Data Warehousing in a Cross-Company Setting
Data warehousing is usually a private affair. Even when external data about your competitors is part of your environment, it’s still your company’s data warehouse, built for your company’s benefit and use.
An interesting trend — one that’s surely noticeable at the executive boardroom level, primarily because those folks are steering corporations in this direction — is to have multi-company cooperation.
Two or three pharmaceutical companies might share the research-and-development expenses on a new generation of drug products, or two manufacturing companies might work together in a partnership to develop a product. A commercial bank and a brokerage institution might work together to offer jointly a series of financial products to the mass market, with the bank administering some products and the brokerage administering others.
Whatever the specifics of the industry and the situation, your company has a good chance of being involved in a multi-company partnership in which cross-company cooperation and sharing of information is a key part of success.
To that end, an interesting spin on the theme of data warehousing as a breaker of barriers is to have a multi-company data warehouse dedicated to more efficient analytical and information-delivery capabilities in support of a joint effort between your two companies.
As you might guess, a multi-company data warehouse is a slightly more complex creature than one dedicated to the support of a single company. Although you experience all the wonderful challenges of data extraction and transformation, tool selection, performance support, and other aspects of data warehousing, you also have to consider these issues:
Corporate standards: Two (or more) sets of corporate standards can affect how you deploy your data warehouse and its tools. For example, one company might be a Business Objects environment, and another might be a Cognos environment. Whose standards will you use, and what’s the effect on the other company’s users?
Proprietary and sensitive data: The business case for the multi-company data warehouse, and the accompanying functionality and data necessary to support the data warehouse’s business mission, might require that sales history information is made available for predictive sales forecasting.
Sales history information involves a breakdown by region, territory, and possibly even customer. What are the effects of revealing customer lists, and strengths and weaknesses in various regions, to a partner that’s also a competitor?
Security concerns: In addition to data security issues, any linkage between two environments can open up one environment to any security weaknesses in the other, such as unauthorized outside access and hacking.
Support costs: Which organization has primary support responsibility for the data warehouse? Does it bear the full burden of support costs, or will the two organizations share those costs? If the organizations share the costs, how do they calculate support costs, and how do they bill and pay those costs?
Development methodology: One organization might develop its data warehousing applications and environments by using Kimball techniques, but the other organization uses Inmon techniques. Who controls the development processes? Do individuals from one organization have to figure out and use the other’s methods and techniques?
Dispute resolution: Resolving any type of dispute in which parties are from different companies always proves an interesting challenge.
Ongoing enhancements: What cross-company management structure must you put in place to prioritize and approve enhancements to the environment?