You're drowning in data points. How do you decide which trends to prioritize?

Finding a way to focus in on what goals are a priority and which data points tell the story can help you focus. Be careful not to look to confirm any biases going in. An example of this could be the effectiveness of outbound calling. You may find data to say that your team is making plenty of calls but not seeing results. If you stop here, you may miss data showing call length, call times, and call frequency when there is no contact that may show a lack of skill on the calls instead of calls themselves not being useful.

In the defined state, you need to ask what your end goal is. Are you aiming to reduce defects, increase product, increase sales, etc? A fishbone diagram would set up categories (machine, environment, etc) to assign the different data points. From there, we can focus on what category would help us the most in achieving our goal. Once focused on the category, let's say machine, for example, we can then aim to gather whatever data we have that relates to it. Everything else is put to the side. You'll most likely find certain trend(s) that will pop right out (downtime). Breakdowns, automation errors, etc. What needs the most attention or what could have the biggest impact if fixed?