Channel: Jacqueline Bloemen 
Five Steps to Predictive Performance Management
15 Juni 2009
Admiring the interactive shopping experiences offered by Disney and similar organizations, a New England retailer set in motion a strategy to create a similar environment in order to attract a larger share of their target market. The retailer believed the strategy would create differentiation, which would translate into higher profits.
This strategy had two important aspects: layout of stores and interactions between employees and customers. As part of the strategy, employees would actively engage customers to increase cross-sell opportunities. However, this strategy omitted a critical dependency – employee capabilities. As a result, while the new strategy was successful in stores with high employee capabilities, it was damaging for the rest. Statistical analysis later proved this correlation and quantified the impact on objectives and financial results. Early recognition of the correlation between these measures could have saved this retailer considerable time and money.1
There are many stories just like this one. As several studies have verified, executives' concerns are focused squarely on the successful execution of strategy.2 Effective performance management closes the gap between strategy and its execution. Predictive performance management not only closes this gap, but helps you iteratively perfect both the strategy and its execution.
Strategies for Performance Management
When managers think about performance management, they often consider the culture and people in their organizations. How will they cultivate an environment of accountability? How will they secure executive sponsorship? How will they help others embrace process changes that will be required?There is no doubt that overcoming cultural issues is paramount to successful performance management. But people represent just one of several success factors. Organizations must also view performance management strategically and holistically. In support of the organization's goals and objectives, managers develop strategies, initiatives, measures and performance indicators, and provide a means for proving department effectiveness. Taking a holistic approach to performance management, managers embrace technology to better develop successful strategies that support objectives and drive culture and process changes.
Scorecards and strategy maps help them understand performance today, detail and communicate the strategy, and ensure alignment with and accountability to organizational goals. With scorecards that draw from analytics, managers can not only manage against objectives, but also drive real improvements. They can predict performance to validate and modify strategies and retain focus on the measures and activities that really matter.
Analytics Can Improve Performance
Predictive performance management involves strategically applying analytics – a suite of statistical methods that sifts through vast amounts of data to discover meaningful correlations, patterns and trends. Analytic techniques for modeling, forecasting and simulating potential outcomes answer critical questions such as:- Which measures drive the business and which do not?
- Why did these problems occur?
- Where do I need to improve and by how much?
- Are employees aligned with the strategy?
- Which is the right course?
- How should I adjust my strategy and modify initiatives?
Organizations that take advantage of a wide range of analytics to support performance management will have confident answers to these key questions.
Here are five steps to making full use of analytics and improving performance.
Step 1: Gain transparency with strategic context.
Without transparency, managers and stakeholders cannot effectively manage performance. We've all heard the story about the executive meeting in which each manager brought his/her own set of numbers. Everyone then scrambles to find out who's right and who's wrong. Sound business decisions rely on accurate, consistent and comprehensive data. However, according to a global online survey of almost 2,500 managers, led by Tom Davenport of Babson College in collaboration with Harvard Business Review, 39% of respondents indicated that questionable data quality/accuracy/consistency was a primary barrier preventing the adoption of analytical, quantitative performance management. Thirty-one percent referenced the inability to aggregate data as a primary barrier.
To achieve financial and operational transparency, data can be integrated from siloed ERPs, transactional systems and third-party data. This data is then cleansed (for accuracy), normalized (for consistency) and rationalized (for relevancy). Not every manager is interested in the same information.
Managers bring the relevant information into focus by viewing performance as it relates to their strategy and objectives within their scorecard, with the confidence of consistent and accurate information.
Managers can show strategic context, while inhibiting silo thinking, by displaying metrics and objectives within a strategy map. This visual, macro diagram illustrates how metrics and objectives are connected in a cause-and-effect manner across multiple perspectives. Specific to the manager's needs, these perspectives can vary widely, demonstrating the support of methodologies (such as the balanced scorecard), initiatives (such as sustainability) or even show how teams' programs support a department's objectives. The balanced scorecard perspectives, for example, include learning and growth, internal processes, customers and finance, and span multiple business units.
Step 2: Validate the strategy and establish metrics that matter.
As we learned with the retail example, a strategy, no matter how brilliant, is only successful when appropriately supported and executed. To be most effective, each manager's view of performance should be limited to 10-25 key performance indicators (KPIs). Instead, many managers oversee hundreds of KPIs, diluting priorities and strategic focus. So how can managers focus on the metrics that matter?
Interconnected strategy maps are a good place to start. These allow for various views (organization, business unit, dept, team, individual) that help filter which KPIs should be shown at the various levels. They use hierarchical logic to show how metrics at the low end (individual) roll up or connect to higher-level metrics/objectives at the organizational level.
In this way, no single view breaks the 25-KPI rule, helping managers focus.
The initial strategy, however, is a hypothesis. It's important to begin execution, but there are no guarantees that you will have chosen the right metrics and objectives to support your strategic intent. This may become visible by seeing many individual targets being met with little impact on overall performance. The construct of the strategy map (showing the connections across multiple perspectives) will help management ask better questions. These may ultimately lead to the selection of better metrics or a complete rethink of the strategy. This is what happened with the retailer.
A faster approach builds on a strategy map to apply statistical analysis of results. These can accurately prove relationships with facts – not mere conversation or eyeballing of trends. What might take two years of discussing trends before making a change, could be done with as little as twelve months of data using statistical analysis.
Interconnected strategy maps are a good place to start. These allow for various views (organization, business unit, dept, team, individual) that help filter which KPIs should be shown at the various levels. They use hierarchical logic to show how metrics at the low end (individual) roll up or connect to higher-level metrics/objectives at the organizational level.
In this way, no single view breaks the 25-KPI rule, helping managers focus.
The initial strategy, however, is a hypothesis. It's important to begin execution, but there are no guarantees that you will have chosen the right metrics and objectives to support your strategic intent. This may become visible by seeing many individual targets being met with little impact on overall performance. The construct of the strategy map (showing the connections across multiple perspectives) will help management ask better questions. These may ultimately lead to the selection of better metrics or a complete rethink of the strategy. This is what happened with the retailer.
A faster approach builds on a strategy map to apply statistical analysis of results. These can accurately prove relationships with facts – not mere conversation or eyeballing of trends. What might take two years of discussing trends before making a change, could be done with as little as twelve months of data using statistical analysis.
Step 3: Align resources to strategy.
To be effective, the strategy must be clearly articulated, translated into measurable results and communicated so employees can understand the vision and how day-to-day activities contribute to it. With KPIs and targets set and mapped to the strategy, a scorecard can be put in place to continually monitor results. Yet most large organizations function in a constantly changing environment of interdependencies - related variables that affect each other on parallel levels, feed into higher levels and ultimately determine overall performance and profitability. How do you gain visual clarity of such complexity?
As mentioned earlier, a strategy map helps individuals see how everything is connected. Analytics helps managers understand relationships and patterns by probing historical data with exploratory data analysis (e.g., clustering and correlation) and predictive analysis (e.g., regression and decision trees).
Advanced modeling techniques can then isolate and highlight these causal relationships so managers can base hypotheses on a more fact-based understanding of how KPIs affect each other. With cause-and-effect relationships determined and validated, managers can see the strength of their relationships to one another and better predict potential outcomes based on achieving certain results. Based on facts, they are also more likely to act on them faster.
A manager will use this information to better allocate people and money toward higher-value activity. But with strategy and relationships understood, the focus switches to understanding how cost, profit and value are created. Think of it this way – the cost of achieving an objective may outweigh the value received, but that information is rarely communicated in a strategy map. With that in mind, organizations benefit from applying cost and profitability optimization techniques to determine what drives cost, then model or simulate how changes in resource allocations (primarily money and people) impact overall results. In addition to displaying performance against goals, insights from analytics (such as root-cause analysis) and cost drivers (from activity-based costing) can be displayed within the scorecard to be understood within its strategic context.
As mentioned earlier, a strategy map helps individuals see how everything is connected. Analytics helps managers understand relationships and patterns by probing historical data with exploratory data analysis (e.g., clustering and correlation) and predictive analysis (e.g., regression and decision trees).
Advanced modeling techniques can then isolate and highlight these causal relationships so managers can base hypotheses on a more fact-based understanding of how KPIs affect each other. With cause-and-effect relationships determined and validated, managers can see the strength of their relationships to one another and better predict potential outcomes based on achieving certain results. Based on facts, they are also more likely to act on them faster.
A manager will use this information to better allocate people and money toward higher-value activity. But with strategy and relationships understood, the focus switches to understanding how cost, profit and value are created. Think of it this way – the cost of achieving an objective may outweigh the value received, but that information is rarely communicated in a strategy map. With that in mind, organizations benefit from applying cost and profitability optimization techniques to determine what drives cost, then model or simulate how changes in resource allocations (primarily money and people) impact overall results. In addition to displaying performance against goals, insights from analytics (such as root-cause analysis) and cost drivers (from activity-based costing) can be displayed within the scorecard to be understood within its strategic context.
Step 4: Predict potential outcomes with a high degree of confidence.
True competitive differentiation requires more than querying and reporting on past events and basic calculations that deliver hindsight. To make improvements that can differentiate the business, organizations must operate from more than gut instinct. By performing analysis based on events from the past, managers can better understand not only what happened and when, but also why. Regression analysis can reveal previously unknown relationships between KPIs in an easy and intuitive way. In the process, the analysis reveals broader and deeper insights into how an organization or department operates. To further hone strategy, a variety of analytical methods – e.g., neural nets, genetic algorithms, experimental design and optimization - can be applied. Understanding how changes in strategy and policy will impact future performance can help managers set priorities by highlighting areas that require the most attention.
Leading organizations use analytics to create forecasts, perform what-if simulations and scenario modeling in order to understand complex relationships in data, and also to model behavior, systems and processes.
These technologies help managers predict the path ahead so they can appropriately adjust pricing, inventory, staffing, etc., for effective operations. Managers can anticipate future challenges and opportunities, assess the impact of changing KPI values and promptly respond with fact-based decisions.
These technologies help managers predict the path ahead so they can appropriately adjust pricing, inventory, staffing, etc., for effective operations. Managers can anticipate future challenges and opportunities, assess the impact of changing KPI values and promptly respond with fact-based decisions.
Step 5: Monitor and correct course for continuous improvements.
To iteratively improve strategies, analytics helps managers test alternative strategies through modeling what-if scenarios and running simulations, so they can confidently select a positive course of action. In this way, predictive performance management becomes self-learning and self-tuning.
With the understanding of the full context and impact of historical actions, managers can better identify early indicators of success or failure and can make effective decisions as part of a strategic-learning loop. This self-tuning capability supports the ability to be proactive rather than reactive in daily operations.
With predictive insights, managers can identify best strategic actions. Analytics embedded into the performance management process empowers organizations to better focus on developing and supporting an effective strategy and ensuring successful execution. Scorecards and strategy maps that leverage analytical and predictive insights help managers learn from the past – and then identify the best strategic actions for today and the future. With these capabilities, executives, managers and employee teams can focus on what really matters, see outcomes sooner, set more effective directions and change course quickly.With the understanding of the full context and impact of historical actions, managers can better identify early indicators of success or failure and can make effective decisions as part of a strategic-learning loop. This self-tuning capability supports the ability to be proactive rather than reactive in daily operations.
End Notes:
- Radical Action for Radical Times by Jonathan Hornby, SAS Institute. In press.
- 2006 Monitor Analysis' Survey of 354 Executives and The Execution Premium, 2008 by Robert S. Kaplan and David P. Norton.
Umfangreiche Quellen für Business-Intelligence- und Data-Warehousing-Profis.
NETWORK WEBSITES: BeyeUNIVERSITY | BeyeRESEARCH | BeyeBLOGS | BeyeSURVEY | BeyeNETWORK | SearchDataManagement.com | SearchBusinessAnalytics.com
PARTNERSEITEN:
TDAN.com | DSSResources.com | DataGovernance.com | Decideo.fr
Twitter
LinkedIn
YouTube
Pressemeldungen | Über uns | Kontakt | Site Map | Privacy Policy | Mediadaten
All Rights Reserved. Copyright 2004 — 2013, TechTarget
BeyeNETWORK™, a TechTarget company
Kommentare
Möchten Sie den Beitrag kommentieren? Login oder Registrieren Sie sich heute!