1.1 Introduce the Data Lifecycle - Video Tutorials & Practice Problems
Video duration:
5m
Play a video:
Video transcript
<v Narrator>All right, in this sub-lesson</v> we're gonna introduce you to the Data Lifecycle, where we're gonna examine all the people, process, technology, and data components that surround the management of a business process. Now, some examples of business processes are things like sales, inventory management, and payroll. So payroll has processes around it like time sheet entry and then invoicing. But a lot of it really starts around the time sheets that people enter and all the different tracking you're doing and all the downstream processes that end up coming out of it. So those are some example business processes. So pick one of those or a business process that you have in your organization and frame it around this data lifecycle. All right, so typically there's some type of technology that we have in our organization that is used to help manage our business process. So we may have spreadsheets, we may have CRM, we may have an ERP, all depending on how large our organization is, how complex the business process is. Then the day we have some type of technology here that's helping us manage this business process. Then we have data producers on the left-hand side. So these are people that are in our organization that help manage that business process. And then those people end up having some type of process around that, where they do data entry. So maybe there's data entry happening around inventory coming into your shipping doors. Maybe there's a process around entering sales into the point of sale systems as customers are buying things. So at the end of the day, we have data producers that are in our organization and they're using some type of process to enter data into our systems. And that's really what comes in next. So after that process is followed and hopefully we have a standard and consistent and ultimately simple process because when we have those things in place we typically find that organizations end up with higher quality data. So if the process is understood and is followed then the result in data that gets stored in our technologies is usually of a higher degree of quality. And you may have heard the term garbage in, garbage out. This is the end part. So we want to make this as high quality as we possibly can. All right. So on the other end once this process is being managed on a daily basis. So as we're selling things, as we're bringing inventory into organization is inventory is moving out the door. As people are logging hours and time sheets and things like that, which will eventually move out into payroll and perhaps invoicing depending on the type of organization you have. The next thing that happens is we have data consumers on the other side that say, Hey we'd like to how she measured this business process. How well are we doing in terms of sales? What are our inventory levels? Do we need to replenish our inventory? What are our time sheets looking like? Are we actually logging as many hours in a month as we thought we might are we on track to actually hit our revenue targets? So there's all kinds of business questions that come out of any one of these processes here. So the data consumers are typically the ones that are asking those questions. And sometimes the data producers and data consumers are the same people. And then the data consumers want to use some type of process around analytics and reporting to actually go ahead and measure that business process so we can see how well it's actually performing. <v ->Okay. So next we'll talk</v> about a general analytics process flow. And what this is is the general stages of answering a business question. So a lot of the business questions which I just mentioned a moment ago there's stages that people need to go through when actually answering that question. So it usually starts off with a question being asked to you as an analyst, and then you go through some type of process that looks like this. So step number one, when a business question is asked of you is trying to understand where the data might be in organization that actually supports that. So you have to go through some type of acquisition step. Maybe you understand that data, maybe you don't. The second step is to go, go ahead and actually try and understand what that data is telling us. Often, we see organizations bringing subject matter experts into the mix here because often they understand the source systems how the data's really being recorded from a timing perspective, what some of the business rules are around that data. Then to the day we go through some type of understand process. Then often we run into a phase of doubt. So we get the data we start working with and it doesn't quite seem to be adding up to what we'd expect. So there's some doubting that happens. And ultimately we end up going back and working with those subject matter experts to really understand what's happening. And at some point we overcome that doubt. And actually we have data that we're ready to work with. Next, what ends up happening is we go through some type of transformation process. So maybe that data's not quite in the form that we want for reporting, or maybe there are some quality issues that are found in that data. We want to go ahead and do some cleansing on that. So we typically find ourselves going through a transformation process. Then we go through, after our data's in a report ready state, and we start doing some analysis on it to ultimately try and answer that business question. Once we have our findings in place, then we're ready to go off and start presenting the findings of our business questions to our business users. So we go through some type of presentation process. And at the end of the day, we draw this out as a very linear process, but it's highly iterative. We're moving back and forth between these different circles all the time as we learn new things, new business questions are being asked and we learn different things about our data. All of a sudden find a quality issue that we may need to go back in our transformation step, for example, and fix. So this is not a linear process. We show you the linear process here from an, a learning perspective, but this is highly iterative. All right, so that brings us to the end of this sub lesson. Let's move on.