PROGRESS MADE, WORK TO BE DONE June Dershewitz unpacks results to showcase not only collective progress but also what needs to be improved and even so thoughts on how to get it done.
The data and analytics organization must be able to communicate with the business in a way that they can understand. And if they’re If data leaders are coming to business leaders with deep technical details - such as, here’s the algorithm we use in our new predictive model - it may not be the best way to communicate with people in a way that will convey value. When I see that “most analytics and data science issues at my organization are caused by data quality,” while I empathize with my colleagues I feel I should point out that the “data quality” issue might
“It’s clear in these results that we still have a lot of work to do to address the potential that we see for using data throughout our organizations.” be the symptom of another problem. For instance, if as a business, there is a certain kind of data that carries a lot of risk with it, and if you have a data quality problem in that data set- that’s extremely dangerous to your company. You’re going to invest in making sure that that data set is high quality and that issue is solved in that use case. I’ve heard stories long ago about companies recognizing that fact. Organizations should be finding risk areas and standing up teams to ensure talent, technology and processes are in place to get those data sets to be as high quality as possible. You can’t do that everywhere. You just need to prioritize and decide which areas carry the highest risk and address them. So to do that for, ”issues caused by data quality,” the question is what’s the impact? Are they small things that are inconsequential or do they carry a lot
Powered by FlippingBook