How to Make a Waterfall Chart in Looker with AI

A waterfall chart tells a clear visual story about how you get from a starting number to an ending one, showing all the positive and negative contributors in between. It’s perfect for understanding a monthly P&L or a campaign budget breakdown. This article will show you exactly how to build a waterfall chart in Looker, step by step, and discuss how AI is making this kind of reporting much more accessible.

What is a Waterfall Chart? Answering the Important Questions

Before jumping into the builder, let's make sure we're on the same page about what a waterfall chart does and when it’s the right tool for the job. Misusing a chart type can create more confusion than clarity, a problem we definitely want to avoid.

So, what are we looking at?

Imagine your company's bank balance at the start of the month. Several things happen: you make sales (money in), you pay for ad spend (money out), you pay salaries (money out), and maybe get a refund from a vendor (money in). At the end of the month, you have a new balance.

A waterfall chart visualizes this exact story. It starts with an initial value, and then each subsequent bar "floats," showing how it either adds to or subtracts from the previous total. The final bar usually represents the grand total, connecting back to the baseline. The result looks like a series of cascading steps, hence the name "waterfall."

Why should you use a waterfall chart?

The main job of a waterfall chart is to explain the journey between two points, not just show the points themselves. It excels at breaking down a total into its individual components, which makes it incredibly useful for:

• Financial Analysis: The classic use case. You can quickly visualize a company's profit and loss statement, showing how revenue transforms into net income after accounting for all costs and expenses.

• Marketing Campaign ROI: Start with gross revenue from a campaign, then sequentially subtract ad spend, creator fees, and software costs to arrive at your net profit. Each cost becomes a negative bar, making it easy to spot where your budget is going.

• Sales Pipeline Changes: Show how your sales pipeline value changes over a quarter. Start with the initial pipeline value, then show additions from new leads, and subtractions from closed-lost deals, culminating in the final pipeline value.

• Inventory Tracking: Begin with your starting inventory, add new stock, subtract sold items and returns, and end with the final inventory count.

When to use a waterfall chart (and when to stick with something else)

Waterfall charts are fantastic for illustrating a process or a flow over time, where each step contributes to a cumulative result. They clearly answer the question, "What happened between our starting and ending numbers?"

However, they're not ideal for comparing a handful of unrelated, static categories. If you just want to compare website traffic from five different countries, a standard bar or column chart is simpler and more effective. If you wanted to show how your total traffic number is a sum of contributions from different channels, that's when a waterfall could work.

Building Your Waterfall Chart in Looker: A Step-by-Step Guide

Now for the fun part. Looker has a built-in visualization option for waterfall charts, which simplifies the process quite a bit compared to older methods that required complex table calculations. Let's build one together, using a simple monthly performance report as our example.

Step 1: Ensure Your Data is Correctly Structured

Your data is the foundation of your chart. For a Looker waterfall chart, the ideal structure is simple: you need at least one dimension and one measure.

• The Dimension: This will be the label for each column on the x-axis. It describes each step in the flow. For our example, let's call it "Metric."

• The Measure: This is the numerical value for each step, representing the change. We'll call ours "Amount." Outgoing expenses should have negative values.

Your data might look something like this table from your database:

Notice how expenses are entered as negative numbers - this is fundamental for the chart to function correctly. Also, we have explicitly defined "Starting Cash" and "Ending Cash" as line items. This gives you full control over how your chart starts and finishes.

Step 2: Start a New Explore and Select Your Fields

Navigate to your chosen Explore in Looker. In the field picker on the left, find and select your dimension ("Metric") and your measure ("Amount"). Click "Run" to pull the data into the table below.

At this point, you'll just have a simple data table. Our next step is to turn it into the visual we want.

Step 3: Choose the Waterfall Visualization

In the Visualization pane, click the ... menu on the far right. A list of chart types will appear. Scroll through and select Waterfall.

Instantly, Looker will try to render a waterfall chart. It might look a little strange at first, but don't worry. The real magic happens in the settings.

Step 4: Configure Your Waterfall Chart's Settings

Click the "Edit" gear icon in the top right of the visualization area to open the settings panel. Here, you'll find tabs for Plot, Series, and Y. Let's tackle the important ones.

Plot Tab:

• Up Color / Down Color: This is where you set the colors for your bars. Typically, positive changes (like revenue) are green, and negative changes (like costs) are red. Pick colors that are intuitive for your audience.

• Total Color: This setting is key. It lets you assign a different color to the bars you want to act as totals - like "Starting Cash" and "Ending Cash". This makes them stand out visually as anchors.

Series Tab:

This is arguably the most important section for dialing in your waterfall chart.

• Enter Name(s) of Total Bars: This field tells Looker which of your dimension values should be treated as a total. A total bar, unlike an incremental bar, will always connect to the 0-axis line. In our example, you would type Starting Cash, Ending Cash into this box. Looker will apply the "Total Color" you chose in the Plot tab to these specific bars.

• Show Value Labels: If you want the numerical value to appear on each bar, make sure this box is checked. It's almost always a good idea for clarity.

Step 5: Get Organized and Interpret Your Chart

Your final task is to make sure your columns appear in logical order. Your waterfall chart should tell a story from left to right. If your "Starting Cash" column is in the middle, it won't make sense. Go back to your data exploration, and for your dimension ("Metric"), click the gear icon and select "Sort..." to arrange the values in the correct sequence.

Once you've configured your settings and sorted your columns, you should have a clean, easy-to-read waterfall chart. Now you can clearly see how your starting cash was impacted by revenue and various expenses, leading to your final cash position. It tells a much richer story than just saying, "We ended the month with $61,000."

The Old Way vs. The AI-Powered Future

The process above is quite powerful once you know it. But getting there requires you to be familiar with Looker's interface, understand how your data is structured in the underlying tables, and manually click through menus to define the properties of your chart.

This is the reality for most traditional business intelligence tools. There's a steep learning curve involved just to get comfortable with the software, let alone mastery. For many, this complexity is a barrier that keeps them from getting the insights they need. If you don't already have deep technical data skills or the time to watch hours of tutorials on data modeling, you're pretty much stuck.

This is where AI is changing the landscape of data analytics completely. The bottleneck of technical knowledge is starting to disappear. Instead of needing to know which table contains "ad spend" and which configuration menu holds the "total color" setting, you can use natural language to simply state what you need. AI doesn't force you to learn the software, it understands you.

Imagine being able to just type, "Show me a waterfall chart breaking down last month's profit, starting with revenue and subtracting costs," and having the chart built for you in seconds. The AI would already understand the semantic layer of your data - it knows what "profit" means and how it relates to "revenue" and "costs" without you having to define it manually. This frees up your whole team to find answers and make better decisions, regardless of their technical background.

Final Thoughts

Waterfall charts are a superb tool for explaining the story behind a number, showing all the positive and negative steps that contributed to a final outcome. With Looker's native visualization, building one is a manageable process of structuring your data correctly and carefully configuring the chart's display settings.

But building reports and dashboards shouldn’t have to involve so many manual steps. At Graphed, we’ve built the data platform we always wished we had - one that lets you get answers by just asking questions in plain English. We turn hours of complex data analysis into a 30-second conversation, allowing you to create real-time, interactive dashboards without knowing any SQL or spending time inside a complex BI tool. Just describe the chart you want to see, and Graphed builds it for you instantly.