|
There is always a question on how much we should outsource our BI capabilities. Typically organizations tend to out-source most of their BI work. This includes most of the IT work as well as business analysis work. While the out-sourcing of the IT work follows a conventional model, to ensure that one is able to become skill-independent on varied technology platforms.
For the BI modeling and analysis, we recommend building internal capability. The key reason is that apart from few IT related rules and techniques; this is purely a business subject. It needs a significant business domain knowledge and understanding of business information needs. Just like you will have the users in your organizations, who will be defining the detailed functional specifications for your transaction systems, the same is needed for BI platforms as well. We have seen in some organizations, the entire business analysis and modeling piece is also outsourced. In that scenario, the role of the client organization stays limited to stating the business requirements. Few reasons for this are:
- The business analysis and modeling skills are difficult to get.
- An organization will take time to build these skills and BI initiative is not able to wait till the internal skills reach a level of maturity.
- Vendors are stated to have comprehensive modeling and analysis skills, borne out of their experience of multiple implementations.
All of these statements are largely true. At the same time, one can hugely benefit from having these skills internally as they end-up being most knowledgeable resources on business thinking and organization-specific details. The modeling and analysis skills are largely technology-independent and will be least impacted due to the BI platform that you are using. This knowledge is not only important for BI initiative implementation, but also maximizing the usage of the information post-implementation.
Here are few tricks that you can deploy to make the best out of the benefits and the constraints:
- Consider BI as an ongoing agenda: As we have mentioned in Enterprise level big-bang data-warehouse is a pipe-dream, a BI capabilities of an organization evolve over time. This has just one exception and that it the selection of plumbing or core elements (ETL, Data Warehouse, OLAP and Metadata...). These core components will be generally decided for once and they will be vehicle to carry your entire BI payload. Apart from this exception, typically, BI agenda is met through multiple projects over a period of time. One can rely on the vendor for first few initiatives, and use them as an opportunity to train your own staff.
- Use the industry knowledge of the vendor, while sustaining your own skills: While you train your own staff of modeling and analysis, it’s not an either-or game. One should allocate some money to pick-up the industry knowledge and best practices from the vendor.
- Let the internal resources play the role of reviewer: If organization does not want to retain a team of modelers and analysts, for doing the job, they can have a smaller number playing the role of reviewers. This means that Vendor does the modeling and analysis and this internal super-skilled team will review the vendors work, get it whetted by the business owners and represent organizational sign-off.
- Intensive class-room training and coaching: Though there are not too many organizations, which focus on training your staff on modeling and analysis, one can try to find one. One can also ask the vendors to take the training courses.
- Get BI analysis and modeling capability in one domain first (like customer analysis and sales analysis...), so to focus your efforts.
|
