How to Create a Weekly Report with ChatGPT

Manually compiling weekly reports often feels like a race against time that you’re destined to lose half a day to. You download CSVs, wrangle data in a spreadsheet, and try to piece together a coherent story before your next meeting. This guide will walk you through how to use ChatGPT to automate the grunt work of analysis and writing, turning that hours-long task into a much faster process.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

First, A Reality Check: What ChatGPT Can and Can't Do

Using an AI tool like ChatGPT for reporting is a huge time-saver, but it's important to set the right expectations. Think of it as a brilliant but junior data analyst who is incredibly fast but needs clear instructions and supervision.

What it's great for:

• Summarizing Data: Quickly digesting a CSV of marketing or sales data and spitting out bullet points of key takeaways.
• Basic Analysis: Calculating week-over-week changes, identifying top-performing channels, or segmenting data based on your instructions.
• Generating Text: Drafting the narrative for your report, like an executive summary or channel-specific commentary.
• Creating Simple Visuals: Producing basic charts (as static images) to illustrate trends in your data.

What it's not great for:

• Connecting to Live Data: ChatGPT can't connect directly to your Google Analytics, Shopify, or Salesforce account. It relies entirely on the static data file you upload.
• Complex Visualizations: It cannot create interactive, filterable dashboards like you would in Power BI or Looker.
• 100% Accuracy: It can sometimes make calculation errors or "hallucinate" insights. You must double-check the important numbers.
• Understanding Business Context: It doesn't know your company's sales targets, previous campaign performance, or unique industry terminology unless you provide that context in your prompt.

Step 1: Get Your Data Ready for Analysis

ChatGPT's output is only as good as the input data you provide. Before you even think about writing a prompt, you need to gather and clean up your data in a format it can easily understand. An organized CSV or spreadsheet is your best bet.

Gather and Export Your Raw Data

Start by pulling the raw data you need for your weekly report. Your process might look something like this:

• Website Performance: Export a weekly report from Google Analytics showing key metrics (Sessions, Users, Bounce Rate, Goal Completions) broken down by Source/Medium.
• Sales Data: Export the week's orders from your Shopify or Stripe account, including order value, product sold, and customer location.
• Ad Performance: Export campaign data from Facebook Ads or Google Ads, focusing on Spend, Impressions, Clicks, and Conversions.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Combine and Clean Your Data

Next, consolidate this data into a single CSV file or a tab in a Google Sheet. The goal is to create a simple, flat table that is easy for a machine to read. Stick to these best practices:

• Use Clear Headers: Make sure column names are simple and descriptive (e.g., "Date," "Traffic Source," "Revenue," "Ad Spend"). Avoid ambiguous names or special characters.
• Keep Formats Consistent: Ensure dates are all in the same format (e.g., YYYY-MM-DD), currency is just a number without symbols, and percentages are decimals (e.g., 0.25 for 25%).
• Sanitize Your Information: This is critical. Do not upload any personally identifiable information (PII) like customer names, emails, or addresses. Replace them with generic IDs or remove the columns entirely. Treat any data you upload as public.

Here's an example of a simple, clean table structure combining data from multiple sources:

Date,Channel,Sessions,Conversions,Spend,Revenue 2023-10-23,Google Ads,1500,75,500,2250 2023-10-23,Facebook Ads,1200,60,400,1800 2023-10-23,Organic Search,2000,10,0,300 2023-10-24,Google Ads,1650,80,550,2400 2023-10-24,Facebook Ads,1100,55,380,1650 2023-10-24,Organic Search,2150,15,0,450 …and so on for the rest of the week…

Step 2: Upload Your Data and Craft Your Prompts

With your clean CSV file ready, it's time to put ChatGPT to work. Note that uploading files requires a subscription to ChatGPT Plus.

Upload Your Spreadsheet

In a new chat, click the paperclip icon and select your cleaned CSV file. Once it’s uploaded, you can start giving ChatGPT instructions.

Start with a Broad Prompt for an Overview

Begin with a general request to make sure ChatGPT understands your data structure. This helps you get a quick summary and confirms the AI is on the right track.

Example Prompt:

I've uploaded a CSV with our performance marketing data from last week. Please act as a senior marketing analyst. First, provide a high-level summary of the key metrics in the file.

ChatGPT will likely respond with a summary of total sessions, overall conversion rate, total ad spend, and total revenue. This is your foundation.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Ask Specific Questions to Dig Deeper

Now, drill down into an area of interest. The more specific your prompts are, the more useful the output will be. Think of this as a conversation where you guide the analysis.

Example Prompts for Specific Analysis:

• To analyze channel performance:
• To identify trends:
• To add business context:

Request Specific Formats for Your Report

One of ChatGPT's greatest strengths is structuring information. You can ask it to format its findings in a way that’s easy to copy and paste directly into your report document or email.

Example Prompts for Formatting:

• For an executive email summary:
• For specific report sections:

Step 3: Refine and Visualize

Your first pass gets you 80% of the way there. The final 20% is about refining the language and creating supporting visuals to make your report clear and compelling.

Iterate on the Content

Don't be afraid to ask for revisions. If an answer isn't quite right, you can give it feedback.

Example prompts for refinement:

• "That's a good start, but can you make the tone more concise?"
• "You mentioned that Organic Search revenue was low. Can you offer two potential hypotheses for why that might be?"
• "Please recalculate the ROAS as 'Revenue / Spend', not the other way around."

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Create Basic Visualizations

While ChatGPT is not a purpose-built business intelligence tool, it can generate basic charts using its Code Interpreter feature. These visuals are static images but are often just what you need for a quick weekly email or slide deck.

Example prompts for charts:

• For a bar chart:
• For a line chart:

You can then screenshot or download these charts and add them to your report. It saves you the trouble of building them yourself in Excel or Google Sheets for simple demonstrations.

Putting It All Together: A Sample Workflow

Here’s what a full workflow might look like from start to finish:

1. Monday Morning (9:00 AM): Export weekly data from Google Analytics and Facebook Ads. Combine them into a single, clean CSV and remove any sensitive user data.
2. Monday (9:15 AM): Upload the CSV to ChatGPT. Use an initial prompt to get a high-level summary and check for understanding.
3. Monday (9:20 AM): Ask a series of specific follow-up questions to analyze channel ROAS, daily trends, and top performers.
4. Monday (9:30 AM): Ask ChatGPT to generate a bar chart of revenue by channel and a line chart of daily sessions.
5. Monday (9:40 AM): Request a final, structured output: an executive summary, bulleted takeaways for each channel, and a concluding sentence.
6. Monday (9:45 AM): Copy and paste the text and charts into an email or Google Doc. Review the numbers and add your personal insights and recommendations. Your weekly report is ready to send before 10 AM.

Final Thoughts

Using ChatGPT to build weekly reports dramatically cuts down on the manual labor of data aggregation and summary generation. By preparing clean data and using specific, iterative prompts, you can turn a tedious reporting task into a quick conversation, freeing you up to focus on the strategic insights that actually drive decisions.

While this method is a huge leap forward from manual spreadsheet work, it still relies on static, exported files. For a truly seamless view, you need your tools to talk to each other automatically. This is where we built Graphed to be different. Instead of uploading CSVs, we connect directly to your live data sources like Google Analytics, Shopify, and Facebook Ads. This means you can just ask, "Show me a dashboard comparing Facebook Ads spend vs. revenue last week," and get a live, interactive dashboard that updates automatically, saving you from ever having to download a file again.