DIG | A Framework for Data Analysis with AI
Most professionals today work with data. Few were ever taught how to analyze it. And while AI is often marketed as a shortcut, the real breakthrough comes from applying frameworks—ones that clarifies context, prioritizes insight, and lets AI serve its most useful role— thinking with us. This is where the DIG Framework comes in—an approachable, three-step method for making sense of any dataset using ChatGPT.
“The true value of data is not in its volume— but in the clarity it can create.”
Why This Matters
In my own work—across product marketing, consulting, and leadership—I’ve often been handed datasets with little context and short timelines. The old model— clean, sort, pray. The new model? Partner with AI to analyze like a pro—with speed, accuracy, and clarity.
What changed was learning how to coach the model through a structured lens.
That lens is DIG—
Description → Introspection → Goal Setting
This proven model has been adapted from principles of Exploratory Data Analysis (EDA), made more accessible through modern AI. It allows anyone—regardless of data fluency—to explore new datasets, surface insights, and make better decisions.
Description: ‘What Am I Looking At?’
Start by asking ChatGPT to show its understanding of the data—you’re prompting for context.
Prompts:
— “List all the columns in this spreadsheet and provide a sample from each.”
— “Take 5 more random samples to check consistency and data format.”
— “Run a data quality check. Highlight missing values, anomalies, or odd formats.”
This phase does two things:
• Increases your understanding with minimal effort.
• Surfaces early problems before they derail your analysis later.
— Ask follow-up questions like “What does this column mean?” and make ChatGPT show its work. Meaningful insight starts with this level of clarity.
Introspection: ‘What Questions Can This Data Answer?’
Once you understand the structure, prompt ChatGPT to brainstorm questions the data might answer.
Prompt: “Suggest 10 interesting questions we could answer using this dataset. Explain why each would be valuable.”
Then follow up with: “Which columns are required to answer each question, and is the data sufficient?”
This phase reveals: • What’s possible • What’s missing • What’s worth exploring
And crucially—it helps you manage expectations. You’ll know what can’t be answered, so you don’t waste time chasing insights that aren’t there.
Goal Setting: ‘What Am I Trying To Decide?’
Once you know what’s in your data, clarify what decision you’re trying to make. Now, its time to give Chat GPT a purpose and clear goal via prompt—
““My goal is [insert goal]. Based on the dataset, which metrics or columns should we prioritize?”
In response, you’ll receive not just numbers, but narrative. Suggestions. Trade-offs. Scenario thinking. This is where ChatGPT shifts from a tool into a partner.
⎯ Why This Works
The DIG method doesn’t require technical skills. It requires thinking in frameworks. It levels the playing field between analysts and operators, founders and freelancers. It creates a repeatable process for turning ambiguity into clarity—without over-engineering the workflow.
What you’ll find isn’t just faster insight—it’s better questions. And in data as in life, better questions change everything.
#EssentialAI #dataanalysis #TheAvantLife