Common challenges

- [Instructor] Data can be messy. I see it all the time in my consulting practice. An organization will want to conduct an analysis but then suddenly realize that they've not been collecting the data necessary for the project. This is where data governance comes into play. Data governance is a process of collecting data effectively and storing it in a way that you can use later for analysis. Data governance shows up differently for different sized organizations. First, you want to ensure that you're effectively tracking each transaction. This enables you to study what was purchased. Time of purchase, channel, location, and category. All these point points will become very valuable when it comes time to do analysis. Ironically, if you are a smaller business, then you probably won't have a hard time tracking your sales transactions. You can simply use a payment processing system like Stripe or Square, which will collect all your sales data automatically. All you need to do is download the raw data file into Excel, and then you're off to the races to do your analysis. If you're a larger business, you may be selling multiple types of product lines to various different types of customers. This can become very complex, very quickly. My recommendation to you, if you fall into this camp is to be very clear about how you track your data and potentially implement a customer relationship management tool like Salesforce. In this platform, you can track when purchases were made, but also data points like industry, buyer persona, days to close, and the channel that the lead came through. Next, you want to track demographic information about your customers if it's possible. If you're a small business, a simple survey asked upon checkout on your website will do the trick. This is effective, if you're an e-commerce store and not doing in-person sales. On the other hand, larger organizations will probably be using a CRM system and might have an in-person sales team. Just make sure your sales team is proactively tracking as much data as possible. Last be mindful of potential data quality issues. A common type of data quality issue is data duplication. This is when your sales are being counted twice. This could skew the final recommendations that you come up with after you're done with your analysis. Sometimes this can come from a mistake in the data collection process or when an analyst needs to join two data sources together. One way to help solve this challenge is to have an employee really own the role of data governance administrator, but also document how the data flows. Another type of data quality issue is corrupted data, or data that's incorrectly gathered. The most common form of this data quality issue is due to human error, resulting from manual data entry. For example, if your sales team is using Salesforce you'll want to have the lead source data point as a dropdown menu, instead of having the sales rep manually type in that data point. I've now walked you through some of the common challenges of working with data. Is your organization struggling with any of these issues?