Continue in 2 seconds

Capitalizing the Data Warehouse

By
  • Sid Adelman, Suzan Dennis
Published
  • July 01 2005, 1:00am EDT

"Assets are probable future economic benefits obtained or controlled by a particular entity as a result of past transactions or events." - FASB Concept Statement No. 6

An asset is something that will have value in future periods, and the data warehouse (if it is any good) certainly fits into that definition. Companies capitalize their enterprise resource planning (ERP) systems and the guys with the green eyeshades don't object. Why not do the same for the data warehouse? A number of us have long maintained that the data warehouse (DW) is a real asset that has significant value to an organization -- some of us even believe that the DW will make the difference between a company living and dying. Because the DW provides value in future periods, it represents an intangible asset that should be capitalized on a company's balance sheet along with the tangible assets of cash, accounts receivable, inventory plant and machinery.

Current and Future Accounting Rules

The current financial accounting rules are heavily criticized as not reflecting the true value of IT assets such as those of the ERP systems and data warehouses. Under the current rules of the Financial Accounting Standards number 141 and 142, a data warehouse, as an intangible asset, could be capitalized at its fair value only when a company with an internally built data warehouse is acquired by another corporation. Thus, the CIO would be rewarded for managing a high value asset only when the data warehouse has been purchased.

Today, a company with an internally created data warehouse would be shown as an intangible asset only at its capitalized historical cost. This historical cost probably does not capture the fair value of the completed project, nor does it capture all the costs associated with the data warehouse. The current accounting system in the United States allows one company to show a high value on the data warehouse that was purchased and a second company to have a lower value on a data warehouse that was internally developed. Thus, the company is not able to reflect the fair value created by the successful implementation of an internally developed data warehouse.

The Financial Accounting Standards Board (FASB) is working on a new standard to address the comparability issue between companies that have purchased data warehouses and companies that have built a data warehouse internally. The first step is a proposal to disclose in the footnotes of the financial statements the quantitative value of the substantially built up intangibles in a business. The eventual goal is to book the intangible assets at their fair value to allow the financial statements of different entities to be comparable and, to address our interests, for companies to accurately reflect the asset value of a data warehouse.

The FASB is also integrating the U.S. accounting rules to the International Accounting Standards (IAS). More than 90 countries have moved to the IAS standards in the past year, including all of Europe. The IAS standard No. 39 requires the company to currently book the data warehouse at the fair value of the assets, and now the data warehouse must be carried on the company's books. In most situations, this rewards management for successful efforts in creating and maintaining a data warehouse.


What Did the DW Cost?

The expenses for any DW will vary widely. The cost will be dependent on the size of the database, the number of users, the complexity and quality of the source data, the software tools employed, the need for consultants and contractors, the capabilities of the team, and how well the system is supported and maintained.

It's necessary to understand how costs will be accounted for. Some costs will be expensed immediately and others amortized over the expected life of the system. Costs will appear in different accounts, and all these factors will become important when the actual total costs are tabulated and the DW is capitalized.

The accounting approach to classifying the costs as current expenses or capitalized as assets is a three characteristic definition. If the associated costs fail any of the characteristics, then the cost should be expensed under current accounting rules. The first characteristic is future benefits. The future benefits are defined as the capacity, singly or in combination with other assets, to contribute directly or indirectly to future net cash inflows. The second characteristic is whether the enterprise has control. Control is defined as the ability to both derive future benefits and to deny that ability to others. The third characteristic is that the cost accumulation is based on a past event or transaction that gives rise to the future benefit.

Hardware: For the data warehouse, you will need CPUs, disks, networks and workstations. Some vendors, such as Teradata, usually bundle the hardware along with the relational database management system (RDBMS), and the DW appliances bundle the hardware, operating system and RDBMS. If existing desktops and laptops are adequate to support end users, no additional costs should be charged; however, if upgrades or new machines are required, the additional costs should be assigned and depreciated over the expected life of the system. Three years is often used as the expected life, even though the system will probably last longer. The calculation is the cost to purchase or upgrade times the number of anticipated users. The cost of the hardware should always be capitalized.

Software: The data warehouse always needs an RDBMS. Most installations employ end-user access and analysis tools such as BusinessObjects, Cognos and MicroStrategy. Many installations choose an extract, transform and load (ETL) tool such as Informatica or DataStage rather than writing their own ETL code. Add-ons with the ETL tools could include additional costs for each different type of source file or target database. These tools are often priced based on the operating system and size of the machine. Additional tools are often needed for data cleansing and performance monitoring. Initial software costs should always be capitalized.

Internal Staff: The fully burdened rate (salary plus taxes, benefits, support costs, etc.) for the IT folks associated with the project should be included in the project cost. Business personnel are usually not included in calculations for personnel costs, but any help desk staff in the business organization should be. Include the fully burdened costs of the people on your project and capitalize these costs.

Consultants and Contractors: Consultants are engaged to help determine requirements, help plan the project, create the scope agreement, cost justify the project, help select the software, and establish the initial and long-term architectures. Consultants are typically more expensive than contractors but usually don't remain on projects as long. Contractors are brought in to supplement technical skills, specifically for software such as the database management system (DBMS). The cost for contractors will be dependent on how deficient the organization is in the required skills, how fast the organization needs the system implemented and how long it will take to transfer skills once the implementation is complete. The costs of the consultants and contractors for the initial implementation and for any major enhancements (not maintenance) should be capitalized.

Training: You could make a case for the value of training, the intellectual capital, the knowledge and the increased capability of those who went through the training becoming an asset that would have value in future periods; however, because of employee turnover and other factors, accountants are reticent to claim the value of training as an asset. Training costs should be expensed as they are incurred and should not be capitalized.

Training fails the control test for being a capitalized asset. There are past expenditures for training and probable future benefits, but without enforceable work contracts, there is no entity control of specifically trained labor.

Operations and System Administration: This is a grab bag of roles and costs including monitoring the system performance, executing backups, administering security, administering the meta data repository, dealing with the vendors and assigning charge-backs. The initial costs for operations and systems administration should be capitalized. The ongoing costs of both operations and system administration should not be capitalized unless there are major enhancements (because this represents day-to-day operations and not probable future benefits).

Data Quality Improvements: Every data warehouse implementation demands improvements in the quality of the data. The source files far too often contain data elements that are outside the valid values, data that is missing, incorrect data types and data that violates business rules (i.e., men getting hysterectomies, date of birth in the 22nd century, non-unique values for primary keys and incorrect data types). Most of this data must be cleansed before it's loaded into the data warehouse. Software is available to profile the data, which would provide information on most aspects of the quality of the data, and some software is specific to cleaning up names and addresses. Customer relationship management (CRM) data could now be de-duplicated, names and addresses would be corrected, deceased customers would be deleted and customers would have activity codes that are more correct.  This is an expensive process, but the result is data that is much more valuable to the organization. The value comes from fewer wasted mailings, fewer customer interactions that cannot result in sales, more customized and more appropriate marketing, and a much better image of the organization to the customer. The improvements in data quality should result in more sales, more cross-selling, greater profitability per customer interaction, and lower mailing and printing costs. For the DW, the improvements in data quality will mean far less checking and rechecking of results. The improved data quality will mean that the organization's strategic and tactical decisions will be supported by better information, and those making the decisions will be more likely to take action on results they trust. The value of the improved CRM data will be realized in future periods and should therefore be considered as an asset, and the following costs should be capitalized: data quality software purchase, consulting and internal personnel costs.

Meta Data: Capturing and maintaining meta data is an expensive effort, but it is generally recognized as a critical success factor for a data warehouse. Business meta data should include data definitions, domains (valid values), business rules, uniqueness, data source, security, timeliness and the owner of the data. Some meta data can be generated automatically from the modeling tools, the ETL tools and the BI tools; however, architecting how the meta data will be captured and maintained is not simple. It will require smart internal people or perhaps consulting help. In addition, an organization may choose to purchase meta data products. Meta data will be valuable to analysts, report developers and report recipients because the meta data will reduce their efforts to research and gather the data. It will also be valuable in giving them a better understanding of the results. CRM meta data capture codes would indicate a customer's value, previous purchases (retailing) and average balances (banking).  CRM for law enforcement would have codes for a person's violence level, an insurance company would have codes representing the likelihood of a customer discontinuing paying their premiums. Meta data will have value in future periods (as long as it's properly maintained) and should therefore be considered an asset, and the following costs should be capitalized: meta data software purchase, consulting and internal personnel costs.

Data Modeling: The effort to develop the data models to support the DW should be capitalized, but not the ongoing maintenance of the models.

Performance and Availability: Supporting the service level agreements (SLAs) for performance and availability will require new processes and procedures, possibly new monitoring software, possibly more hardware, and possibly some consulting and contracting effort. The costs for this area should be capitalized, but not any ongoing maintenance.


What Costs Can Be Capitalized and Disclosed?

Cost accumulations that have probable future benefit, controlled by the enterprise, and are based on a past transaction or event will be capitalized under the current accounting rules.

What should be disclosed in the footnotes of the financial statements about the intangible asset? Possible qualitative and quantitative disclosures about unrecognized intangibles are:

  • Major classes of intangibles assets and their characteristics.
  • Expenditures to develop and maintain.
  • Values of these assets.
  • Significant events that change the anticipated future benefits arising from intangible assets.

Is the Value of the DW Greater Than Its Cost?

The real value (the accountants may call it fair value) of the database can be significantly greater than the costs.  The new proposals for fair value measurements, for accounting purposes, create a means to fairly value an asset or liability.

If observable (quoted) market prices for identical or similar assets or liabilities are not available, the estimate of fair value should be determined based on the results of multiple other valuation techniques. Valuation techniques consistent with a market approach (for example, multiples based on prices from market transactions involving reasonably comparable items), an income approach (for example, a present value technique or option-pricing model), and, if applicable, a cost approach (current replacement or reproduction cost, adjusted to reflect the current condition of an asset) should be used whenever information necessary to apply those techniques is available. For a data warehouse, the market approach would use the purchase price of other acquired data warehouses. The income approach would use present value of an estimated cash stream attributed to the data warehouse. The cost approach would use an estimated recreation cost.

The objective of fair value measurement for accounting purposes is to estimate the single agreed-upon exchange price between willing parties in a transaction other than a forced liquidation or distress sale. The willing parties are all hypothetical marketplace participants (buyers and sellers) that have utility for the item being measured and that are willing and able to transact, having the legal and financial ability to do so. For a company's data warehouse, potential participants would be companies that are savvy and not affiliated with the company owning the data warehouse, with financial means and qualifications to buy the property and having full knowledge of the data warehouse. An outside and disinterested party, possibly a data warehouse consultant, could provide an analysis of what the data warehouse is worth.


Amortizing the DW

The data warehouse is amortized over the remaining useful life of the asset. The life of the data warehouse is not an item of certainty. Like many uncertain lives of assets such as oil deposits and the life of a utility plant, the life of a data warehouse could be determined using actuarial methods. The inputs into the calculation will be the life cycle of the products, customers, vendors and industry. However, it is more likely that an organization would use an amortization schedule that would conform to the norms they use for application package or ERP amortization.

What About the Costs of Expanding and Maintaining the Data Warehouse?

Expansion could include adding new source databases, adding additional data warehouse subject areas, new applications such as HR, adding whole new departments of users or expanding toward an enterprise data warehouse. The costs of expanding the data warehouse should be evaluated under the same capitalization tests as the original implementation. If the expansion does provide future benefits - and the example provided surely would (see: Totally Teen Togs below), then the costs should be capitalized.

Maintenance would include running the ETL processes, monitoring the quality of the data and ongoing data cleansing, monitoring and tuning the databases, supporting the security requirements, adding and deleting users, user support, installing new releases of software, and addressing software bugs and installing fixes. If the maintenance just supports the original implementation and adds no new value, these costs cannot be capitalized.


Selling the Concept of Fair Valuation to Your Organization

Why should senior management be interested in fairly valuing the DW or in capitalizing it as an asset? As mentioned, the current accounting rules actually require this capitalization, but in addition, by capitalizing the DW assets, corporations are much more likely to commit the resources and the management attention that are so critical to the success of a data warehouse. For corporations wanting to minimize their expenses in the current period, capitalizing allows them to amortize the value of the DW over future periods and not have to absorb all the expenses right away. By capitalizing and amortizing the cost of the DW, organizations are forced to capture the true costs of each major DW project and are thus able to more accurately compare the estimated with the actual costs. This then leads to a more exact and repeatable method of DW cost estimation with fewer projects that are not cost justified. The idea of capitalizing the DW should appeal to the CIO who wants to be able to get the budget needed for the DW projects and infrastructure, to the CFO who wants to fairly represent the corporation's expenses, profits and assets, and to the CEO who wants the company's stock price to include value that could otherwise be underestimated. 

Totally Teen Togs (T3)

Totally Teen Togs (T3) has 157 stores. Its target market is teenage girls between the ages of 13 and 17. This is a very fashion-oriented business and fashions are time and season dependent. T3 has four seasons: Spring, Summer, Fall (Back to School) and Winter (Holiday). Success in this business requires knowledge of the customer base, current and future fashions, inventory and what the suppliers can provide. Time delay in any of these areas means loss of a marketing and advertising advantage, lower sales and stale inventory.
T3 has a loyalty program with an estimated 68 percent of their clientele as members. Information about the members includes name, address, phone number, e-mail address, date of birth and the last two years of purchase history. In addition, there is a supplier database, an SKU database and an inventory database. T3 is considering implementing a DW primarily for marketing including market segmentation, promotions and cross-selling, and they are attempting to cost justify the effort. The CIO and the CFO are aware of the new accounting guidelines, and they would like to understand how the DW could be capitalized.

Data Quality

The quality of the data is substandard, especially the names, addresses, phone numbers and e-mail addresses. T3 knows that any marketing program will need a much-improved level of data quality. If the quality of this data is poor, members will be placed in the wrong marketing segment, they will be receiving inappropriate promotional material and the promotion will miss those who should be receiving it.

Meta Data

There are differences of opinion on what certain fields mean. The headings on the report are not clear as to their meaning. Sizes are sometimes expressed with U.S. size designations and sometimes with the European size. The marketing department is often reticent to take action because they neither trust nor fully understand what the reports are telling them. Clean, understandable and timely meta data is required for the marketing effort.

Performance and Availability

Marketing has tried to pull data for promotions, but the performance of the queries has been terrible. Each query has had to search the member database looking for a specific profile of SKUs. The design of the databases resulted in inefficient queries that took many hours. If the results were wrong or inappropriate, a new long-running query would have to be rerun. Because marketing always needs to be able to respond quickly, the existing database designs were suboptimal and performance was grossly inadequate. When the databases were being updated, it was not possible to run reports or queries. The marketing department had to wait. To support a marketing initiative, the databases would have to be redesigned.

Infrastructure

This new DW initiative will require an additional DBMS license, new hardware and new software for ETL and for queries and reports. Because T3 does not have staff skilled in the new tools nor in DW database design, contractors and consultants will be required for some period of time. A number of internal staff will be committed to the project.

Capitalizing These Costs

The CFO determined that all of the costs associated with data quality, meta data, performance and availability, and with building the infrastructure could be capitalized because T3 planned to have a continual and active promotional program coming out of the marketing department and that all of these initiatives would have value in future periods.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access