bus from one actor to another as the status changes.

A Telecommunications Common Service Platform (CSP) block diagram is shown in Fig 2. This is an alternative, more physical representation than that of Fig 1 of some of the software building blocks within a Telco and includes decision buses (or data transfer channels), and prior-art and novel modules. The CSP integrates all the systems controlled by a Telco at both the data and process level in order to support the design, development and execution of new products (as well as many other business functions) using a workflow approach. Many of the systems will already exist in discrete functional silos. The DME intelligent workflow augments these systems and integrates them so that data and processes are integrated. Fig 2 shows various processor modules that make up a typical version of the invention. The concept of the decision bus is implemented as a series of workflow services (closely hatched lines) and publish and subscribe services (unhatched lines). These services connect the application workflow modules, the support workflow modules and the existing systems together to allow data and processes to interact throughout the enterprise. Fig 2 identifies DME workflows and related processor modules (components; shown as boxes 203-208 and 212, and underlined in the list) by means of a sparse sloping hatching. Other systems, included here for completeness, are unhatched. Two separate interconnecting buses 201 - unhatched, and 202 - hatched are shown.

List of module names used in Figure 2.

203 Product execution workflow 204 Product Development workflow

205 Product Desifin workflow 206 Workflow Services

207 Application Services 208 Registration Services 209 OSS integration services 210 Production services

211 Development Services 212 Analytical services

213 Convergent translation and Control Transfer services 214 Third party marketing and services partners.

215 Native Connectivity and EUAD. 216 Data Stores for Production services 217 Data Stores for Product Development 218 Data Stores for Product Design

219 OSS integration services.

Workflow Applications are next described in tables 1, 2 and 3.

TABLE 1 : Product Execution Workflow Components, and Definitions.

PRODUCT EXECUTION: EXECUTION WORKFLOW.

product design and forecast workflow as a forecast overlay for automated execution response.

TABLE 2: Product Development Workflow Components, and Definitions.

PRODUCT DEVELOPMENT: DEVELOPMENT WORKFLOW.

TABLE 3: Product Design Workflow Components, and Definitions. PRODUCT DESIGN; PORTFOLIO DESIGN AND MONITORING WORKFLOW.

Market Describes the general market conditions and assumptions which either Definition constrain or validate the Portfolio Strategy and individual Portfolio Member performance. Sets a baseline for overall market performance monitoring.

Customer Describes customer purchase and usage including lifestyle, community, Behaviour value and purchase attractors. Associates specific customer behaviours to Analysis. specific portfolio members for prediction, qualification, retention and ampaign messages.

Portfolio Describes, forecasts and monitors the entirePortfolio. The Portfolio consists Maintenance. of all products, services, promotions, proce plans, bundles, campaigns, offers, commerce options,(debit and credit) and connectivity options. Maintenance includes capturing all portfolio assumptions, building product families, designing custmer communication rules of engagement and setting all baselines.

Portfolio Describes each portfolio member from a connectivity, functionality, pricing, Member Design. ;arget customers, existing portfolio impact analysis and execution disciplines. Defines new capabilities required to be added to the capability canon. Produces automated forecast and campaign requirements.

Market Catalogue Describes the portfolio catalogue. The portfolio catalogue consists of three Maintenance. eparate catalogues: customer, product and asset. The portfolio catalogue associates customer behaviour to specific portfolio members and provides ;he customer "qualification" criteria for product recommendation. Competitive portfolios and "what if portfolios are supported.

Publish to Describes the necessary activity to convert a portfolio member design into a Development and market offer. This includes functional specifications (functionality, billing, Execution. integration, EUAD), provides a target customer population along with an integrated campaign execution plan and details all associated work required for product release to forecast specification.

TABLE 4: Workflow Support Services - See tables 4, 5, 6, and 7.

WORKFLOW SERVICES COMPONENTS, DEFINITIONS, AND DATA SETS

TABLE 5: APPLICATION SERVICES COMPONENTS, DEFINITIONS AND DATA SETS:

TABLE 6: REGISTRATION SERVICES COMPONENTS, DEFINTIONS AND DATA SETS:

TABLE 7 Analytical Support Services: components, definitions and data sets: part 1 of 3.

Universal Customer Portfolio Member ■ustomer Behavioural Identification Identification Analysis.

The analytical capability to The analytical ability to The analytical ability to combine customer profile create a product topology and describe specific customer nformation derived from provide unique identification behaviours or combine different OSS and/or ODS methods for individual multiple customwer data stores, into a single portfolio members, portfolio behaviours into an dentification method. This families, marketing based aggregate behaviour profile. service has two distinct portfolio segmentation and The analysis sets a number behaviours - legacy the association of customer of baselines based on onversion of historical data, behaviours to specific pecific events or time and creation of an products and product constraints to allow the dentification service to families. analysis and mode resetting propagate an unique customer while market behaviours number into all OSS and ODS hange or when new nvironments as part of the products are introduced into ustomer acquisition process. a behavioural population.

Segmentation Services. Analytical rating services Product Forecast Services

The ability to group any The ability to create a rating The ability to forecast the member in the Portfolioo apability^'for analysis only. performance (customer atalogue (product, customer Analytical rating is used for penetration, revenue and asset) as required for price point analysis, price ontribution, asset predictive services , forecasted lasticity, product design, ontribution,) of any services (targeted customers), price point analysis (barrier portfolio member. Based on behavioural qualification or to entry or to usage) churn a method of normalising performance analysis. prediction or new price-plan brecast performance based Segmentation includes introduction impaact on customer asset community analysis for analysis. Existing price plans, constraints, market position lifestyle determination and 'what-if price plans, bundles clarity, pricing and network ommunity based campaigns. and designer price plans are asset constraints (ability to included in this analysis. grow with customer demand). TABLE 7 Analytical Support Services components, definitions and data sets: part 2/3.

TABLE 7 Analytical Support Services components, definitions and data sets: part 3/3.

All labels are Data Storage descriptors, related to the previous twopaiis of Table 7.

Workflow Services;

These workflow services are required to support the application workflows.

Workflow Role Assignment; Who has access to each workflow is defined by the combination by declaring a «role» by login ID. Roles have granularity and "eyes only" capabilities

Workflow Declared Alarms and Escalation; Each workflow may support multiple alarms based on analysis, time delay, failure to complete a workflow, assumption changes that may affect downstream workflows

Workflow Event Declaration; Workflows generate events which are published to «wake-up» other workflows or start monitoring services. Customer behaviours may generate events requiring workflow administration

Workflow Dynamic Navigation; Workflow behaviour is dynamic based on specific catalogue entries, canon conditions or specific logic based on Portfolio Member types or analytical changes.

Workflow Registration; Workflows may be added at a later date. Registration services allows new workflows to be introduced at the role/login level (different Workflow versions) or at the macro level Workflow Library and Flow Content; Captures the workflow logic, description and relationships. The workflow library acts as a macro navigation service coupled with logic checking to make sure workflow flow control passes specific conditions

Administrative Data, Role Data, Workflow Propagation Log, Workflow Publication Gates, Work in Progress, Workflow Catalogue.

Application Services;

These application services are required to support the application workflows.

Login Services; Login declaration, password assignment and maintenance

Administrative Services; Miscellaneous controls and services for technical support

Version Control and Historical Leveling; Due to the historical nature of the workflow, different version of the same application service may require historical consideration

Testing and Performance Monitoring; Service Level Agreement Testing and Standard Analytical Tests for Reasonable data and analysis results

Application Service Registration; Registration of application publication results for use by other applications or for workflow support

Application Service Subscription; Subscription to specific analytical results or granular data to perform local logic or used to publish analytical results

Datasets; Administrative Data, Login and Security Control, Published Applications, Subscribed Application Data, Work in Progress, Error Log and Operation Scripts

Registration Services;

These registration services are required to support the integration of the application workflows with other systems based on publish and subscribe.

Registration Services; Each workflow, application, catalogue, canon or analytical object is required to be registered for multi-use and conflict avoidance.

Publication Services; The services to control publication of analysis or work. Publication Services deals with issues of "recalled" publication, publication logging, and publication taxonomy and object creation control.

Subscription Services; The services to control subscription of analysis or work. Subscription Services deals with issues of "recalled" publication for items already subscribed, and subscription logging.

Logging; The physical logging of data creation, methods, and subscriber.

Datasets; Registered Providers, Publication Catalogue, Subscribed Objects, Logs, Error Log and Operation Scripts.

A nalytical Support Services;

These analytical services are required to support the application workflows.

Universal Customer Identification; The analytical capability to combine customer profile information derived from different OSS and/or ODS data stores into a single identification method. This service has two distinct behaviours - legacy conversion of historical data and the creation of an identification service to propagate a unique customer number into all OSS and ODS environments as part of the customer acquisition process.

Portfolio Member Identification; The analytical ability to create a product topology and provide unique identification methods for individual portfolio members, portfolio families, marketing based portfolio segmentation and the association of customer behaviours to specific product and product families.

Customer behavioural Analysis; The analytical ability to describe specific customer behaviours or combine multiple customer behaviours into an aggregate behavioural profile. This analysis sets a number of baselines based on specific events or time constraints to allow the analysis and model resetting as market behaviours change or new products are introduced into a behavioural population.

Product behavioural Analysis; The analysis to describe specific product behaviours at an individual or community level. The product can be any member of the portfolio and/or any combination of members. Product behavioural analysis allows for the qualification of customer behaviour to specific products and provides the guideline for new product development.

Behavioural Predictive Services; The ability to predict either a product or customer behaviour for a set criteria. The criteria includes, churn, retention, acquisition, product recommendation, product impact performance, product migration, product penetration and asset catalogue trend analysis. Campaign services are included (predict which customers would be interested in what campaigns along with the message "type") .

Baseline Monitoring Services; The ability to either automatically or manually set various baselines without destroying existing baselines (baseline over lay). Baseline deltas allow the alarming of deviation from baseline. A forecast (such as customer growth, product penetration, and customer churn) are specialized baselines and are included in baseline services

Segmentation Services; The ability to group any member in the Portfolio Catalogue (product, customer and asset) as required for predictive services, forecast services (targeted customers), behavioural, qualification or performance analysis. Segmentation includes community analysis for lifestyle determination and community based campaigns. ό

Analytical Rating Services; The ability to create a rating capability for analysis only. Analytical rating is used for price point analysis, price elasticity, product design, price point analysis (barrier to entry or barrier to usage), churn prediction or new price plan introduction impact analysis. Existing price plans, «what if» price plans, competitive price plans, bundles and designer price plans are included in this analysis

Product Forecast Services; The ability to forecast the performance (customer penetration, revenue contribution, asset contribution) of any portfolio member. Forecast services is based on a "deductive" method of normalizing forecast performance based on customer asset constraints, market position clarity, pricing and network asset constraints (ability to grow with customer demand)

Revenue Assurance Services; The services to compare end to end revenue or funding record generation through to accounting services.

Analytical Object Services; The ability to produce a standard set of analytical objects consistent across different levels of granularity of reporting (detail, time, segmentation, workflow specific requirements). Analytical objects are used in workflow management, reporting (analysis, performance, management, exception and reference) .

Catalogue Management Services; The ability to control catalogue object creation, update and retirement and set specific alarm conditions. This services include all event creation (manual or automatic).

TABLE 8: BELIEF MODEL. List of Factors Liable to Increase Preventable Churn

Premise 1 : Preventable Churn can occur if the Telco does not keep its "Promise" to the Customer.

Excessive Network Quality issues with dropped calls or throughput call quality.

Premise 2: Preventable Churn can occur if you fail to differentiate your products, or your products cause customer frustration.

;hanges will increase preventable churn.

21 A lack of Entanglement products and offers will increase preventable churn.

22 Product Design and Content that do not support market messages will increase preventable churn.

23 Lxcessive difficulty in using products and services will increase preventable churn.

Belief Models perform two functions. They are used during the assessment stage of decision making to isolate independent variables for analysis purposes, in order to determine root cause of a problem and during the decision execution stage to define the actions required to correct a problem.

For example, as per Table 8 above, analysis of the data shows that the Telco has failed to provide the customer with their monthly bonus. This is a contributing variable to the possibility that the customer will consider that the Telco has not kept its promise to them, and hence churn. The Telco will need to take action that ensures the systems and processes responsible for paying each customer their monthly bonus are functioning correctly. Please refer to the flow-chart diagrams of Decision Trees shown as Figs 3 a and 3b, 5 a and 5b and table 10.

TABLE 9: Decision Tree process type classification table, with descriptions.

Point insights achieved. The specification level of an Assessment Point is an Interpretation Library component, ("if you see this, the most likely casual point is this") Assessment points always lead to decision points.

DM Data Mining A specific point in a decision tree is to be retrieved and analyzed to and feed an assessment point (AssessP).

Analysis.

MCPB Monitoring A placeholder in the decision tree to take a snapshot of current and Control business process or measurement performance to be able to Point- with determine a "before and after" picture. A MCP point usually baselines. precedes an Action Point (ActP) or Decision Point (DecP)

DecP Decision A placeholder in the decision tree where the analysis and insights Point have been presented and a decision is required to proceed.

ActP Action Point A placeholder in the decision tree where Actiion is to be taken, based on a specific decision.

Decision Trees define a series of decisions and action steps that are used to automate common business operations. Decision Trees have links to relevant contextual, industry specific knowledge held in the KnowledgeSphere. Table 10 sets out the types of process for the steps in a decision tree.

DME

The workflows required to implement a Decision Making and Management Intelligent Workflow Engine will now be illustrated. This engine integrates the right knowledge at the right decision point with a workflow process for all decisions across the enterprise. As a result, decisions will have one of the following status, Decision Making, Signed Off, Execution Phase, Performance Monitoring Phase, Learning Captured. The workflow engine first controls all decisions being made, then it controls the management and execution of them. The DME Workflow Engine ensures that the right knowledge from the KnowledgeSphere (101- Fig 1) is provided to the actor at the right decision point. Then, the DME engine transports decisions by means of the decision bus from one actor to another as the status changes. Each decision is tracked and the performance impact of its execution is stored in the KnowledgeSphere as learning for future use, thus forming part of the corporate memory. For workflow applications, view figures 3a with 3b, 4, 5a with 5b, and Tables 1, 2 and 3.

For example in the flow chart of figure 3 a with 3b, "Price Plan Migration" a series of decision points (diamonds) and actions (rectangles) step through the decision tree to first decide if a price plan is performing as expected and then action changes. Depending on the outcomes of the performance analysis a number of decisions are possible, i.e. the decision may be to shutdown the existing plan and migrate all the customers to a new plan, or keep some customers on the existing plan while migrating others to a range of new plans.

Explanation of the flow chart of Figure 3 (as Figs 3a and 3b).

Box 301 is entitled "Historical Performance Analysis if Prior Price Plan Migration Occurred" (lesson learned). Output goes to box 302 (Predict Customer Population at Price Plan Transition Date), then to all three destinations: 303 (Analytical review Decision Tree), 304 (Communication Plan decision Tree), and test box 305 (Is the population on the Rate Plan Significant?). If yes (Y), box 307 performs a "Price Plan Portfolio Analysis"; otherwise (N) box 306 performs a "Forced Migration at Least Cost". The Price Plan Portfolio Analysis continues to test box 308 "Is the migrating price plan performing on par with other price plans based on customer behaviour in the portfolio?". If evaluated as "No" (N) the flow chart goes to the inter-figure connector 313 leading to Fig 3b. If "Yes" (Y) box 309 performs a "Price Plan Economic Performance Analysis" and the output is evaluated in 310: "Is the price plan economically sustainable?". If "No" (N), flow proceeds to connector 313 leading to Fig 3b. If yes (Y) flow proceeds to box 311 "Perform Price Plan Network Impact Model" and the output is tested in box 312: " Is the price plan sustainable, based on network capacity"?. A "Yes" (Y) flows to the connector 314 leading to Fig 3b. A "No" (N) flows to connector 313 leading to Fig 3b. Figure 3b operates on information received at either connector 313 or connector 314 and terminates at box 316, box 326 or box 328. Box 315 performs Risk Analysis on received data, and box 317 asks "Is the general level of risk acceptable?" If yes, then box 318 receives data via (317A) and "Performs Benefit Analysis on each Option and Decision Prediction. If no, then data flows via line (317B) to a termination 316 entitled "Do Not Migrate", which also receives all incoming flow from Fig 3a via connector 314. Reverting to output from box 318, this is passed to evaluation box 319. Output 319 A is "Extent and Embrace Existing Price Plan " and is passed to box 316, previously described. Output 319B is "Forced Migration" and is passed to box 328. Box 328 is a termination and is entitled "Perform Migration to Existing Other Price Plans". Output 319C is "Incented Migration" and is passed to box 320, which applies the test" "Will any new price plan be available to the entire market?". If "Yes" (Y), box 321 performs market analysis based on the new price plan proposal, and the result is tested in box 323 "Does the proposed price plan have a positive impact?" If "Yes" (Y), flow continues to box 324, "Create New Price Plan based on Customer Usage Behavior", then to box 325 "Create Incentive Plan to Migrate to New Price Plan, then to termination box 326 "Perform Migration to New Price Plan". Reverting to "No" (N) output from box 320: this is evaluated in box 322 with the test "Does the Proposed Price Plan have a Positive Impact on Migrating Customers?" If it does (Y) then flow is taken to box 324, previously discussed. If not (N) flow goes to box 327 "Create Incentive Plan to Migrate to Existing Price Plan" and then to previously mentioned termination box 328 entitled "Perform Migration to Existing Other Price Plans".

Figure 5 (as 5a with 5b), is a second flowchart related to "Preventable Churn" and customer- related decisions according to the principles of DME, namely, distinguishing between relevant 'and non-relevant variables in the process of implementing biusiness decisions. An early step is to determine which pattern is seen in the data (styles are shown as 503a .. 503e in

"Analytical Results" box 503). To traverse Figure 5 in detail, start box 501 in Fig 5a is entitled "Form preliminary analysis", and leads to box 502; "Perform general Churn Pattern Analysis", based on episodes, to find primary and secondary patterns. The pattern seen in the data will determine which actions are taken to prevent churn. For example, a decreasing pattern (as seen in the bar graph of 503 a) indicates that customers are churning because the Telco is not keeping its promise and is confusing the customer. An increasing pattern (503b) indicates that the customer is/are becoming more and more frustrated over time until they decide to churn. Box 504 assesses the general product offer, according to business rules. Box 505 is intended to determine if there is a fundamental problem with performance in the product offer, and evaluation box 506 tests "Is the offer breaking any promises?" If yes, flow continues to box 507, "Assessment of Actions and Impacts to correct Fundamental Product Offer Problems", which continues to box 508;an Action Process to review "Product Offer Performance". Output from 508 is merged with the "No" (N) output from evaluation box 506 at box 509; which assesses whether the Product Offer is too confusing to the customer. Then box 510 determines whether there is unreasonable customer behavior with regard to "offer purchase" or "offer usage efficiency". The evaluation box 511 "Are customers exhibiting unreasonable behavior?" forwards a "Yes" (Y) output to box 512 "Assessment of unreasonable behavior" and then box 513; "Unreasonable behavior action process" then to connector 514 (to Fig 5b) while a "No" (N) output from box 511 goes directly to connector 514. In Fig 5b, the only input is connector 514, leading to box 515 "Assess product offer Attracting Casual Customers". Flow continues to box 516; "Determine what Components Attract Casual Customers" and the result is tested in box 517 "Can Churn be Associated with a Specific Offer Component?" A "yes" (Y) output is assessed in box 518 "Assessment of Component Attraction of Casual Customers" then through box 519; "Casual Customer Attraction Action Process" the output oif which is merged with a "No" (N) output from evaluation 517 and fed to box 520; "Assess your offer as complementing a Competitive Offer", then to box 521 "Determine if your offer design is playing to a competitor's offer strength" the output of which is tested in box 522 "Can churn be associated with a specific Competitive Offer?. A "Yes" (Y) output leads directly to termination in box 523; "Assessment of churn due to competitor's offer". A "No" (N) output leads to a string of processes: box 524 "Assess if the product offer plays to the competitive offer?"; box 525 Assess the Calling Community as reflected in the offer"; box 526 "Assess if there is Fraud or unreasonable Sales Channel Behavior", and terminates in box 527 "Assess Product Quality Issues". It will be clear that these flow charts embody the principles of DME as herein defined.

TABLE 10: Example of an analytical data set, holding customer information.

Spend Rate based on cellular services only.

Customer Account Average Daily Average daily amount of money transferred Transfer Rate from this episode to another customer.-- ^■

Customer Account Average Daily The average daily amount left in the account. Residual Account Amount

10 Customer Account Episode The recharge amount used to start the episode. Starting Recharge Amount

11 Customer Account Starting Residual account balance immediately before Residual Account Amount he recharge event that started the episode.

12 Customer Account First-Billed The amount of time between theepisode start Front Latency Duration date and any event that generated revenue.

13 Customer Account Number of The count of unique numbers called during the Unique Called Numbers ipisode.

14 'ustomer Account Maximum Call The difference between longest and shortest call Duration Variation duration, in minutes and seconds.

15 'ustomer Account Number of The count of unique numbers messaged during Unique Messages :he episode.

16 'ustomer Account Average The average time between revenue-generating Revenue Generating Front ivents. Frequency

Product Design and Monitoring;

This category includes the workflows required to design a new product and maintain the product portfolio catalogue used to draw new product components from.

Market Definition; Describes the general market conditions and assumptions which either constrain or validate the Portfolio Strategy and individual Portfolio Member performance. Sets a baseline for overall market performance monitoring.

Customer Behaviour Analysis; Describes customer purchase and usage behaviour including lifestyle, community, value and purchase attractors. Associates specific customer behaviours to specific portfolio members for prediction, qualification, retention and campaign messages.

Portfolio Maintenance; Describes, forecasts and monitors the entire Portfolio. The Portfolio consists of all products, services, promotions, price plans, bundles, campaigns, offers, commerce options (debit and credit) and connectivity options. Maintenance includes capturing all portfolio assumptions, building product families, designing customer communication rules of engagement and setting all baselines.

Portfolio Member Design; Describes each portfolio member from a connectivity, functionality, pricing, target customers, existing portfolio impact analysis and execution disciplines. Defines new capabilities required to be added to the capability canon. Produces automated forecast and campaign requirements.

Market Catalogue Maintenance; Describes the portfolio catalogue. The portfolio catalogue consists of three separate catalogues: customer, product and asset. The portfolio catalogue associates customer behaviour to specific portfolio members and provides the customer "qualification" criteria for product recommendation. Competitive portfolios and "what if portfolios are supported .

Publish to Development and Execution; Describes the necessary activity to convert a portfolio member design into a market offer. This includes functional specifications (functionality, billing, integration, EUAD), provides a target customer population along with an integrated campaign execution plan and details all associated work required for product release to forecast specification.

Product Development;

These are the workflows required to develop and test a new product in preparation for its execution into the market:

Forecast and Capacity Management; The ability to predict portfolio performance as a consumer of network or service capabilities and to automatically create dimensioning specifications (including pricing, existing equipment lifecycle and architectural introductions into the production environment).

Product Development; The ability to convert product specifications into market ready solutions. This includes Customer Experience management measurements as an automated criteria, integrated customer help, all training and operational documentation and integration services across all departments. Product Testing; The workflow and automated testing scenario generation to ensure the product performance as defined from all aspects (integration into native networks, revenue assurance, billing assurance, performance assurance, operational logging and native instrumentation) .

Operation Certification; The workflow to certify all operational systems are certified to support all portfolio members. This includes training, OSS integration, Call Centre integration, monitoring, outage resolution, trouble ticket collection and all product acquisition methods (IVR, sales distribution, etc.).

Business Certification; The certification process to ensure the business is ready for the production of a portfolio member. This includes all marketing communication, packaging, problem resolution, and customer access request handling.

Publish to Production; The workflow to move all certification processes and the product into the production environment (ready for market purchase). This includes adding the member to all catalogues, setting all behavioural monitors, modification of access control databases and translation services and all operational monitoring and alarming.

Product Execution;

The workflows required to execute the new product into the market and monitor its performance.

Architecture, Strategy and Standards; The workflow to define an architectural strategy including future capability, redundancy, existing architecture changes as a function of growth, replacement of architectural components due to new technology or cost benefits. This workflow includes the methods for setting all standards and network settings.

Financial Management; The workflow to capture pricing information to allow the «financial» analytical objects to be created as part of product design and performance forecasting. This workflow also provides all ERP reporting, and balanced scorecard services.

Product and Capacity Integration; The workflow to integrate new products or sizing requirements into an existing production environment. This workflow validates the business, operational and testing certification processes as part of the installation or change management process.

Project Services; The workflow to assign and track assets to accomplish a specific goal within a specific timeframe or budget target. Performance Monitoring and Optimization; The workflow to ensure the Customer Experience is executed as defined by the architectural and product standards. This includes outage monitoring, individual component performance, capacity growth monitoring and ensuring all GOS and QOS targets are maintained and achieved.

Forecast Execution; The workflow to execute an integrated product introduction under the rules of customer communication, target customer product execution and campaign execution. AU forecast activities executed are monitored and triggered for customer behavioural response and fed back to the product design and forecast workflow as a forecast overlay for automated execution.

EXAMPLE OF DME, AS APPLIED TO A PRE-PAID TELEPHONE BUSINESS

Table 11 below, in relation to our PortfolioExpress Churn Management Software for Prepaid Mobile Telephone Users, compares current practice with the proposed new practice delivered by this software as previously described in this section; together with corresponding examples. This may be referred to as "Intelligent Workflow (IW) computer technology.

Table 11

These analytical support services offer an improvement over current software due to the way they treat events as multidimensional in nature, as opposed to the current state of the art where an event has only one dimension. A multidimensional event is called an Event Tensor. The tensor variables are loosely based on Chaos mathematical definition and concepts. Although it may seem counter intuitive, there isn't an infinite set of usage behaviors and variables. Thus through the use of Belief and Segmentation models we create a "closed" behavioral environment - and therefore, the analysis can use techniques to look for patterns inside patterns. If the analysis is predictive in nature, the Event tensor acts as the predictive agent. If the analysis is backward chaining, the Event tensor acts as an isolation agent - weeding out arbitrary, ambiguous or confounding data.

PortfolioExpress adds multidimensionality to the concept of an Event.

PortfolioExpress analytical solutions use the concept of event as an inter-relationship between non-related, yet logically consistent data values (example: logically, a Customer must execute a "Top Up" event to fund their prepaid account. A billable call event cannot occur unless there is funding in the prepaid account) — therefore, although the two events are completely different — there is an implied relationship between the two: the Shadow between the Data,

By using Event Tensors (see above), PortfolioExpress can support the following conceptual models:

• A Business Measurement or Business Process is an artifact of a Business Decision

(in PortfolioExpress a business decision is catalogued as a Decision Event Tensor).

• Customer Behaviors are driven by Motivational Purchase Decisions A behavior is what took place, a purchase decision is why it took place

Table 12: clarifies further terminology used in this area.

Software for use in support of the making and carrying out of decisions in an enterprise as claimed in claim 1 or in claim 2, characterised in that the software provides an organisation capable of supporting a physical or virtual layout of interconnected modules named as a set of Decision-making and Management Intelligent Engine (DME) processor modules (including 203, 204, 205, 206, 207, 208, 212); the modules being interconnected for the transfer of workflows by interconnecting data links (202) as further described previously in Tables 1, 2 and 3; in addition to pre-existing modules (including modules 214^211 and their interconnecting data links (201)). Combining equations (1) and (2) derives the general equation (3) for achieving results symmetry according to the principles of DME.

(Analys is ^² *De cis ion²)/D e cis ion Exe cutio n

Actual Business Results = Confounding Factors²

(3)

The equation is deceptively simple: The efficiency of your Decision Framework (DME) defines the accuracy and aggressiveness of achieving the value of Desired Results. The more that confounding factors are isolated and mitigated (converted to asymmetrical determinants), the less risk there is of actual result achievement not matching desired results.

Note that Confounding Factors cannot achieve a value of "zero". If a "zero" value were possible, actual business results would reflect an infinite value. In practice, asset constraints (an asymmetrical determinant) will always be in play.

The equation is applied recursively until the Resultant decision expires (the Desired Result Decision must state a boundary for desired result achievement. The boundary can be anything, time, budget, value, expenditure, etc.). The aggressiveness of the equation is based on the granularity of the Decision Framework (DME) and the period of recursive review (how often desired results are compared to actual results or how often confounding factor values are updated and mitigated)

By way of example we apply these analytical support services in the PortfolioExpress Churn Management Software for Prepaid Mobile Telephone Users in Table 13 as follows;

Table 13: The part played by Analytical Packages.

Types, data

VARIATIONS

Thus far, the Examples presented herein have mainly related to telecommunications, wherein discussions on matters such as "preventable churn" were convenient for illustration of the invention. It will be apparent to the skilled reader that the same or a similar design can be applied to planning for other kinds of organisation apart from telecommunications. Some possible organisations include banks, airlines, general merchants, supermarkets, local bodies (councils) and national governments.

INDUSTRIAL APPLICATIONS AND ADVANTAGES

DME intelligent workflow across the telecommunications enterprise represents the next evolution of business management and systems. The high-level advantages are thus reduced costs of operation, reduced capital costs and increased revenue. The improved economics all result in competitive advantage to the enterprise. Customer satisfaction gives a competitive edge. Specifically the DME intelligent workflow enhances all 44 areas of value creation identified in figure 4, the Telecommunications Value Creation Map. These are named in table 14, which also provides text labels for the blocks in Fig 4. It is worth noting that in yr 2007, the pre-paid option for mobile telephone accounts is far more commonly used in most countries than the formerly preferred credit accounts. Table 14: Telecommunication Value Creation Map

Leakage

403, Revenue Realisation Revenue 4020 Reduce Fraud and Bad Debt Leakage 403, Revenue Realisation Revenue 4021 Reduce Billing Arbitration Errors Leakage 403, Revenue Realisation Revenue 4022 Reduce Settlement, Interconnect or Roaming Leakage Errors 403, Revenue Realisation Barriers to 4023 Reduce Revenue Loss from Termination Issues Revenue 403, Revenue Realisation Barriers to 4024 Reduce Revenue loss from Late Service / Re Revenue Orders 403, Revenue Realisation Barriers to 4025 Reduce Revenue loss from Network Revenue Congestion/ Outage

403, Revenue Realisation Barriers to 4026 Reduce Revenue Loss from Learned Customer Revenue Behavior

404 Supply Chain Network 4027 Reduce Network Costs Optimisation Management

404, Supply Chain Network 4028 Reduce Cost of Service Fulfillment Optimisation Management

404, Supply Chain Network 4029 Reduce Cost of Routing Optimisation Management

404, Supply Chain Forecast 4030 Reduce Costs from Increased Forecast Optimisation Management Accuracy

404, Supply Chain Forecast 4031 Reduce Asset Overhead and Reclaim Assets Optimisation Management

404, Supply Chain Regulatory 4032 Reduced Costs from Regulatory Reporting and Optimisation Management Management Services

404, Supply Chain Vendor 4033 Reduced Vendor Costs and Improved Optimisation Management Management

405, Risk Management Operational 4034 Reduce Service Disruptions / Process Failure Risk 405, Risk Management Operational 4035 Reduce Capital Costs from System Instability Risk or Failure 405, Risk Management Operational 4036 Reduce Sabotage or Hacking Failure Costs Risk 405, Risk Management Operational 4037 Reduce Supply Chain Disruptions/Costs Risk 405, Risk Management Operational 4038 Risk Mitigation through Disaster Recovery Risk 405, Risk Management Forecast Risk 4039 Improve Forecast Accuracy of market, competition, regulatory and product mix/price

405, Risk Management Forecast Risk 4040 Improve New Product Performance Prediction (Revenue and Uptake)

405, Risk Management Forecast Risk 4041 Improve COC/COG and Financial Assumptions Forecast Accuracy, (including Budget)

405, Risk Management Transition Risk 4042 Reduce Costs / Risk of New Technology Introduction

405, Risk Management Transition Risk 4043 Reduce Costs / Risk of New Revenue Streams Introduction

405, Risk Management Transition Risk 4044 Reduce Costs / Risk due to Mergers or Acquisitions

Finally, it will be understood that the scope of this invention as described and/or illustrated herein is not limited to the specified embodiments. Those of skill will appreciate that various modifications, additions, known equivalents, and substitutions are possible without departing from the scope and spirit of the invention as set forth in the following claims.

Claims

WE CLAIM:

1. A system based on DME (Decision Making and Execution; a decision discipline) for use when creating and when carrying out decisions in an enterprise, characterised in that the system (100) when in use together with digital computer equipment and software, includes means to provide (a) a centralised store of knowledge (102) including knowledge of past events relevant to the or each decision to be made; (b) a logical evaluation centre (101) for evaluating said knowledge in relation to a decision; (c) data links connecting (104) between parts (a) and (b); and (d) at least one connecting link (103) between part (b) and an external environment (including at least one of 113, 114, 115, 116, and 117); the system including functional means to return selected items of knowledge pertaining to past events when making decisions, and including means to provide an effective amount of integration between included decision-forming software routines and included decision-implementation software routines, so that, when in use, the results of decisions carried out shall more closely and accurately reflect desired results, namely those results predicted prior to taking the actions correlated with a given decision.

2. A system as claimed in claim 1, characterised in that the centralised store of knowledge (102) stores the knowledge in a set of functional groups including (a) immediately relevant memory as "customer analytical data sets" (105) (b) one or more databases of product and portfolio assets (106), (c) sets of beliefs (107), and (d) procedures, in the form of decision trees (108), and includes a function of analytical support having a plurality of components that is capable when in use of providing analytical support services.

3. A system for use by an enterprise for rationalising and/or modifying the making of decisions by customers within an environment, characterised in that the system includes software that provides at least one algorithm capable of describing and recording a particular customer's motivation and behaviour during an episode of contact with the business, with reference to previous customer behaviour and thereby allowing an increased understanding of the customer's motivation and behaviour, so that the business may: enter into a unique dialogue with the customer in relation to a purchase of goods or services under consideration, and prepare products and/or services for sale that are clearly differentiated from those of the business's competitors, so that the business can increase revenue from sales and/or services.

4. A system as claimed in claim 3, characterised in that the system is adapted for use in a telecommunications business.

5. A system as claimed in claim 4, characterised in that the enterprise using the system is a telecommunications business using the system to control aspects of customer behaviour, including churn (defined as the effect whereby dissatisfied customers swap between suppliers).

6. A system as claimed in claim 5, characterised in that the system includes at least one analytical data set capable of describing aspects of an episode involving a customer; the analytical data set being exemplified as Table 10.