Developing sales forecasting models

Brief ChatGPT on your business

To create an effective sales forecast, start by providing ChatGPT with a comprehensive overview of your business. Include details such as industry, market, products/services, customer base, and sales history.

You’re going to help me develop sales forecasts for my business. I’ll start by sharing some details about my business. Let me know when you’re ready to proceed.

Business: [summarize business].
Operating history: [add historical detail].
Products: [summarize products].
Sales history: [summarize sales history].

ChatGPT may respond with a preliminary plan. If it doesn’t, you can follow-up with:

Our plan is the following. Please let me know if there are other steps you think need to be taken:

1. Data collection - I will collect historical sales data and share it with you.

2. Trends - I will share any observations about the business’ trends. You can also offer suggestions or ideas as to what might be influencing the business.

3. Data analysis - you will review the data to find interesting patterns.

4. Forecasting model - you will explain different forecasting models suggest the one to use. Once I confirm, you will proceed with the forecast.

Step 2: Data preparation

Accurate sales forecasting depends on high-quality data. Data preparation involves collecting sales data and cleaning it to ensure accuracy.

Sales data

You probably have a wealth of data available but to start, have to hand your monthly sales data in a simple format in a CSV file.

Other important notes about your business

As mentioned in the reply from ChatGPT, you’ll want to share business decisions, events, and changes that might help ChatGPT understand your data, and provide better forecasts.

Use the following prompt, along with your sales data attached:

I have attached the past 5 years of sales data.

Other notes:[include other important notes about your business decisions, events, and changes]

Step 3: Model building

Now that ChatGPT has ingested your historical data with notes, it can start developing a forecast for your business. Rather than simply proceeding, for your own education it’s worth understanding the different models available to you. ChatGPT may suggest an overly simple forecast model when you may believe there’s something better suited for the complexity of your business.

To get more details, ask ChatGPT to explain some models and suggest one to use.

Yes, but can you first explain different types of forecasting models and the one you suggest using?

With data analysis, ChatGPT can sometimes run into an issue with processing the data. It’s often self-corrective though. It make take a few minutes to complete its analysis and initial predictions.

You can also ask ChatGPT to reformat the data into a spreadsheet and provide more recommendations.

Reformat the projections into a spreadsheet I can download.

Provide more recommendations for increasing sales over the next 3 years.\

Congratulations! You’ve developed a 3-year forecast that will help you make critical business decisions.

Bonus: Advanced analysis with ChatGPT API

The ChatGPT API allows for automation and integration with other tools, enhancing your forecasting capabilities. Every company stores and analyzes its data differently. We suggest starting with this prompt:

Explain how to use the ChatGPT API to automate sales forecasting processes and integrate with other business tools.

ChatGPT can take you step-by-step through scripts, CRM integrations and other advanced forecasting methods.

This tutorial was created by Jonah.

‍