“Fewer than 15% [of senior finance executives] are confident that they get helpful information for making decisions about the future.”
- Corporate Executive Board, March 2012
In addition to the fact that Chief Executives feel as though they lack data for good decision-making, over 40 percent of CFOs rate themselves as ineffective in their quest to measure and monitor their organization’s performance. Nonetheless, for decades, leaders in progressive organizations have been using data to optimize their organization’s performance.
In studying best-practice organizations over the past year, we boiled down the five critical success factors that organizations have used to maximize value derived out of big data.
1. Define business requirements: Unlike the promise of some commoditized technology solutions, implementing advanced analytics with big data is not a one-size-fits-all solution. Although there are commonalities amongst organizations within similar industries, similar to most other successful IT implementations, it is critical to start with well-defined, organization-specific business requirements before starting. The most effective business requirements will have well-defined out-come oriented performance measures.
2. Big Data entry point: To help get the most value out of an organization’s big data initiatives, it is recommended that organizations start with reasonably-sized pilot projects. While conducting the first pilot, organizations should develop a roadmap for long-term success. The approach of piloting big data initiatives help to increase the probability of success.
3. Plan to augment and iterate: When starting a big data initiative, best-practice organizations need to neither start from scratch nor start big, they often begin by building off of existing initiatives. The roadmap developed by organizations should include a number of pilots that when combined together, help to realize the organization’s long-term vision. This roadmap will likely include multiple:
- Dimensions within the organization, such as technology, process, and people;
- Business units within the organization, such as finance, IT, and customer service; and
- Measurable business outcomes realized by each participating business unit.
4. Identify gaps: Given the scope of the roadmap and approach in using pilots, it is recommended that each pilot has measurable performance outcomes, identifying and filling gaps throughout the organization. Any one pilot doesn’t need to attempt to fill a large number of gaps. Pilots should be chosen as a portfolio that when combined together identify and fill multiple gaps throughout the organization. Furthermore, pilots can be sequenced along two dimensions:
- Critical path: For some initiatives, there is a natural additive sequence where one pilot is in the critical path of another.
- Return on investment: In determining the sequence of pilots, initiatives with the lowest cost and highest return should be sequenced before higher-cost pilots.
5. Iterate: Similar to most successful system implementations, the successful big data implementations tend to utilize an Agile-like implementation methodology that identifies gaps in the initial requirements and iterates as it fills these gaps. This enables early successes while obtaining buy-in from stakeholders throughout the organization. The gaps are filled via iterations as the platform and organization develops capabilities.
When you put both of the above major take-aways together, what it really says is that “big data” doesn’t have to be so big that it can’t be implemented in reasonable, bite-sized iterations. Big data should be implemented with one eye on enabling short-term value and the other eye operationalizing for long-term returns.