CPM: What it is and how it is different from traditional approaches? -- Part One
CPM: Systems and Steps to CPM -- Part Two
CPM: 12 Best Practices in Implementing a Solution -- Part Three
CPM: Selecting the Right Technologies -- Part Four (A)
CPM: Selecting a CPM Vendor -- Part Five
CPM: Emerging Technology Systems -- Part Six
In the last article [Part Four (A)], we took a look at the role of technology in supporting a CPM initiative. In this article we will look at how to evaluate CPM software.
10 Rules of CPM Systems
CPM systems are fundamentally different from traditional systems used for planning and reporting. To deliver effective Corporate Performance Management, CPM solutions abide by the following ten 'rules':
Rule 1: Multi-dimensional
CPM systems are multi-dimensional in nature. "Multidimensional" describes the way in which numerical information can be categorized and viewed. For example, consider the number 57. The number must be defined to have meaning. In this case it is revenue. It could be further described as having occurred in the West region in the month of June for the year 2002. To define it further, 57 could be described as an actual result for product X.
This example describes six aspects of the number 57. These are known as "dimensions". The values of each dimension -- e.g., east, west, south, north -- are known as "members" of the relevant dimension. CPM systems are multidimensional; that is each piece of information held by the model is qualified in terms its dimensions and the appropriate member of those dimensions. Advantages of multidimensional systems include simple maintenance -- to modify a structure only requires the changing of the appropriate member; and users can view results from any valid perspective -- e.g., by product, by unit, by time.
Rule 2: Common business rules and data
CPM models have a common set of dimensions, dimension members, business rules and data. Each item is only held once -- e.g., only one set of base data is held even though there may be multiple management structures in operation. This means that if the number changes, it automatically updates analyses in any of the other dimensions.
Business rules may be restricted to specific versions and processes -- e.g., the rule for revenue may be price*units sold in the budget, but for actual results, revenue is whatever we received for the product.
:Having a common set of rules and data eliminates the time and effort in moving data or updating rules in multiple systems. This has the result of greatly improving integrity as only one version 'of the truth' is ever held.
Rule 3: Built-in Financial Intelligence
CPM models have built-in financial intelligence that automatically understands the different types of measures when processing data. Measures can be defined as both financial and non-financial. This means that measures such as ratios will not be consolidated, while non-financial measures such as headcount and volume will not be converted to a base currency.
CPM systems can also distinguish between P&L and balance sheet accounts. The model then uses this information to correctly calculate year-to-date totals in both reports and ad hoc analyses.
Similarly, they 'understand' debit and credit assignments and use this information to produce correct better/worse variance reporting.
Often there is a need to support multiple currency perspectives for global planning and reporting. CPM models are able to translate accounts at different rates, detect and calculate exchange gains/losses, and then consolidate results into a base or multiple base currencies.
The more sophisticated CPM models are also able to convert measures at multiple sets of rates, such as budget and actual rate, enabling the comparison of results to assess the impact of exchange fluctuations.
The embedded financial intelligence of a CPM system saves time and effort in the setting up and subsequent maintenance of a system. It ensures that end users have the right answers when they prepare their own ad hoc reports and analyses.
Rule 4: Unrestricted rule access
Business rules can be assigned to measures that are able to access values in any other measure and in any other dimension and dimension member combination. This capability allows the setting up of allocation rules that, for example, spread the total cost of the marketing department across all sales cost centers, based on the volume of product that was budgeted by each. Another use of this capability allows the setting up of central drivers that can be used by budget holders -- e.g., the calculation of revenue by taking the volume sold by each sales department and multiplying it by the price entered by the marketing department.
To evaluate these kinds of rules correctly, the CPM model must be able to perform multiple passes of the model. In the previous allocation example, the model must first calculate the total cost of the marketing department before it can allocate the cost, which will then require a second consolidation. These rules can be recursive: that is, the result of a rule may be used by itself in a second pass of the same rule. Without this capability, the CPM model would not be able to cope with allocations, the calculation of minority interests, and the generation of cash flow statements.
Rule 5: Time Intelligence
CPM systems understand the concept of time. First, they support financial accounting cycles of any length. For example, budgets may be entered monthly, cash flows generated quarterly, and actual results reported weekly. Where time matches between versions of data, the model will allow direct comparisons.
Next, a CPM model can hold data for any length of time both in the past and in future years, to enable historical comparisons and trend analysis.
Some measures may be defined as "opening balances." Built-in financial intelligence enables CPM models to automatically populate these from the appropriate closing balances. This is an essential requirement when forecasting or planning into the future.
The CPM model can also accommodate the relative referencing of time. For example, when reporting the last six months, the CPM model will know how to roll over to last year when reporting data in the first quarter of a new year.
CPM models can generate year-to-date figures without having to write rules. These summations automatically handle the correct treatments of P&L and balance sheet accounts.
Finally, to reduce maintenance time, CPM models have an indicator of the current period. This indicator is used to determine the focus of any report. For example, if a report shows the current month and last month, setting the indicator to "May" will tell the model to automatically show results for May and April. Built-in time intelligence makes CPM models much easier to set up and helps organizations cope with the move towards continuous planning.
Rule 6: Integrated Processes
CPM systems consist of functionality that manage and support the core CPM processes of strategy formulation, scenario analysis, tactical planning, budgeting, communication, monitoring, forecasting, and reporting.
These processes are fully integrated allowing organizations to focus on implementing strategy and avoid the strategy gap. CPM systems also help organizations to become event or trigger-based rather than calendar based, as users can move from planning to budgeting to reporting as results determine.
Rule 7: Collaborative
CPM systems support the communication of objectives, goals, timetables and allow access to strategic planning information as it affects users. They also allow users to communicate back to management on their ability to meet or exceed targets.
To do this they integrate with organizational portals as well as email and other communication systems. The result is reduced effort in controlling an enterprise wide application and better focus as everyone is aware of their responsibilities.
Rule 8: Guided Analysis
The amount of data available today for making decisions is growing at an enormous rate. Left unassisted in this environment, users can waste time looking for exceptions that do not exist or that cannot be found because they are buried within a sea of data.
The real issue facing many users overwhelmed with information is knowing what is critical and needs attention, and what can be safely ignored. CPM systems overcome this by highlighting critical exceptions through a range of analytical and visualization techniques that include:
Color-coding is used to help draw users' attention to performance areas that need investigation. A column or row of numbers can be color-coded according to some simple rule, such as "show all numbers less than minus 5 in red, between minus 5 and positive 10 in yellow, and above 10 in green."
Even with color-coding, the user still has to look through the list to find exceptions. By producing a sorted list, such as "show the top 10," the user now only has to look at the top of the report to find the exceptions. This makes it almost impossible for the user to miss exceptions as the most important information is presented first.
Hierarchical exceptions reveal exceptions in the context of their position within a structure. In the example shown, the hierarchy illustrates actual budget margin variance by product. The coding of each variance is shown in relation to where it fits within the product hierarchy. In this way, it is possible to view and understand thousands of exceptions quickly.
Because business is changing so rapidly, organizations rely on CPM systems to provide proactive alerts to bring exceptions to their notice without having to search and analyze reports. Software agent technologies do this searching automatically without the user even needing to be present. When an exception is found, the software generates an alert, often in the form of an email that is then sent to the appropriate user. Upon opening the email, the user can select the alert and view the place in the database that generated the alert. Exception rules can be quite sophisticated, such as "generate an alert when sales for three consecutive months have decreased, while at the same time advertising expenditures have increased." Like warning systems in a car, this means users will only be alerted when there is an issue that needs immediate attention, instead of having to spend time monitoring results.
Once an exception has been highlighted, users can -- security permitting -- access data online in any time period or version without advance notice. Reports do not need to be pre-configured. Users are able to view and analyze data across any appropriate dimensions, without limitations, such as by initiative, product, line of business, and so on. They are able to rotate and nest dimensions as well as drill down to lower levels of detail within the model. These drill-downs use the most current structures. When the lowest level of the business model is reached, drill-downs are then capable of going back to the underlying data source.
CPM systems also allow end users to produce their own, unrestricted (security permitting) analyses. These analyses include sorting, color-coding, charting, and ad hoc calculations. These analyses can be saved and recalled by users at a later date but will then feature the most current data.
These analytical capabilities are essential if users are to detect variances and their causes. With them, CPM systems prevent surprises. Users are always aware of current and potential exceptions and have time to evaluate alternative courses of action. These analytical capabilities also greatly reduce the time and effort the typical finance staff spends in supporting end-user queries.
Rule 9: Central, mainstream technology database
CPM systems are built around a central database to which everyone attaches. The technology used in these databases can be multidimensional, relational or a hybrid of both. Each has unique characteristics and needs to be carefully chosen to match the organization's requirements.
Multidimensional databases were designed specifically to ease the setting up of business models and to enable interactive multidimensional analysis. In a multidimensional database, data is stored in "cubes" that combine the various business dimensions of an organization. The advantages of multidimensional databases are that they perform extremely well in complex data analysis and are relatively easy to set up and maintain. The disadvantages are that they are number based and need additional technologies to handle information in text and date form, which is essential for CPM solutions. They also lack standards, meaning that many applications featuring multidimensional databases are proprietary.
Relational databases have been around for over 30 years and are common in every organization. They underpin the general ledger, ERP, CRM, and HR systems and are used by IT departments to create customized systems. Vendors such as Microsoft with SQL Server, Oracle, and IBM with DB2 dominate the industry. Because of their prevalence, standards have emerged that all vendors comply with in terms of updating records and providing access. In a CPM system based on a relational database, information is stored in relational tables rather than in cubes. Tables are two-dimensional in that they consist of records, and each record consists of fields or columns. For each table, the number of fields will be fixed. A relational database can consist of many tables that can have different numbers of fields. Multidimensional analysis is achieved by creating a number of tables containing specific fields that are related. This type of design is known as a star schema or snowflake schema. In the past, the disadvantage of the relational approach was poor performance due to the complex queries required to mimic the functionality of multidimensional cubes. However, use of star schemas combined with the dramatic performance improvements in relational technology has transformed the way these applications perform. For many organizations the performance is comparable to multidimensional technology. The strength of relational databases lies in their openness, that is, their ability to integrate with other systems such as the general ledger and ERP systems. Another strength is their scalability, that is, the ability to support hundreds of users with large volumes of information. Openness and scalability are two key requirements in the deployment of CPM solutions.
In recent years, there has been a great deal of activity among the relational vendors, particularly with Microsoft, Oracle, and IBM, in developing multidimensional capabilities within their relational products. Microsoft, for example, offers Analysis Services with SQL Server 2000. This product is multidimensional in appearance and maintenance but uses relational technology to store data and metadata. Oracle 11i and IBM DB2 OLAP Services provide similar capabilities.
These technology leaders have recognized that organizations need both relational and multidimensional capabilities but would prefer them to be combined in a single technology. This combination frees organizations from potential integration and maintenance issues. By using a hybrid approach from a single vendor, users benefit from the performance and reduced setup time of multidimensional technology, and from the openness and scalability of relational technology. Given this move towards hybrid solutions by the major database vendors, it seems that pure multidimensional technology products will be confined to niche markets in the very near future, and may even die out altogether.
Rule 10: Web Architecture
The final rule relates to how users get access to a CPM system. CPM systems employ a web architecture that allows simple, intuitive access that is independent of machine, location or the technology being used. A web architecture means more than just accessing the system via a web browser. Technologies such as Microsoft's .Net initiative means that the same protocols used in the web can also be used for other user interfaces -- e.g., PDA devices, Spreadsheets -- bringing the benefits of a central database and 'anywhere, anytime' access to users in the most appropriate form.
The benefits of a web architecture are many and include eliminating the need to maintain end-user software. The web allows, perhaps for the first time, enterprise-wide applications that are easy to manage.
The architecture of an application is often hidden from end-user view, but it will have an impact on the maintenance of a CPM solution and the processes that can be effectively supported.
It is not sufficient for CPM systems to have most of the features discussed because the benefits of one feature may not be fully realized unless it is accompanied by others. For example, the Web by itself will not make a system easier to support unless it is also accompanied by a central database, full process support, and end-user analysis.
It is only by taking all 10 rules together that make true CPM applications easier to set up, maintain, and able to cope with continuous planning, budgeting, forecasting, financial consolidation, management reporting, and analysis.
In the next article (Part Five) we will look at how to evaluate CPM software vendors.
About the Author:
Profile of Michael Coveney
The Strategy Gap: Leveraging Technology to Execute Winning Strategies
Michael Coveney, Brian Hartlen, Dennis Ganster and David King
John Wiley & Sons, 224 pages, US$27.97 / £22.98 (Amazon.com)
In North, Central and South America, contact:
In the United Kingdom and in other countries, contact:
[Return to Index of Corporate Performance Management (CPM) -- Comshare® Incorporated]
[Return to Main Index] [Return to Home Page]
227 Fuller Street
Revised: April 10, 2003 TAF
© Copyright 2003 Michael Coveney / Comshare, All Rights Reserved