“Data scientists, according to interviews and expert estimates, spend from 50 percent to 80 percent of their time mired in this more mundane labor of collecting and preparing unruly digital data, before it can be explored for useful nuggets” reported New York Times.
One of the biggest difficulties with Big Data is that it is, well, very big. That wasn’t a problem when big data was only in the realm of IT and Data teams, whose primary role is to deal with data, make sense of it and make it actionable. However, in the last couple of years big data and analytics has been increasingly making its way into the hands of other business-led teams. According to a survey by the CEB advisory group, non-IT spending on technology equals 40 percent of the IT budget and 70 percent of CMOs, non-IT Directors and VPs are more than willing to run their own technology without help from the CIO.
However, unlike CIO’s whose primary role is to play with, and make sense of data and drive process efficiency, other business executives don’t necessarily have the resources, or the time to ‘wrangle’ with data.
CMOs are leading all other executives in technology spending, especially analytic tools as reflected in CEB's survey. Understably so, given that data from marketing automation tools and increasingly available customer data are not only copious but also have to be combined to provide any insight. Handling all that data, let alone combining them in a usable way, is a major pain point.
The CMO and her team deal with increasingly available sources of data from marketing automation tools, behavior on website, social media comments, time spent in physical stores, scanned point-of-sale data, weather reports, competitors’ pricing data, mobile app engagement, demographic information, etc. All this data has to not only be integrated but also be cleaned and deduplicated beforehand.
Maintaining an accurate profile of a customer is just as important as collecting the customer data - things change and data needs to be renewed and refreshed. A customer may have moved and not updated their address. Another customer may be using different emails to shop on your website but you need to be able to recognize that they are the same customer, so you don't send them marketing materials twice, especially if the messaging is different. Other times, even something as little as a typo can really hamper the accuracy of data. It's critical to make sure that customer data is always up-to-date and as accurate as possible.
Ensuring that all customer data is in one place, in the correct format, usable and accurate isn’t something that can be done manually. It would take a team of data scientists weeks to look at just one set of data. But data from all those sources need to be consolidated, de-duplicated, cleaned, and updated constantly so CMOs can get a real-time view of customer profiles. Instead of looking for a huge team of data analyst or scientist who takes week to process every set of data CMOs should look into getting software that can continuously accept, clean and analyze new data.
While it's okay for data scientists to spend 50 to 80 percent of their time cleaning up the data, CMOs and their team cannot dedicate that much time to data. The marketing team should be spending time creating content, and stories that help build a connection with customers instead of being mired in data.If you are looking to learn more about how you can automate your data analytics using machine-learning take a look at this book: