Data scientists are inseparable with modern business but they must be guided by more holistic strategy. A common fallacy is that marketers or managers are not aware of how to ask data from such scientists. There exist some better ways to ask the usual questions. It must be pondered what question to ask as the business analytics performed may have far-reaching repercussions. Questions posed must be specific and actionable. Also what kind of data needed is important. Cerner requires data from the US Department of Health and Human Services as it supplies health care solutions. Similarly, i-Medicare selects policies based on data from Centers for Medicare and Medicaid Services. Thus only relevant data must be taken out though it needs to be processes as not all desired information may be contained in raw data. Public data may also be useful in some circumstances. Private lending data has great vintage value due to apt comparisons to be made. Another question is how to obtain the data with what all being the sources. Facebook’s perfectly legal experiment to assess customer emotions was not well received by the public leading to a lawsuit even. Also with enormous data warehousing now done, a lot of it unfortunately is junk. In fact ninety five percent of the data worldwide is unstructured which makes it very difficult to process. Also a lot of it is inaccurate. Most importantly, data scientists and their implementers must make use of the KISS principle. The Model must never get too complicated.

Source:https://hbr.org/2016/11/better-questions-to-ask-your-data-scientists?referral=00563&cm_mmc=email-_-newsletter-_-daily_alert-_-alert_date&utm_source=newsletter_daily_alert&utm_medium=email&utm_campaign=alert_date&spMailingID=15983803&spUserID=OTY0OTMwNTk5NwS2&spJobID=904362140&spReportId=OTA0MzYyMTQwS0