{"id":8200,"date":"2021-04-07T21:34:44","date_gmt":"2021-04-07T21:34:44","guid":{"rendered":"https:\/\/wealthrevelation.com\/data-science\/2021\/04\/07\/how-to-make-sure-your-analysis-actually-gets-used\/"},"modified":"2021-04-07T21:34:44","modified_gmt":"2021-04-07T21:34:44","slug":"how-to-make-sure-your-analysis-actually-gets-used","status":"publish","type":"post","link":"https:\/\/wealthrevelation.com\/data-science\/2021\/04\/07\/how-to-make-sure-your-analysis-actually-gets-used\/","title":{"rendered":"How to Make Sure Your Analysis Actually Gets Used"},"content":{"rendered":"
\n <\/p>\n

By Taylor Count<\/a>, Head of Data at Count<\/b>.<\/p>\n

\"\"<\/p>\n

Photo by\u00a0Austin Neill<\/a>\u00a0on\u00a0Unsplash<\/a>.<\/em><\/p>\n

\u00a0<\/p>\n

Beware of Pirates<\/h3>\n

\u00a0<\/p>\n

One of the most universally demoralizing experiences is to see the results of your hard work go unseen, unappreciated, and unused. In the world of data, that is something we experience all too often. Take the following hypothetical situation:<\/p>\n

    \n
  1. Jim submits a request to the data team for a deep-dive analysis for a client presentation the following week.<\/li>\n
  2. You and Jim spend all week working on the analysis, working closely to make sure he has the right visuals and feels confident presenting the findings.<\/li>\n
  3. The day of the presentation arrives, and not a word from Jim. That\u2019s odd.<\/li>\n
  4. When you finally track him down, he tells you he \u201cdidn\u2019t end up using the charts after all.\u201d \u201cThey just would\u2019ve confused them,\u201d he adds in a conciliatory tone.<\/li>\n
  5. You are fuming. A whole week wasted. Another decision was made without data there to back it up. Why did he even ask in the first place?<\/li>\n<\/ol>\n

    I like to call these requestors\u00a0pirates<\/strong>\u00a0because they steal my\u00a0time. Unfortunately, there will always be pirates, but there are ways we can learn to avoid them or at least cope with their existence. Here is a list of tips to make sure your analysis gets the credit it deserves, assembled from my own experience, academic research, and industry best practices.<\/p>\n

    \u00a0<\/p>\n

    1. Ditch the Data Request Forms<\/h3>\n

    \u00a0<\/p>\n

    \n

    We must be consultants, not hired hands.<\/em><\/p>\n<\/blockquote>\n

    Most data teams have a request portal that they use to triage and assign data requests that come from the business. These portals are designed to make it easier for the business and data teams to work together; business users type out exactly what they want, and the data team just makes it happen.<\/p>\n

    Unfortunately, as we saw from Jim, it\u2019s not that simple. Many business users come to the data team with a chart already in mind, including what the numbers on that chart should show.<\/p>\n

    At this point, we\u2019re already doomed. If the data doesn\u2019t match the story the requestor wants or is a bit nuanced, then they\u2019ll never use this analysis.\u00a0We need to know the problem they\u2019re trying to solve.<\/strong><\/p>\n

    As data professionals, we know the data and the statistical methods better than anyone and can advise on the best approach to using the data to answer the question at hand. The business context in partnership with our data expertise can combine to create analysis far more impactful than what we could produce individually.<\/p>\n

    In short, we must be consultants, not hired hands.<\/p>\n

    \u00a0<\/p>\n

    2. Numbers Never Walk Alone<\/h3>\n

    \u00a0<\/p>\n

    \n

    A chart alone cannot possibly convey everything, and that kind of thinking inhibits our ability to influence the business with our work.<\/em><\/p>\n<\/blockquote>\n

    Often we\u2019re expected to send over a single chart or dashboard as a completed request. These are nearly impossible for the business user to interpret without a 1:1 explanation.<\/p>\n

    We\u2019ve been told that data can speak for itself, that a well-crafted chart can communicate all its nuances on its own. This is simply not true. A chart alone cannot possibly convey everything, and that kind of thinking inhibits our ability to influence the business with our work.<\/p>\n

    <\/p>\n

    You can\u2019t rely on charts to convey insights alone. Make use of text to explain your work. Source:\u00a0The Best Player Never to Win a Title<\/a>\u00a0by\u00a0count.co<\/a>.<\/em><\/p>\n

    When sharing any analysis, I try to always include the following information:<\/p>\n