How do you assess the value of your data analysis work?

As analysts our goals and metrics can come from a variety of sources. They may come from a project manager or executive, or you may have to identify and define them yourself. I find it helpful to start by contextualizing the situation/project, identifying the stakeholders, and specifically identifying who is expected to affect the outcome. By identifying those parties and communicating with them, you will have a good opportunity to validate the legitimacy of the initial goals and metrics, adjust as necessary, and brainstorm the output of your project.

Something I noticed when working with people aspiring to become Data Analysts is that a lot of them tend to start a project or analysis with their expectations already set in stone. The nature of this job is pragmatic and objective, but one of the skills that makes a great analyst is being able to notice a subtle impact on our metrics that is not as straightforward as one might initially think. Just don't let your entire project be already defined on the metrics and goals stage, otherwise you might find yourself adapting your analysis trying to meet perfectly your initial expectations.

Having a consistent, systematic approach can help others around you learn at which stage of data analysis you’re and can anticipate turn around time on the request better. We often think everyone will understand how complicated or easy the task is, but that’s not the case. So having a systematic process will help you keep the quality of the work high and also make you a better communicator.

Cross-check your query, data output, calculations, graphs, and tables time & again before sharing it with anyone. I have found that the more detailed oriented I am in an analysis, the lesser the chance others would pinpoint any potential mistakes.

When working in Data Visualization projects I always try to look for samples or other projects people share on communities. It is a great way of comparing your work to check if there are perspectives you could also include in your analysis.

Nishtha's comment is right on. As a program manager who bridged the analytic and healthcare provider worlds, I can attest that data errors cause the customer to question the data integrity as a whole.

Whether it's in an applied environment where analysis informs our decisions, or in a removed environment where analysis informs the decisions of another party, the value of our efforts get the best results from the presence of the human element. Additionally, because every analyst has their own personal approach and experience base to pull from, it's important to seek out feedback to close knowledge gaps we might have as individuals and provide new perspectives from which to apply analysis. For instance, a colleague pointed out a particular aspect of my analysis that I hadn't considered, and now I have the perspective that inventory problems share aspects of fluid dynamics problems and can be solved for using similar methods.

This is a great step to take if you want to zoom out and determine the impact of your data career as a whole. Measuring the numerical value of your work takes additional effort, but it's necessary to know how successful our projects were.

Always gather user feedback from what actions (or lack of actions) came from your analysis. You can do this with a simple email or brief followup call. Assess whether the actions drove the metrics in the right direction, what roadblocks came up that were unforeseen, where there were blind spots that data could better fill in next time, etc. This all helps you understand how data and the business interact and what levers can be pulled with data to create change.

The most valuable data analysis project is one that can be followed by concrete actions which outcomes can be measured objectively in relation to strategic objectives. Therefore, a data analysis report should ideally be concluded with recommendations for such actions, and guidelines for the measurement of their outcomes.

As a senior program manager in healthcare quality analytics, we were all encouraged complete a Lean Six Sigma Green Belt and look at our metrics and data analysis from a Lean Six Sigma lens. The Lean Six Sigma discipline is a way of incorporating continual process improvement into everyday work life. The physicians and healthcare leaders did not always know what analytics they needed (or thought they needed a dashboard for everything!) so having this Six Sigma perspective was very helpful in providing the best analytics and dashboards so that our leaders could make the strategic decisions and could monitor progress toward their goals.