Speaker 1: We've talked a lot about what data is and how it plays into decision-making. So what do we know already? Well, we know that data is a collection of facts. We also know that data analysis reveals important patterns and insights about that data. And finally, we know that data analysis can help us make more informed decisions. Now we'll look at how data plays into the decision-making process and take a quick look at differences between data-driven and data-inspired decisions. Let's look at a real-life example. Think about the last time you searched restaurants near me and sorted the results by rating to help you decide which one looked best. That was a decision you made using data. Businesses and other organizations use data to make better decisions all the time. There's two ways they can do this, with data-driven or data-inspired decision-making. We'll talk more about data-inspired decision-making later on, but here's a quick definition for now. Data-inspired decision-making explores different data sources to find out what they have in common. Here at Google, we use data every single day, in very surprising ways too. For example, we use data to help cut back on the amount of energy spent cooling our data centers. After analyzing years of data collected with artificial intelligence, we were able to make decisions that help reduce the energy we use to cool our data centers by over 40%. Google's People Operation Team also uses data to improve how we hire new Googlers and how we get them started on the right foot. We wanted to make sure we weren't passing over any talented applicants and that we made their transition into their new roles as smooth as possible. So after analyzing data on applications, interviews, and new hire orientation processes, we started using an algorithm. An algorithm is a process or set of rules to be followed for a specific task. With this algorithm, we reviewed applicants that didn't pass the initial screening process to find great candidates. Data also helped us determine the ideal number of interviews that lead to the best possible hiring decisions. And we've created new onboarding agendas to help new employees get started at their new jobs. Data is everywhere. Today, we create so much data that scientists estimate 90% of the world's data has been created in just the last few years. Think of the potential here. The more data we have, the bigger the problems we can solve and the more powerful our solution can be. But responsibly gathering data is only part of the process. We also have to turn data into knowledge that helps us make better solutions. I'm going to let fellow Googler Ed talk more about that.

Speaker 2: Just having tons of data isn't enough. We have to do something meaningful with it. Data in itself provides little value. To quote Jack Dorsey, the founder of Twitter and Square, every single action that we do in this world is triggering off some amount of data. And most of that data is meaningless until someone adds some interpretation of it, or someone adds a narrative around it. Data is straightforward. Facts collected together. Values that describe something. Individual data points become more useful when they're collected and structured. But they're still somewhat meaningless by themselves. We need to interpret data to turn it into information. Look at Michael Phelps' time in a 200-meter individual medley swimming race. One minute, 54 seconds doesn't tell us much. When we compare it to his competitors' times in the race, however, we can see that Michael came in first place and won the gold medal. Our analysis took data, in this case a list of Michael's races and times, and turned it into information by comparing it with other data. Context is important. We needed to know that this race was an Olympic final and not some other random race to determine that this was a gold medal finish. But this still isn't knowledge. When we consume information, understand it, and apply it, that's when data is most useful. In other words, Michael Phelps is a fast swimmer.

Speaker 1: It's pretty cool how we can turn data into knowledge that helps us in all kinds of ways, whether it's finding the perfect restaurant or making environmentally friendly changes. But keep in mind, there are limitations to data analytics. Sometimes we don't have access to all of the data we need, or data is measured differently across programs, which can make it difficult to find concrete examples. We'll cover these more in detail later on, but it's important to start thinking about them now. Now that you know how data drives decision-making, you know how key your role as a data analyst is to the business. Data is a powerful tool for decision-making, and you can help provide businesses with the information they need to solve problems and make new decisions. But before that, you'll need to learn a little more about the kinds of data you'll be working with and how to deal with it. So when it comes to decision-making, data is key, but we've also learned that there are a lot of different kinds of questions that data might help us answer, and these different questions make different kinds of data. There are two kinds of data that we'll talk about in this video, quantitative and qualitative. Quantitative data is all about the specific and objective measures of numerical facts. This can often be the what, how many, and how often about a problem. In other words, things you can measure, like how many commuters take the train to work every week. As a financial analyst, I work with a lot of quantitative data. I love the certainty and accuracy of numbers. On the other hand, qualitative data describes subjective or explanatory measures of qualities and characteristics, or things that can't be measured with numerical data, like your hair color. Qualitative data is great for helping us answer why questions. For example, why people might like a certain celebrity or snack food more than others. With quantitative data, we can see numbers visualized as charts or graphs. Qualitative data can then give us a more high-level understanding of why the numbers are the way they are. This is important because it helps us add context to a problem. As a data analyst, you'll be using both quantitative and qualitative analysis, depending on your business task. Reviews are a great example of this. Think about a time you used reviews to decide whether you wanted to buy something or go somewhere. These reviews might have told you how many people disliked that thing and why. Businesses read these reviews too, but they use the data in different ways. Let's look at an example of a business using data from customer reviews to see qualitative and quantitative data in action. Now, say a local ice cream shop has started using their online reviews to engage with their customers and build their brand. These reviews give the ice cream shop insight into their customers' experiences, which they can use to inform their decision making. The owner notices that their rating has been going down. He sees that lately his shop has been receiving more negative reviews. He wants to know why, so he starts asking questions. First are measurable questions. How many negative reviews are there? What's the average rating? How many of these reviews use the same keywords? These questions generate quantitative data, numerical results that help confirm their customers aren't satisfied. This data might lead them to ask different questions. Why are customers unsatisfied? How can we improve their experience? These are questions that lead to qualitative data. After looking through the reviews, the ice cream shop owner sees a pattern. 17 of negative reviews use the word frustrated. That's quantitative data. Now we can start collecting qualitative data by asking why this word is being repeated. He finds that customers are frustrated because the shop is running out of popular flavors before the end of the day. Knowing this, the ice cream shop can change its weekly order to make sure it has enough of what the customers want. With both quantitative and qualitative data, the ice cream shop owner was able to figure out his customers were unhappy and understand why. Having both types of data made it possible for him to make the right changes and improve his business. Now that you know the difference between quantitative and qualitative data, you know how to get different types of data by asking different questions. It's your job as a data detective to know which questions to ask to find the right solution. Then, you can start thinking about cool and creative ways to help stakeholders better understand the data. For example, interactive dashboards, which we'll learn about soon. Data is great, but if we can't communicate the story data is telling, it isn't useful to anyone. So we need ways to organize data that helps us turn it into information. There are all kinds of tools out there to help you visualize and share your data analysis with stakeholders. Here we'll talk about two kinds of data presentation tools, reports and dashboards. Reports and dashboards are both useful for data visualization, but there are pros and cons for each of them. A report is a static collection of data given to stakeholders periodically. A dashboard, on the other hand, monitors live incoming data. Let's talk about reports first. Reports are great for giving snapshots of high-level, historical data for an organization. For example, a finance firm's monthly sales. Reports come with a lot of benefits, too. They can be designed and sent out periodically, often on a weekly or monthly basis, as organized and easy-to-reference information. They're quick to design and easy to use, as long as you continually maintain them. And finally, because reports use static data, or data that doesn't change once it's been recorded, they reflect data that's already been cleaned and sorted. There are some downsides to keep in mind, too. Reports need regular maintenance and aren't very visually appealing. And because they aren't automatic or dynamic, reports don't show live, evolving data. For a live reflection of incoming data, you'll want to design a dashboard. Dashboards are great for a lot of reasons. They give your team more access to information being recorded. You can interact with data by playing with filters. And because they're dynamic, they have long-term value. If stakeholders need to continually access information, a dashboard can be more efficient than having to pull reports over and over, which is a big time-saver for you. And last but not least, they're just nice to look at. But dashboards do have some cons, too. For one thing, they take a lot of time to design and can actually be less efficient than reports if they're not used very often. And if the base table breaks at any point, they need a lot of maintenance to get back up and running again. Dashboards can sometimes overwhelm people with information, too. If you aren't used to looking through data on a dashboard, you might get lost in it. As a data analyst, you need to decide the best way to communicate information to your stakeholders. For example, what if your stakeholders are interested in the company's social media engagement? Would a monthly report that tells them the number of new followers for their page be useful? Or a dashboard that monitors live social media engagement across multiple platforms? Later on, you'll create your own reports and dashboards to practice using these tools. But for now, I want to show you what a report and a dashboard might look like. We'll start by using a tool we're already familiar with, spreadsheets. Let's see one way spreadsheet data could be visualized in a report. This spreadsheet has data set with order details from a wholesale company. That's a lot of information. From the headers, we can see different things recorded here, like the order date, the salesperson, the unit price, and revenue for each transaction recorded. It's all useful information, but a little hard to wrap your head around. We want a report that's easier to read. Let's say your stakeholders want a quick look at the revenue by salesperson. Using the data, you could make them a pivot table with a graph that shows that information. A pivot table is a data summarization tool that is used in data processing. Pivot tables are used to summarize, sort, reorganize, group, count, total, or average data stored in a database. It allows its users to transform columns into rows and rows into columns. We'll actually learn more about pivot tables later, but I'll show you one really quick. We'll select the data menu and click pivot table button. It can pull data from this table, so we can just press create and it'll pull up a new worksheet. Over here, it gives us the pivot table fields we can choose from. Let's select salesperson and revenue. And just like that, it made a chart for us. At this point, you can play around with how the graph looks, but the information is all there. All right, let's move on to dashboards. If you need a more dynamic way to share information with your stakeholders, dashboards are your friend. You might create something like this Tableau dashboard with interactive graphs that showcase multiple views of the data. With this, users can change location, date range, or any other aspect of the data they're viewing by clicking through different elements on the dashboard. Pretty cool, right? Later in this program, we'll look into how you can make your own data visualizations. We have a lot to learn before we get to that, but I hope this was an exciting first peek at the different visualization tools you'll be using as a data analyst. In one of our examples, the company wanted to see the sales revenue of each salesperson. That specific measurement of data is done using metrics. Now, I want to tell you a little bit more about the difference between data and metrics and how metrics can be used to turn data into useful information. A metric is a single quantifiable type of data that can be used for measurement. Think of it this way. Data starts as a collection of raw facts until we organize them into individual metrics that represent a single type of data. Metrics can also be combined into formulas that you can plug your numerical data into. In our earlier sales revenue example, all that data doesn't mean much unless we use that specific metric to organize it. So let's use revenue by individual salesperson as our metric. Now we can see whose sales brought in the highest revenue. Metrics usually involve simple math. Revenue, for example, is the number of sales multiplied by the sales price. Choosing the right metric is key. Data contains a lot of raw details about the problem we're exploring, but we need the right metrics to get the answers we're looking for. Different industries will use all kinds of metrics to measure things in a data set. Let's look at some more ways businesses in different industries use metrics so you can see how you might apply metrics to your collected data. Ever heard of ROI? Companies use this metric all the time. ROI, or return on investment, is essentially a formula designed using metrics that let a business know how well an investment is doing. The ROI is made up of two metrics, the net profit over a period of time and the cost of investment. By comparing these two metrics, profit and cost of investment, the company can analyze the data they have to see how well their investment is doing. This can then help them decide how to invest in the future and which investments to prioritize. We see metrics used in marketing too. For example, metrics can be used to help calculate customer retention rates or a company's ability to keep its customers over time. Customer retention rates can help the company compare the number of customers at the beginning and the end of a period to see their retention rates. This way, the company knows how successful their marketing strategies are and if they need to research new approaches to bring back more repeat customers. Different industries use all kinds of different metrics, but there's one thing they all have in common. They're all trying to meet a specific goal by measuring data. This metric goal is a measurable goal set by a company and evaluated using metrics. And just like there are a lot of possible metrics, there are lots of possible goals too. Maybe an organization wants to meet a certain number of monthly sales, or maybe a certain or maybe a certain percentage of repeat customers. By using metrics to focus on individual aspects of your data, you can start to see the story your data is telling. Metric goals and formulas are great ways to measure and understand data, but they're not the only ways. So far, you've learned a lot about how to think like a data analyst. We've explored a few different ways of thinking, and now I want to take that one step further by using a mathematical approach to problem solving. Mathematical thinking is a powerful skill you can use to help you solve problems and see new solutions. So let's take some time to talk about what mathematical thinking is and how you can start using it. Using a mathematical approach doesn't mean you have to suddenly become a math whiz. It means looking at a problem and logically breaking it down step by step so you can see the relationship of patterns in your data and use that to analyze your problem. This kind of thinking can also help you figure out the best tools for analysis because it lets you see the different aspects of a problem and choose the best logical approach. There's a lot of factors to consider when choosing the most helpful tool for your analysis. One way you could decide which tool to use is by the size of your data set. When working with data, you'll find that there's big and small data. Small data can be really small. These kinds of data tend to be made up of data sets concerned with specific metrics over a short, well-defined period of time, like how much water you drink in a day. Small data can be useful for making day-to-day decisions, like deciding to drink more water, but it doesn't have a huge impact on bigger frameworks like business operations. You might use spreadsheets to organize and analyze smaller data sets when you first start out. Big data, on the other hand, has larger, less specific data sets covering a longer period of time. They usually have to be broken down to be analyzed. Big data is useful for looking at large-scale questions and problems, and they help companies make big decisions. When you're working with data on this larger scale, you might switch to SQL. Let's look at an example of how a data analyst working in a hospital might use mathematical thinking to solve a problem with the right tools. The hospital might find that they are having a problem with over- or under-use of their beds. Based on that, the hospital could make bed optimization a goal. They want to make sure that beds are available to patients who need them, but not waste hospital resources like space or money on maintaining empty beds. Using mathematical thinking, you can break this problem down into a step-by-step process to help you find patterns in their data. There's a lot of variables in this scenario, but for now, let's keep it simple and focus on just a few key ones. There are metrics that are related to this problem that might show us pattern in the data. For example, maybe the number of beds open and the number of beds used over a period of time. There's actually already a formula for this. It's called the bed occupancy rate, and it's calculated using the total number of inpatient days and the total number of available beds over a given period of time. What we want to do now is take our key variables and see how their relationship to each other might show us patterns that can help the hospital make a decision. To do that, we have to choose the tool that makes sense for this task. Hospitals generate a lot of patient data over a long period of time, so logically, a tool that's capable of handing big data sets is a must. SQL is a great choice. In this case, you discover that the hospital always has unused beds. Knowing that, they can choose to get rid of some beds, which saves them space and money that they can use to buy and store protective equipment. By considering all of the individual parts of this problem logically, mathematical thinking helped us see new perspectives that led us to a solution. Well, that's it for now. Great job. You've covered a lot of material already. You've learned about how empowering data can be in decision making, the difference between quantitative and qualitative analysis, using reports and dashboards for data visualization, metrics, and using a mathematical approach to problem solving.

Speaker 3: Congratulations on finishing this video from the Google Data Analytics Certificate. Access the full experience, including job search help, and start to earn the official certificate by clicking the icon or the link in the description. Watch the next video in the course by clicking here. And subscribe to our channel for more from upcoming Google Career Certificates.