Memos Not Dashboards
Dashboards prioritize data visualization over data analysis. We want self-service data analysis to enable individual data scientists to scale their impact. Pretty visualizations are only useful if they help achieve that, and they usually don't.
But too often Dashboards are the only tool a data scientist can use to create a self-service data analysis at their organization. So they end up using a tool built for visualizing data to create self-service data analysis. It is just the wrong tool for the job.
Suppose you asked a data scientist at your organization to do a data analysis, and they sent you back a couple of graphs. Would never happen, right? You'd expect them to write a document where they explain exactly what they did and their findings. You want a memo! Self-service data analysis should be data analysis. It only de-couples the person who decides on the Premises from the person who decides what the Conclusion is given those Premises. It should not change the fundamental nature of the analysis.
Memo_er makes memos, the standard output of data science.