Use AI to simplify data and get insights | Google AI Professional Certificate

Let's say you're exploring ways to improve your company's services. So, you collect some survey data, pull reports on customer retention, and grab meeting notes from your colleagues. Before you know it, you have a lot to work with. But now, you've got a different kind of challenge. With all of these different sources, how do you make sense of it all? Not everyone has training on how to analyze data, and even for those that do, it isn't always easy to pull clear insights from raw data. But data analysis can be an incredibly valuable tool. Why? Because data can help you make the best and most informed decisions. And it can provide support for your point of view on any given issue. Now, with AI, even people whose eyes would normally glaze over at the sight of a spreadsheet, can take advantage of the power of data insights. Gemini can take messy data and tell you what it all means. It can even make sense of different types of data at the same time, like customer retention numbers and survey feedback. AI can make the most frustrating and time-consuming parts of data analysis just fade into the background. You can even create visualizations to communicate your findings. You can also do more complex analysis in new ways using everyday language with Gemini in Sheets. And you can accelerate data cleaning and uncover patterns between different data sources. It'll even troubleshoot complicated formulas for you. All this means that you can discover valuable insights in a fraction of the time it would normally take. Now, you might be thinking, if Gemini can do all this analysis for me, what do I have to do? Well, first, you still have to ask really good questions about the data. Sometimes it helps to start broad to understand the biggest trends and the most important conclusions, and then drill down into the details. Second, if you make a discovery or find an interesting insight, it's really up to you to understand the implications and make the right business decisions. And lastly, you'll always need to verify the findings to ensure their accuracy. Another thing to consider is data privacy. The golden rule is simple. Never input confidential data into a public AI tool. And always consult your company's policies on acceptable use of AI tools. Now, if you do need to work with sensitive data, you have two main options. First, see if your company has an enterprise-grade AI tool. The second option is to check the settings of the public tool that you're using. You can opt out of data collection or adjust your privacy controls as needed. Understanding these controls isn't just about privacy, it's about giving you the knowledge and confidence to analyze data with AI. By the end of this course, you'll be able to use Gemini to discover the most critical insights hidden in an ocean of data. So, let's dive in. Hey, I'm Owen from the Growth Google Marketing team. My job involves a lot of reporting, and a key challenge is knowing what metrics matter most. When you're getting ready to dig into the data, you have to keep in mind both the questions you want to answer and who your audience is. I'm going to show you how I use Gemini to level up my reporting prep. It can help me quickly identify key metrics, anticipate tough questions about performance, and walk me through how to export data from my data source. Let's take an example. Let's say I work at an e-commerce company as a marketing manager. I routinely report a lot of metrics related to my company's website. But now, I've got my first big presentation for my company's executives next week, and I'm wondering what metrics they might ask about. Let's see how Gemini could help me think through this. I'll start by giving some context. I work at an e-commerce apparel company. I'm preparing for an executive review of website performance. Our Q3 goal is to increase revenue. And now I'll provide my task. What are the top five metrics I should consider for my presentation? Okay, so Gemini gave us a few good options related to financial numbers, but I want my presentation to concentrate on the website's performance. And as a core metric for a website, I expect to get questions about our conversion rate. I want to make sure I'm ready for any questions they might ask about that. So, let's ask, what questions might my leadership team ask about conversion rate? And since I added the context in the first prompt, I actually don't have to repeat it here. Okay, so the team is probably going to ask me about the biggest drivers behind changes in conversion rate, and where the biggest drop-offs are in our checkout flow. I'll make a mental note to dig into that when I start my analysis. Now, if I want to report on my conversion rate over time or against other benchmarks, it'll be helpful to pull my metrics into a spreadsheet so I can build a simple table. My company uses Google Analytics to measure metrics like the number of website visitors we have and the number of orders that are submitted. I know I can get most of my information from there. How do I export website traffic data from Google Analytics so I can analyze it in a Google Sheet? I can see there's actually more than one way, and since this is just a one-off

analysis, I can go with the first method. One of the trickiest things about data analysis is knowing what questions to ask of a data. You really don't know what you don't know. And that's where Gemini can really come in handy as a thought partner. You can even try prompting Gemini, "What have I not thought of? " You can prompt Gemini to take on the persona of a co-worker, your manager, your executives, or even your ideal customer. Whether you're a full-time analyst or just do some occasional reporting, Gemini can be an excellent complement to your skill set. Next time you're digging into your data, try using Gemini to help think through different perspectives and strategies. How's it going, everyone? This is Owen from the Google Marketing team. As a marketer, I work with all kinds of info that I need to analyze, and that can take the form of numbers, text, or even images. Analyzing text-based or image-based info is a bit trickier, though. I have to find some way to get the data into a more structured format before I can even start my analysis. And that's where Gemini has become a really useful part of my toolkit. I can use Gemini to structure that data into a table, and then use Gemini in Google Sheets to quickly process and analyze it. Let's say I've got a retail business that sells pants, and I want to analyze some customer reviews to inform our product development team. I couldn't cleanly and copy and paste the reviews on the website, so I have them in a screenshot. But that's okay. Gemini can extract that data, organize it, and export it to Google Sheets. First, we'll upload the screenshot of our reviews to Gemini. Now, we'll ask Gemini to extract the information we need from the screenshots and organize it. Extract the reviews from the screenshot. If I take the time to specify the format now, it becomes easier to read and use later. I'll add, and turn them into a table with three columns, date, text, and number of stars. Each review shows a star rating based on how many stars are filled in. This is one of my favorite tricks. Gemini converted each of those image-based ratings into a number, which will enable me to do some basic calculations like summing or averaging. Gemini is integrated with Google Workspace, so you can export this table into Sheets. To follow along with me, you'll need access to Gemini in Google Workspace. Okay. We took our screenshot and put the data into Sheets. Now, I want to learn how my customers feel about my products. So first, let's name this column sentiment. Then, let's enter a formula to classify whether the review is positive, neutral, or negative. We can use the AI function in Google Sheets, where you can write a prompt that can use cells as inputs. I'll click here in row two and enter a formula starting with equals AI, followed by the task in parentheses. Analyze the customer feedback and classify it into positive, neutral, or negative. Then I'll write, comma B2, so it knows which cells to analyze. I'll tell it to classify with the words positive, neutral, and negative because those are the categories our product team uses. Gemini will fill in the cell with the sentiment of that review. So this customer said they were disappointed, so it makes sense that it says negative here. Now, I want to know the topic of the review as well. Is it the price or the size of the pants? So, let's click on cell E1 and name it category. Then we'll enter another AI formula. This time it's about the category. Again, I want to limit the categories like I did with the sentiment. Categorize the following comment. Select only one category from these options. Size, color, quality, customer service, shipping, price, or other. This first customer said the fit was awful, so it's a negative review on pant size. To make our formulas work across all the rows, we'll select these two cells, click and hold down at the bottom right of the cell, then drag down and let go at the bottom of the table to populate the other rows. It's really convenient to click and drag like this since I don't have to run one AI cell at a time. Being able to extract info from less structured sources like surveys or reviews or comments, and then work with it using natural language, unlocks a whole new world of analysis. Without Gemini, I'd have to learn complicated formulas or involve our data analytics team. Now, it's easier than ever to access new types of data to help make business decisions. Now that you're more familiar with what you can do in Gemini in Google Sheets, what kinds of messy data do you want to dig into? I'm Owen from the Growth Google Marketing team. My job often involves turning raw data into smart, actionable decisions. I'm going to show you how I use Gemini in Google Sheets to help me with my data analysis. It quickly finds top-performing channels, generates a clear narrative for that performance, and can create a final data visualization. Say I need to analyze the website performance and marketing performance from an e-commerce company.

access to Gemini in Google Workspace. In this case, I have a CSV file that I can import into a sheet. That way, I can view it in a spreadsheet and then analyze it in Gemini. Now, we have our data in sheets. We can prompt Gemini to get a quick summary on the website's revenue and traffic. I'll click the Gemini icon here and prompt, "This is data from my e-commerce website. Provide the total purchase revenue and total number of transactions. " The numbers Gemini's giving me for the total purchase revenue and total number of transactions looks correct. Now, I want to explore the data and figure out what's driving most of this revenue. I wonder if different channels have more revenue per session. "What is the average revenue per session based on channel, regardless of device category? " It looks like email has a much higher revenue per session than other channels, which is great. But, if I want to present this at work, I need an explanation for why email might be the best performer. To do this, we'll switch to the Gemini app and give it a persona. "You are an experienced data analyst. I'll also add some context. So, I want to specifically find out what drives email performance in e-commerce. For an e-commerce site, why might the average revenue per session from email be higher than from other traffic sources? " Ah, I hadn't thought about how email lets us personalize messages to the receiver. Now, I've got the numbers and a narrative, but the final thing I want to do is a data visualization that's easy to understand. And to do that, let's go back to sheets. For my data visualization, I want a bar chart that shows the average revenue per session by channel, which I know will clearly show email at the top. "Create a bar chart that shows average revenue per session by channel. " Now, as expected, the bar chart shows that email has the highest revenue per session. It also shows that organic and social media ads are the lowest performers. This might help inform our strategy on those channels going forward. Data analysis sometimes feels overwhelming, especially without a background in math or knowledge of complex spreadsheet formulas. But, Gemini's great at leveling the playing field by analyzing data, drawing insights, and making clear visualizations. Hi, I'm Carol and I'm a product manager on Gemini. Today, I want to show you how you can use Gemini to help achieve your business goals. I love collaborating with Gemini to brainstorm new ideas for products and to see the impact certain strategic decisions would have on my metrics. So, let's say I run an e-commerce business selling trousers. I have a goal where I want to maximize my profit this year. I also want to take a look at different strategic decisions I can make and see what kind of impact they'll have on my overall business. I can do that with Gemini in Canvas. To start, let's prompt. "I run an e- commerce business selling trousers. My main goal is to increase my profit this year. Acting as a business advisor, analyze my sales data and give me a list of the three most important changes I can make that would have the biggest impact on my goal. " I'll then go down and I'll add a CSV of my sales data and I'll select Gemini Canvas. So, Gemini is going to come up with specific recommendations for my business based on my data. It'll look through the data at different trends and different ways I can improve my profit. So, here's the output. It starts with an overview of my sales data, which is a good snapshot of my business, and then walks through three important changes to improve my profit. The first is tackling the baggy cargo return crisis. So, these are my pants with the highest return rate and it gives me recommendations for to how to drop that return rate and improve my profit. The next is prioritizing products like classic chinos for customer acquisition, given they have such a low return rate. And then finally, it recommends volume-based bundling for low volume categories, like creating a buy two save 15% offer. So, I can implement one of these, I can implement two of these, or I can implement all three of these. And my question now is, what would the impact of these be on my profit and my overall business? To do that, I can use Canvas to create a visualization where I can change my

assumptions for each of these and then look at the impact on my profit. To do that, I prompt, "Based on those insights and my data, create an interactive profit calculator for my trouser business. Let me adjust sliders for each change to show their effect. " Gemini's going to take these recommendations as well as that original data CSV I uploaded and create an interactive visualization of my business. So, you can see now it's coding up this web app, going through all the process of things like formatting and making it interactive. Um it's going to include sliders because that's what I've requested in my prompt, but you have a lot of uh ability to customize and um have different interactions based on what you think is most relevant for your business and your forecast. So, Gemini has taken the data from my first prompts and the recommendations from its last response to create an interactive profit simulator. I can now adjust the impact I think each of those recommendations will have. So, my return rate on my cargo pants is currently 36%. I can drop that down to 23% and see the impact on my profit, which would increase my profit by $1,000. I can then do things like increase my chino acquisition and the number of chinos I'm selling by 50%, which would result in 6K more profit. And then I could also do things like increase my bundling strategy, which would increase with those other changes a total of $15,000, giving me a projected net revenue of $128,000. So, this is really helpful to see the impact of multiple changes at once and to give you an easy way to adjust these assumptions. Let's say I want to take this and share it internally, but the styling doesn't really match my business. I can prompt Gemini to change the styling while leaving the core interactions the same. So, I can prompt something like, "Can you give this cal- culator more of a fashion theme? " And Gemini will take that, leave the core interactions the same, all of the assumptions and data the same, but it's going to change the theme to something that may feel closer to my business. You can also upload your own style guidelines if you have those, um or give it very specific uh styling recommendations in your prompt. Great. Gemini's taken that last output and given it a more fashion style. So, it's changed the font, it's changed some of the colors, but it's left the core interactions the same. So, I can still change my assumptions and see the overall impact on my business. I can also share this with anyone in my company. Um this will generate a link that anyone can open and have the same interactions. So, if you're collaborating with others internally, it makes it really easy to all be working off the same data and the same assumptions. This is one way you can collaborate with Gemini to achieve the goals you have for your business this year.