Académique Documents
Professionnel Documents
Culture Documents
Table of Contents
Executive Overview .......................................................................... 1 Aligned Resource Optimization The Resource Optimization Model ...... 3 Five steps to resource optimization.................................................... 4 The technology to support the Resource Optimization Model ................ 8 Underlying technology to support optimization ..................................11 Closing thoughts ............................................................................12 From SAS, the leader in business intelligence ....................................13 Examples Resource optimization across the enterprise ....................14 Optimizing retail revenue ............................................................14 Optimizing profit ........................................................................15 Optimizing human capital ...........................................................16 Optimizing for sustainability ........................................................17 Optimizing marketing campaigns .................................................18 Optimizing IT performance ..........................................................19
Becca Goren, Ed Hughes, Mary Crissey and others at SAS contributed to this white paper.
ii
Executive Overview
The fast-food franchiser has regional distribution hubs, a fleet of trucks of various capacities (some refrigerated and some not) and hundreds of stores needing on-time deliveries that vary from week to week. Given carton and pallet dimensions, sell by dates, distance, urgency, number of drivers, weather and restrictions on working hours, what is the best way to load trucks and route these deliveries? The catalog retailer wants to better manage its call centers, direct mail and e-mail channels. The millions of customers in its database represent the gamut of buying histories, buying propensities, profitability, demographics and cost to serve. Given capacity and costs for each channel, which customers should receive which offers through which channel? What will happen if you add a channel, trim budget for another or initiate a new contact policy? The manufacturing line has been underperforming on one shift due to periodic shortages of staff and materials, and bottlenecks in product inspection. Should the company invest in a just-in-time inventory system, add third-shift staff, reduce the sample size of post-production testing, cut one shift but add a new production line or outsource the more time-consuming processes? In each case, the answer would be, It depends. The best way to allocate resources depends on the nature of the resources and constraints at hand and the organizations mission. Is it a Six Sigma organization, striving to reduce process variability and increase product quality? Is it a lean manufacturing outfit, driving out every possible cost? Does the organization live and breathe Total Quality Management (TQM), where everyone is tasked to deliver ever-improving value to customers at continually lower costs? Does the organization embrace performance-based budgeting or Economic Value Added (EVA ) principles, which link costs to results yet recognize some costs as investments in disguise? Or has it adopted a balanced scorecard approach, which provides an organization-side approach to measuring and tracking performance against objectives?
By definition, optimization is the design and operation of a system or process to make it as good as possible in some defined sense. It is in the defined sense where things get murky. What is optimal for you, with your goals and values, could very well be suboptimal for the next person. Every performance management paradigm, every mission statement, could point to a different definition of success and therefore to a different way to optimally allocate resources (people, money, technology). Even within one organization, theres no one-size-fits-all proposition. If you legislate a sole method of resource optimization across the organization, you could miss out on the advantages of uniquely tailoring the approach to optimize the attributes of greatest interest in each functional area. Least cost, highest quality, greatest innovationdifferent teams could realistically have very different charters, all under the umbrella of a unifying, top-level strategy.
1
However, on the flip side, when every department handles performance measures and processes in its own way, it can be difficult to determine exactly how resources are being applied to drive organizational strategy. Without a clear picture of resource use across the enterprise, including interdependencies across functional areas, managers cannot know how to allocate resources to optimize organizationwide results. How, then, does an organization do this well? How do you optimize resources in what ultimately is a dynamic and often poorly defined environment or one that is well-defined but ineffectively executed? In the past, these were daunting challenges. Resources, constraints and market conditions continually change. Even if you managed to get the necessary details to develop optimization models, complex models could take days to run. Opportunities might pass before they were even revealed. That was then. Technology has refined the possibilities. Now, strategic visions are shared and managed through scorecards and strategy maps. Hidden costs are transparent, their roots understood. Analytically derived intelligence drives performance improvements. Models that formerly took days to run can now deliver insights in minutes. Reports that once required special requests to the IT department can now be accessed on demand via self-service, Web-based interfaces. Sales targets and performance metrics that were once defined based on instinct and intuition can now be mathematically validated (or invalidated). Results can be automatically woven back into the process for continual improvements. Organizations that have embraced these new technologies report gains of millions of dollars and payback in just a few months. If that sounds good, read on for a look at the people, process and technology attributes that are the basis for aligned resource optimization.
Functional leaders must recognize that their departments are connected to each other, and how they are connected matters. A focus on functional optimization leads a company into the in isolation yes, in combination no sub-optimization trap.
Steve Beeler Director, Special Situations, Production Modeling Corp. (PMC)
Optimization helps you determine the best that can happen, so you can take action in ways that will deliver significant performance improvements. Advances in technology have made this process easier and more powerful. 2
Objective
Data Inputs
Historical or current operational data or analytically derived information.
Decision Variables
Actions or choices that can realistically be carried out in pursuit of the objective.
Constraints
Recommended actions
The optimal course to meet the objective balanced against constraints and decision variables.
Implementation
Execution on recommended actions.
Its not enough for executives to agree on the goals, business rules and constraints, and decisions that will be made. An aligned process will ensure the best choice for each decision variable the recommended actions will actually be implemented, and that requires accountability and commitment from all parties.
Successful organizations identify a champion an executive with power and influence to endorse the resource optimization project. Champions can help scope the project, identify obstacles and implementation issues, and ensure accountability to drive the project to completion. Armed with an effective scorecard and strategy map, this champion can share this information with other executives, managers and employee teams, so everyone across the organization can work together and understand how their daily activities contribute to the companys success.
Test the optimization model for suitability. No software can determine the most appropriate representation of a decision problem by an optimization model. Software is a tool that guides you. Training and experience, often from an optimization modeler, will help you to choose the best model. This means that once youve built the model and used it to produce a solution, you need to consider whether the mathematically derived optimal solution is suitable for the original business problem. At this point it is not unusual to discover that some key element of the model has been overlooked or misconstrued, making the optimal solution (and the decisions that it represents) unsuitable. The model might be correct, but some data used by the model might be incorrect. Or your understanding of the original business problem might be flawed. In these and other such cases, you need to step back through the modeling process, address the difficulty and then move forward with the improved model. This iterative process is quite common and represents another aspect of the art of optimization modeling. Fine-tune the conceptual and descriptive models in an iterative process. Models can and should be updated as needed, and should always be flexible. You wont always choose the best one from the start. Dont let this deter your efforts. Define an initial model and refine it as you move forward and learn more. One of the key differentiators of SAS software is that users are never stuck with black-box calculations. After looking at the output generated from the analytical model, you can go back and tweak the model by relaxing certain constraints or adjusting the primary objective to be optimized. This iterative, what-if process, reassessing assumptions to tweak the formulation, also adds valuable insights into the organization and process at hand. Early assumptions may be overturned by the insights revealed by analytics. For example, you may have initially assumed staff resources were fixed and then find that hiring extra staff yields optimal output that more than compensates for the additional cost. Establish formal mechanisms for learning from past actions. You would want to know how well the model works in the real world, and incorporate the knowledge from previous iterations into future ones. What did the implementation look like in the end? Were decisions made? If so, were they based on facts rather than gut instinct? Were those decisions effective in driving improvement in alignment with organizational goals? If not, why? How can the process be improved? Were we measuring the right things?
7
How did this decision or process affect related business units or the organization as a whole? Did the optimization recommendation make sense? In other words, was it workable or unworkable? Why? If the result was not what you would expect from a mathematical optimization model, revisit the model to determine whether objective, decisions, constraints, resources and so on were properly identified. The results of this assessment should be automatically incorporated back into the system to continuously improve models, metrics and future results.
Minimum cost does not equal maximum profits. Otherwise, companies would have no people and no assets. Missing in many companies are enterprise-level analytical tools to enable collaborative efforts to continuously improve financial and operational performance.
Steve Beeler Director, Special Situations, Production Modeling Corp. (PMC)
Resource optimization must support organizational goals. Scorecards and strategy maps help show the interdependencies among resources and objectives, and how resources do or do not support organizational goals.
Improve outcomes with multiple analytic techniques. With or without optimization software, analytics have a powerful role in resource optimization. With analytic intelligence, you can confidently anticipate the result of a strategy in advance, test various scenarios and use optimization software to select the best of all possible courses. You can explore and understand complex relationships among resources, behavior, systems and processes; assess the impact of changes in KPI values; and respond more quickly with fact-based decisions. Then this technology helps you learn from past results so you can use that knowledge to realign indicators and improve resource allocations at the next iteration. Unlike generic business intelligence software reports on what has happened (forcing you to figure out what will happen next and what to do about it), optimization software and business analytics identify the best forward-looking course of action the best use of limited resources to achieve strategic objectives. Balancing goals against limitations, you can answer questions such as these: Is what were trying to accomplish possible? How are we doing now? How can we do better? Whats the best we can do? What happens if conditions change? The organization that could successfully answer these questions would have a clear advantage, yet few have capitalized on the (readily available) means to do it.
Your chosen technology can be implemented in low-risk stages. Start with a pilot project, prove its worth and expand it as the business case warrants. This phased approach is made possible with a technology platform that is affordable at the startup level yet scalable to the enterprise level.
10
What it does Integrates data from across the organization, transforms and cleanses it in real time, and ensures accuracy and consistency. Answers diverse business problems for which variables, constraints and outcomes can be mathematically defined. Monitors and displays key performance indicators that tie to strategy, with at-a-glance visuals.
How it supports resource optimization Creates a common foundation for delivering trusted information throughout the enterprise. Adds value to corporate data and ensures access to the best possible data for operations and decision support. Mathematically calculates optimum resource allocation to achieve stated objectives, given multiple, weighted decision variables and constraints. Helps organizations focus on performance and opportunities to take appropriate action, align resources and day-to-day activities with corporate strategy, and adapt to meet changing conditions. Helps align the organization and its resources by articulating goals and the initiatives that support those goals throughout the enterprise. Clarifies how resources are consumed by an activity, and the financial consequences; identifies the cost constraints of an optimization exercise. Helps synchronize financial and operational strategy across the organization, to every level with repeatable, sustainable processes for financial reporting, risk analysis and achieving goals.
Optimization
Dashboards/ scorecards
Strategy map
Activity-based management
Helps determine accurate costs and cost drivers at the activity level.
Financial management
Used by business-unit heads for budgeting and planning and by finance executives also for consolidation and reporting.
Manages processes and Delivers function-specific analysis and insights resources at the department that can be incorporated into optimization level to support strategic goals. models and feedback loops.
Provides packaged solutions with prebuilt models and metrics for specific industries.
Delivers analysis and insights that can be incorporated into optimization models for specific industries, such as banking, insurance, retail, government, education and manufacturing.
11
Closing thoughts
Fast, cheap, quality: pick any two. Good for a chuckle but hardly wise counsel for market success. Classic resource optimization questions are a balancing act of all three, within the resources and constraints at hand colored by the organizations unique mission. In the past, this was a daunting challenge. Optimization must address factors that span functional areas and multiple stakeholders. Without high-level sponsorship, the project would lack the necessary data and consensus. Without big-picture perspective, the optimization model would yield suboptimal results. And without buy-in, even the best mathematical models could generate answers that no one actually implements. This document outlines a conceptual model and five-step process that address these organizational challenges of resource optimization projects: Step 1. Define the objective, reflecting organizational strategy and objectives. Step 2. Get buy-in and foster accountability. Step 3. Define the conceptual resource optimization model. Step 4. Define the descriptive resource optimization model. Step 5. Implement and update the model. Organizations that have embraced this process, in one form or another, report gains of millions of dollars and payback in just a few months. SAS provides the technology foundation to make it possible. If the possibilities of resource optimization sound intriguing, be sure to read the supplementary section of this paper, Examples: Resource Optimization Across the Organization, for a high-level look at optimization exercises in support of various objectives, from maximizing profit to improving marketing campaigns to revealing the most productive workforce strategies and sustainability initiatives.
12
To find out more about SAS solutions for performance management, visit www.sas.com/solutions/pm. Visit the sas.com resource center to download the companion white papers: The Aligned Organization: How performance management can align activities and resources with enterprise-level strategy and market conditions. Optimization with SAS/OR : What it is, whats new and how it adds value
Predictive Performance Management: Continually improve performance by applying the power of analytics
13
Decision Variables
Determine optimal price for each product by store.
Constraints
Recommended actions
List of optimal price per product per by item by store.
Implementation
Fed into price execution and core merchandising system.
14
Optimizing profit
Understanding and maximizing profit used to mean little more than reporting the bottom-line profit and loss results of legal entities and ferreting out costs wherever possible. Now it also means interpreting financial performance to predict the future impact of business decisions and having profit-and-loss information for each customer and product, calculated at the individual transaction level. The challenge is that many organizations combine inaccurate cost information from traditional costing systems with other financial and operational data to generate reports on customers and products. This approach doesnt show true profitability, so it does not accurately reveal which customer, product or channel mix scenarios will be optimal. How SAS can help: SAS Financial Intelligence helps businesses improve the financial performance of the entire organization. The suite includes an enterprise business intelligence platform, integrated consolidation, budgeting and planning, scorecards and strategy maps, and cost and profitability management. Sample case: A state department of transportation sought to satisfy diverse mobility needs, address concerns for public safety and the environment, and maximize the use of existing resources within the agencys $430 million annual budget. The agency adopted SAS Financial Intelligence to identify the costs associated with business processes and to determine if activities and resource allocations were aligned with the organizations mission. As a result, employees gained a better sense of how their work contributed to the agencys objectives. In addition, the elimination of various high-cost, low-benefit activities saved the agency $2 million in the first year alone. Example of a profit optimization model
Objective
Data Inputs
Customer dimensional data, individual order line transactions, product dimensional data, customer behavior rates.
Decision Variables
Constraints
Widely distributed, drillable profit and loss reports needed per customer reporting timeliness.
Recommended actions
Determine appropriate service levels to customers based on profitability.
Implementation
How will customers react to different service levels?
15
Objective
Employee location, job titles, salary/salary grades, employee skills and profiles, # of employees needed per site/per season/per hour.
Decision Variables
Allocate X # of Y type of employee to Z location.
Constraints
Cost, supply, demand, time/ season, site location.
Recommended actions
Prioritized # of employees, skills, job titles and location combinations to pursue.
Implementation
Will we need to hire to fill the gap? Do we need to reallocate resources?
16
Objective
Data Inputs
Project-specific inputs such as utility utilization, activity costs, projected fines, waste disposal options.
Decision Variables
Prioritize list of sustainability projects that minimize risk while maximizing ROI.
Recommended actions
$X initial investment with $Y monthly incremental cost will yield $Z return/avoidance of penalty in #T years.
Implementation
Apply resources to develop and execute project plans.
17
Objective
Data Inputs
Business requirements, priorities/service levels, capacity, existing projects.
Decision Variables
Which customer segment should be targeted with which offer for which type of campaign.
Constraints
Recommended actions
Which combination of offer channel/customer to use.
Implementation
Determine campaign timing, prioritize by segment purchase behavior.
Update model based on campaign response. Add loyalty and house-holding data to customer segments.
18
Optimizing IT performance
CIO responsibility extends far beyond keeping the IT infrastructure afloat, juggling data and applications, and delivering on service level agreements. Todays CIO is expected to contribute strategic thinking about how to add value to corporate data, create new insights to drive success, and optimize IT resources in alignment with the organizations mission and goals. In the quest to optimize IT performance, CIOs face a host of conflicting objectives. They are pressured to provide more processing power, servers, storage space, redundancy, bandwidth, power and self-service capabilities than ever. And they have to deliver it on less than ever: fewer people, dollars and days. To succeed, CIOs need a full understanding of resource utilization and costs, optimized in alignment with business requirements and service level agreements. How SAS helps: SAS IT Intelligence is a comprehensive solution for IT that helps you optimize resources, services and financial impact, all in support of strategic business goals. Sample case: A major European financial services organization, with more than US $694 billion in assets and 56,000 employees, had been growing rapidly through acquisitions and attained more than 5 million retail banking customers. But the IT organization faced major challenges in integrating the acquired systems and keeping more than 1,900 applications running smoothly especially as Internet banking grew by 40 percent. Capacity management was a process of crisis management. With SAS IT Intelligence, the team was able to align IT direction with the corporate bottom line, while bridging organizational gaps. According to the institutions IT operations manager, the bank can now ensure that adequate resources are available and functional at the required time, and that everything performs according to specifications, while correctly accounting for and allocating all costs. Example of an IT optimization model
Objective
Data Inputs
Business requirements, priorities/service levels, capacity, existing projects.
Decision Variables
Match activities resources, service levels and support capacity.
Constraints
IT capacity, budget # of resources, project length, available technology, service level agreements.
Recommended actions
Prioritized list of IT activities/projects and associated resources and service levels to support business requirements.
Implementation
Build recommendations into documented, monitored and automated processes/controls for example ITIL, CMMI, PRINCE2 and update/document/communicate IT goals within scorecard/performance management application.
19
SAS Institute Inc. World Headquarters +1 919 677 8000 To contact your local SAS office, please visit: www.sas.com/offices
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. indicates USA registration. Other brand and product names are trademarks of their respective companies. Copyright 2008, SAS Institute Inc. All rights reserved. 103387_465057.0308