Graphed LogoGraphed
Login

How to Use What-If Analysis in Google Analytics

Cody Schneider•

Ever wish you could ask Google Analytics questions about the future? Questions like, "What if we increased our ad spend by 25% next month?" or "How would a 2% lift in conversion rate affect our annual revenue?" This type of forecasting is called what-if analysis, and it helps you turn historical data into a strategic roadmap. While Google Analytics doesn’t have a built-in "what-if" button, this article will walk you through exactly how to perform this powerful analysis by combining your GA4 data with the flexibility of a simple spreadsheet.

What Exactly is 'What-If' Analysis?

What-if analysis is the process of changing the value in one or more input variables to see how it affects a dependent outcome. It’s a simple but powerful way to forecast the potential results of your decisions before you actually commit to them. Instead of basing your strategy on gut feelings, you can model different scenarios to make informed, data-driven choices.

For marketers and business owners, this translates to answering crucial business questions like:

  • Impact of Traffic Changes: "If our new SEO strategy increases organic traffic by 30%, what is the likely impact on total leads?"

  • Budget Scenarios: "What happens to our customer acquisition cost if we shift $10,000 from Google Ads to Facebook Ads?"

  • Conversion Rate Optimization (CRO): "If our A/B test on the checkout page improves the conversion rate by just 1%, how much additional monthly revenue will we generate?"

By exploring these questions, you can set more realistic goals, allocate your budget more effectively, and better understand the levers that drive your business growth.

The First Step: Exporting Data from Google Analytics

Google Analytics is a treasure trove of historical performance data, but it’s designed to tell you what happened, not what could happen. To perform what-if analysis, you need to export the relevant data out of Google Analytics and into a tool that allows for flexible calculations and modeling, like Google Sheets or Microsoft Excel.

The beauty of this approach is its simplicity. You don’t need a complicated business intelligence platform to get started. A spreadsheet is the perfect environment for this kind of work. The specific report you export will depend on the question you’re trying to answer. For most marketing scenarios, you'll start with a report like the Traffic acquisition report in GA4.

Step-by-Step Guide: Your First 'What-If' Analysis

Let's walk through a common and practical example: Projecting the impact of increased website traffic on total revenue. Our core question is, "If we successfully increase our user traffic by 20% next quarter, what will our projected revenue be?"

Step 1: Get Your Baseline Data from GA4

First, we need to establish our current performance benchmarks. This data will be the foundation of our model.

  1. Navigate to your Google Analytics 4 property.

  2. In the left-hand navigation, go to Reports > Acquisition > Traffic acquisition.

  3. Set your date range. For a stable forecast, it's best to use a period that represents your typical performance, like the last 30 or 90 days.

  4. Make sure your report includes the essential metrics. We need:

    • Users

    • Sessions

    • Conversions (specifically your primary one, like 'purchase')

    • Total revenue

  5. Click the Share this report icon (top right corner) and select Download File > Download CSV. We'll use this file in the next step.

Pro-Tip: The default view groups data by Session default channel group. You can leave this as is if you want to model a general traffic increase, or change the primary dimension to "Session source / medium" for more granularity on specific campaigns.

Step 2: Set Up Your Google Sheet

Now, open a new Google Sheet and import the CSV you just downloaded. We'll organize this data to build our analysis model cleanly.

Your imported data might look a bit messy. For this analysis, we just need the summary row at the bottom that shows the totals. Create a clean section in your sheet with the following layout:

In the 'Current Metrics' column, fill in the total values from your GA4 export. Then, calculate your two core baseline metrics which will be key for building your model:

  1. Conversion Rate: This tells you what percentage of your users make a purchase. The formula is Total purchases / Total users.

  2. Average Revenue Per User (ARPU): This tells you how much each user is worth on average. The formula is Total revenue / Total users.

Using the example numbers, the formulas in Google Sheets would be:

For Conversion Rate (in cell B5):

=B3/B2

Remember to format this cell as a percentage.

For ARPU (in cell B6):

=B4/B2

Step 3: Build Your 'What-If' Model

This is where the magic happens. We'll use the 'Scenario Modeling' section to build formulas that automatically update based on one simple input variable.

  1. The Variable: The first input is our "what-if" lever. In cell D2, type in a target percentage for your user growth. Let's start with 20%.

  2. Projected Users: This calculates the new total number of users based on your projected growth. In cell D3, enter the formula:

=B2 * (1 + D2)

  1. Projected Revenue: Assuming your average revenue per user stays the same, you can now project total revenue. This is the most important output of our model. In cell D4, the formula is:

=D3 * B6

This formula multiplies your new, higher user count by your existing average revenue per user to predict the outcome.

  1. Revenue Uplift: A useful metric is your net revenue growth—you need to subtract last reporting period's revenue from your forecasted revenue. It is an extremely easy way to conceptualize the magnitude of your initiatives:

=D4 - B4

Step 4: Play with the Scenarios!

You’ve built your model! Now you can easily test different scenarios. Simply change the percentage value in the "Projected User Growth" cell (D2). What happens if traffic only increases by 10%? What about an ambitious 50%? Your "Projected Revenue" and "Revenue Uplift" immediately and automatically update, giving you instant forecasts for each scenario.

This simple model allows for quick, clear conversations about goals. Now when someone asks, "What do we need to do to hit $300,000 in revenue a quarter?" you can plug numbers into your model and respond, "Based on our current performance, we’d need to increase user traffic by approximately 50% to hit that target, which is a massive goal."

More Practical 'What-If' Scenarios

The framework above can be adapted to answer all sorts of strategic questions.

Scenario 1: Measuring CRO Impact

  • Question: "If we increase our site's overall conversion rate from 2.5% to 3%, what is the potential return?"

  • How to model it: Tweak the user-based model to be built from the conversion event metric instead. With a forecasted user base, a revenue model can be developed. Build logic on top of last report's performance that includes some factor or range, for instance +/-10%. This allows building probabilistic models based on real-life scenarios.

Scenario 2: Understanding Ad Spend Efficiency

  • Question: "Our cost-per-click on Google Ads is $2. If we can get that down to $1.75 through optimization but with our paid users and purchase rate the same, how much bottom line can we expect to add?"

  • How to model it: Use a similar architecture as above. Use your new lower CPC as a driver for the scenario and observe the effects. Including margin in your models for each scenario provides valuable insights into potential ROI and profitability.

Pro Tips for Better What-If Models

  • Isolate Variables: When starting, change only one variable at a time. This helps clearly see its impact without confounding effects.

  • Pressure-Test Your Assumptions: Ensure your assumptions are realistic. If ARPU is projected to stay constant but external factors suggest it might decline, adjust accordingly.

  • Use Historical Data: Base your projections on actual growth trends rather than overly optimistic estimates.

  • Remember It is A Forecast: These models produce projections not certainties. Use them as planning tools, not crystal balls.

Final Thoughts

Diving into your business performance and understanding what drives it is incredibly important. By exporting your Google Analytics data and using a spreadsheet for what-if forecasting of different scenarios you want to try, you can help de-risk potential capital expenses. Modeling business performance makes decision-making easier for all stakeholders in a marketing team. Setting realistic expectations around your paid spend can help your agency with client relations, or in an in-house context, help avoid overspending inefficient budgets or managing inventory.

These exercises can take a bit of manual setup, especially because models in the business world get pretty complex and involve exporting from multiple platforms—think Google Analytics, Shopify, and Facebook Ads files all mixed into one workbook. To shortcut that process, we built Graphed. After you connect your data sources just once, you can ask those "what-if" questions using plain English. Instead of exporting CSVs and typing functions, you can just ask, "Show me our projected revenue if traffic from Facebook went up 30%" and get an instant visual answer that's already connected to your live data. This helps you get straight to the strategic insights and saves hours of exporting data and building a spreadsheet with potential vulnerabilities of human error, no matter what.