At Yellowfin, we understand that implementing a Business Intelligence (BI) solution, particularly for the first time, can be a daunting process. Getting it right from the beginning is critical for:
- Fostering a favorable business culture around reporting and analytics
- Securing ongoing executive backing
- Your career
We’ve compiled a shortlist of non-negotiable cultural and technical elements that must be practiced and adhered to to ensure success.
1. Identify potential issues before they become a reality: Develop a comprehensive roadmap to success, carefully sign-posting potential project stumbling blocks, so that appropriate solutions can be devised. Consider such issues as: Are your data sets complete and accurate? Do you have a master data management framework in place? Is your delivery team adequately staffed and appropriately skilled? Do we have sufficient hardware to support our BI goals?
2. Start the evaluation process early: Most BI vendors will assert that their product contains all the latest technological wizardry and addresses the specific needs of your industry or organization flawlessly. However, this will not be the case. Taking the time to conduct a thorough proof of concept will reveal a myriad of underlying differences. There is no such thing as “the best” BI tool – select the one that best suits your needs.
3. Let business users make the final purchase decision: BI is designed to increase the capacity of business people to meet and exceed their workplace goals. The BI solution that they perceive to be the most intuitive and helpful is the right one for your business.
4. Fast, easy and seamless – select a single-integrated BI application: Many BI tools are an amalgam of moving parts, creating a system that is difficult to integrate, manage and navigate. Reduce user resistance and implementation timeframes by selecting a single-integrated BI solution.
5. Ensure your chosen BI vendor will/can support you: Determine your vendor’s capability (and willingness) to deliver adequate services support and training throughout the implementation process. Obtain documented agreement on the level of services, support and training incorporated in the initial purchase order as well as expected timeframes to resolve issues.
6. Define user groups and requirements: Carefully divide your user community into definitive user groups (finance, marketing, sales, etc). Project requirements can be driven and defined according to the most pressing/immediate reporting needs of each defined group.
7. Under promise and over deliver: Create and follow a realistic delivery schedule – don’t try to do it all at once. Promising analytics for all and then delivering a handful of inactionable reports for a few will reduce user confidence in the usefulness of the BI project as well as the likelihood of ongoing executive sponsorship. To avoid this situation, develop reports for one user group at a time, say the sales team, moving onto the user group of next highest priority in a systematic fashion.
8. Secure support from data source owners: Your BI tool will combine data from a range of different data sources. For your BI project to deliver deep operational insight and the value of near real-time reporting, secure cooperation from all departmental or individual system data owners.
9. Incorporate representatives from defined user groups into the project delivery team: Let user demand drive product development to ensure that deliverables are always connected to business needs. Not only will this collaborative approach ensure the continued usefulness of your BI project, doing so will nurture a feeling of ownership within your BI community, which will help you achieve sustained user adoption.
10. Ignore testing at your peril: A thorough testing process prior to any, but particularly your first, phased rollout is essential. If the project is launched with glitches and slow query response times user drop-off will swiftly follow.
Where to next?
Can you think of some other critically important factors to consider when deploying BI for the first time?
