I want to start tackling an elephant in the room when it comes to data analytics in business. When leadership decides to pull the trigger to stand up an analytics team, several challenges need to be addressed. Most importantly, leaders need to integrate analytics with other business functions to be most successful. Put simply, analytics is not a bolt-on business function. In this article, I’ll discuss the following:
- Characteristics of the most useful analytic projects
- Problems created by bolt-on analytics
- Best practice considerations for analytic integration
Characteristics of The Most Useful Analytic Projects
Data analytics covers a broad spectrum of activities, from simple description to complex predictive modeling. Sure, businesses survived before without analytics. The power of good analytics, however, lies in removing uncertainty, increasing efficiency, and making better decisions.
The most helpful analytics projects often share several common characteristics:
- Begin with the end in mind
- Are highly targeted
- Focus on policy-amenable processes and outcomes
- Are not developed as an afterthought
- Integrate analytics as part of the project team
Begin with The End in Mind
Every business function serves a purpose. You wouldn’t create a product development team without having an idea of what you want them to do. Data analytics, as a business function, is no different. During the preparation phase of creating an analytics team, leaders need to consider the specific use cases for analytics.
Are Highly Targeted
The more targeted your analytics projects is, the better chance you have of answering your questions. Organizations use focus groups and customer surveys to better understand how clients use their products. Similarly, it’s easier to identify healthcare processes that improve patient safety in a small number of hospitals rather than nationwide.
Focus on Policy Amenable Processes and Outcomes
Just because you can measure something, doesn’t make it useful for analysis. From an organizational standpoint, it’s best to focus on processes and outcomes you have some influence over. You might track some metrics that aren’t in your control to understand the context of your operations. You want your deep analytic efforts, however, to yield insights you can DO something about.
Are Not Developed as an Afterthought
Data analytics are best used to provide answers that help you make decisions. As your key decisions become clear, you should identify any assumptions or contextual factors that you don’t understand with certainty. These uncertainties become the high-value targets for your analytic inquiry. Learn by leveraging analytics before making decisions, rather than backing into justifications for your decisions after the fact.
Integrate Analytics as Part of The Team
Every project team works through phases that include planning and troubleshooting as the project evolves. Your analytic team should be part of this process to help tackle questions and provide insights along the way. Reaching out to analytics only when there are specific questions excludes them from the problem context. You may still get your question answered, but integrated analytics provides a much better answer.
Problems Created by Bolt-On Analytics
Organizations taking the plunge into analytics do so in a variety of ways. One of the more common ways is to add an analytics team – which may be one person – as a standalone business unit. The bolt-on approach makes sense when you want to isolate the spend and monitor work within the new unit. Bolt-on analytics, however, can result in many negative side effects as well. These include
- Resistance to change
- Turf wars
- Misaligned expectations
- Poor communication
- Minimal or negative ROI
Resistance to Change
Staff in existing units may choose not to engage with a bolt-on analytics unit. Existing units already have processes in place to accomplish their work, even if less efficient and narrower in capability. Staff will attempt to continue using their old processes saying, “That’s not how we’ve done it in the past.”
Turf Wars
In more extreme cases, existing staff may view bolt-on analytics as competition for their jobs. Resistance in existing units escalates to actively demonstrating any weakness or mistake made by the new analytics unit. Staff will rationalize their arguments about analytics saying, “They don’t know what they’re doing” or “They need to stay in their lane.”
Misaligned Expectations
Existing staff may misunderstand the capabilities and responsibilities of the new analytics unit, making unrealistic requests. Existing units may request analytics to generate results for every question and curiosity they think of, saying, “We just need [X].” Alternatively, existing units may minimize the scope of requests saying, “It shouldn’t be that hard for you to [X].”
Poor Communication
Leaders adding bolt-on analytic units may find resistance to change, turf wars, and misaligned expectations also come with poor communication. Existing staff may be passive-aggressive in their communication tactics, or actively avoid communication. Regardless, analytic units may find themselves at a disadvantage in planning and completing work efficiently.
Minimal or Negative ROI
When leaders add bolt-on analytics, the combination of possible negative side effects results in weaker effectiveness overall. The less that existing units engage with analytics, the less likely the organization is to receive a solid return on investment (ROI). In extreme cases, bolt-on analytics may generate substantial costs and negative ROI.
Best Practice Considerations for Analytic Integration
Leaders looking to effectively integrate analytics into their teams need to work through several considerations. Here are a list of six (6) of the key elements to think carefully about:
- Use cases
- Budgets and staffing
- Organizational structure
- Governance
- Analytic process requirements
- Cultural change management
Use Cases
The word analytics covers a broad spectrum of activities. Effective integration begins when leadership clearly defines the use cases for which analytics will be responsible. Which problems will analytics take on, which questions will they answer, and which clients will they work for?
Budges and Staffing
It almost goes without saying that you need a budget to create an analytics team. Look at your use cases to determine the number of people and levels of experience you’ll need. Then create the plan to source, recruit, hire, and integrate that team.
Organizational Structure
Where will your analytic unit sit in the organizational structure? Will it be centralized and report to leadership? Will it be decentralized and report to key business units? Leadership should draw out a new org chart to decide where and how to house analytics.
Governance
Leadership needs to decide on governance policies and procedures to put in place before creating an analytics team. Determine who will have access and control over data, collection systems, and reporting. Leaders also need to identify analytic governance and the processes for requesting analyses and managing workloads. Learn more about analytic governance here.
Analytic Process Requirements
Leaders need to consider how they want analytic projects structured and documented. Identify the key elements analysts need to document in every project for future reference. The organization also needs to develop a template for recording and reporting those elements efficiently.
Cultural Change Management
A final consideration leadership needs to consider is how they will manage cultural change when the analytic unit begins operations. Existing staff need to be prepared for a change in process, and to release some of their control. Additionally, leaders will need to mitigate any objections raised by existing staff to prevent the problems outlined with bolt-on analytics.
Conclusion
As more organizations decide to take the leap into building and integrating analytic units, integration challenges loom large. By taking a step back and developing a plan addressing key integration elements, everyone can enjoy a smooth integration.
The other option for leadership? Bolt-on analytics: a recipe for structural and cultural problems in the organization.
