You can now analyze complex spreadsheets like a pro, even if you’ve never taken a data course in your life. I often get handed data files with zero context and feel a bit lost on where to even begin. I just stumbled upon this incredible video from an AI professional that lays out a simple, repeatable framework for turning ChatGPT into your personal data analyst, and it’s brilliant.
The mind behind it calls this method the DIG framework, which stands for Description, Introspection, and Goal Setting. It’s basically a simplified version of a professional technique called Exploratory Data Analysis (EDA). The whole idea is to guide ChatGPT through a structured analysis of any dataset, which helps you understand data you’ve never seen before in minutes instead of hours. The creator showed how with each prompt, your understanding goes from 0% to nearly 100%, uncovering insights you’d likely miss otherwise.
The DIG Framework in Action
The creator demonstrates this using a real dataset of Apple TV+ content. He walks through each step, showing the exact prompts he uses and explaining the logic behind them. It’s super practical.
Here’s a deeper look at how it works:
📌 The ‘Description’ Phase: Your First Look
This first step is all about getting a solid grip on the data you’ve been given. Imagine a coworker just quits and leaves you with a spreadsheet with no explanation. Instead of panicking, you start here. The creator uses a sequence of prompts to have ChatGPT describe the file. First, he asks it to “List all the columns in the attached spreadsheet and show me a sample of data from each column.” This gives a quick, digestible overview without overwhelming you. Next, he asks for five more random samples to spot potential inconsistencies that a single sample might hide. The most crucial prompt here is: “Run a data quality check on each column. Specifically look for missing or empty values, unexpected formats or data types, outliers or suspicious values.” In the example, this prompt instantly revealed that the ‘available countries’ column was 99.7% empty, a critical finding that tells him not to perform any geographical analysis. This step alone prevents you from wasting time on dead-end analyses.
💡 The ‘Introspection’ Phase: Asking the Right Questions
Once you understand what the data is, the next step is to figure out what you can do with it. This phase is all about brainstorming potential insights. The creator starts with the prompt: “Tell me 10 interesting questions we could answer with this data set and explain why each would be valuable.” The quality of ChatGPT’s questions tells you if it truly understands the data. The video shows it suggesting great questions like which genres dominate the catalog. But my favorite prompt from this section is: “What questions do you think someone would want to ask about this data but we can’t answer due to missing information?” This is genius because it surfaces gaps in your dataset and helps manage expectations with your boss or team. The creator takes it a step further by showing how to handle these gaps. He uploads a second, fake dataset (with viewership and cost data) and asks ChatGPT to merge it with the original using a common field, the IMDb ID. It’s a fantastic demonstration of how you can enrich your analysis on the fly.
✅ The ‘Goal Setting’ Phase: From Data to Decisions
This final step is what separates a technically correct analysis from a genuinely useful one. It’s about aligning your work with a specific business objective. As the creator points out, you don’t want to spend hours creating 20 slides only to find out your manager just wanted a simple ‘yes’ or ‘no’ on discontinuing a product. He uses a clear mission-briefing prompt: “My goal is to understand what content Apple TV should invest in next. Given this goal, which aspects of the data should we focus on?” This directs ChatGPT’s focus. The AI then provides a step-by-step roadmap, suggesting things like building a genre scorecard and ranking opportunities. The creator also shares an absolute gem of a final prompt to use before any presentation: “What are the key questions someone reading my analysis would ask and how should we proactively address them?” This simple question helps you anticipate challenges from managers and peers, making your final report that much stronger.
I was blown away by how this simple, structured process makes data analysis so accessible. You don’t need to be a data scientist to start pulling valuable insights from spreadsheets anymore. This approach levels the playing field for all of us.
For the full walkthrough and all the prompts, you have to check out the original video from this talented creator!