What's the most important step in how organizations can get started with advanced analytics? Setting goals. The overarching goal for analytics within an organization is positive impact. This can be measured several ways depending on the nature of the organization, e.g., for-profit vs. nonprofit, and the desired objective(s), e.g., lowering cost, increasing revenue, expanding services. Regardless of the method of measurement, setting the right goal(s) is critical for the success of an analytics organization. Goals must align with the organization’s strategy and objectives to maximize the potential for positive impact.
Analytics is a team sport and, therefore, the goals of the team are important to understand. Within an organization, there will be people in different roles who all contribute to the success of any big data analytics project. These include:
- Analytics Producers: those who do the work of analyzing data and developing decision support systems that leverage analytics techniques for making better decisions.
- Analytics Consumers: the members of the organization that will be responsible for executing the results of the analytical work.
- Analytics Champions: the leaders in the organization that act as sponsors of projects. They typically lead a component of the organization that will benefit from the application of analytics.
- Analytics Enablers: the supporting parts of an organization, such as information technology, data stewards, and graphical user interface designers, who provide supporting functions for successful analytics projects.
Analytics enablers can be found in different roles in an organization. If a project is strategic (e.g., supply chain network redesign), then senior executives would need to support the results and execute the large and typically tough changes. If the project is operational (e.g., optimal scheduling of manufacturing tasks), then production supervisors would need to support the results from the new algorithm. Therefore, aligning the goals of the analytics work with objectives of the teams responsible for executing the solution is a critical step that should be completed as early as possible in the project. A change management plan should always be part of the overall project plan in order to minimize the risk of misaligned goals.
New data science approaches can often require a fundamentally different way of thinking and require different tasks to perform the new analytics-based process. In these cases, a broader change management plan should become part of the project. This could include new and/or changing job descriptions, role expectations, and organizational structures. These are understandably tough changes, but these changes must also be considered to minimize the risk of misaligned goals.
Given the significant relationship between change management and analytics goals, the role of executive champions and sponsors becomes extremely important. For larger initiatives, steering committees that bring multiple stakeholders with multiple perspectives into the same conversation also becomes very important. Analytics tools and models often clarify and expose critical tradeoffs that may negatively impact one business function to achieve a net positive result for the larger organization. Executive sponsorship and fora for these critical discussions help to validate goal alignment. There is no short cut to this process. Analytics leaders within an organization must act as continual stewards of these relationships to keep an analytics strategy on track.
For the analytics team itself, goals take on a slightly different form. While alignment with business impact should be the primary objective for individuals on the team, creating an environment where they can be successful is critical for the team to succeed. Relevant system access for data collection and the presence of sufficient computing resources are obvious starting points. Beyond that, giving the team flexibility to explore ideas makes for a work environment that is attractive and fulfilling. This can lead to breakthroughs on solving new problems, solving old problems in new ways, as well as increased job satisfaction reducing turnover for resources that are in high demand.
One often overlooked but critically important goal for analytics teams is getting the data scientists embedded into the business fabric, where they gain a deep understanding of how the business operates. Creating goals that ensure time spent working with and understanding specific roles or broader functional areas has large returns in terms of the relationships and trust that can be developed, as well as the quality of the business intelligence models (built to address real business cases). This holds true across the board: from simple databases to fully integrated, real-time machine learning solutions.
Finally, it may be worthwhile to establish a planning template for analytics projects within the organization.
This template can include the following goals:
- Identify new business processes that can realize value from analytics capabilities
- Define detailed functional designs for analytic models and supporting data structures
- Define detailed technical designs for analytic models and supporting data sources
- Assess the volume, velocity, and variety of an analytic subject area
- Build analytic models and data staging area to support incoming data sources
- Provide training and education for business analysts throughout the organization on the types of analytic models and data sets available
- Test and deploy new analytic models into production
Sustain current analytic models in production by refreshing the models to fit changes in business processes and/or underlying data models and retire analytic models that become obsolete.
Up Next
Where Should the Analytics Team Live Within the Organization?