ChatGPT prompts for supply chain: 25 prompts for planners and buyers

These prompts are written for planners, buyers and operations managers. Each one says who ChatGPT should be, what it should do, what you are giving it and what you want back. Copy a prompt, fill in the brackets, and check every answer before it leaves the building. For where a chatbot fits in your work and where it does not, read our guide to ChatGPT for supply chain.

Three rules before you paste anything

• Only paste what you are allowed to. Check your company policy and your supplier and customer contracts. Use a business account with the data controls your IT team approved.
• Give it the numbers. ChatGPT cannot see your ERP. If you ask it for your lead times or your stock, it will invent believable ones.
• Check the math. Ask it to show the formula, then check the result in a spreadsheet. It is a writing tool that can do arithmetic, not a calculator.

How to write a supply chain prompt that works

You are a [role]. [Task]. Context: [what is going on]. Data: [paste]. Give me [format and length].

The more specific the role, context and format, the less you will rewrite. Ask for a table when you want to compare, and give a word limit when you want an email.

Supplier communication

• Chase a late order. “You are a buyer. Write a short, polite email to [supplier] about PO [number] for [item], due [date] and not yet shipped. Ask for a confirmed ship date and tracking by [day]. Under 120 words.”
• Answer a price increase. “A supplier has announced a [x]% increase on [items] from [date]. Draft a reply asking for the cost drivers behind it and whether a longer commitment would change it. Firm but friendly.”
• Request corrective action. “Draft a corrective action request to [supplier] for [defect] found on [quantity] units of [part] received [date]. Ask for the root cause, containment and a date for the fix.”
• Make a long email usable. “Rewrite this supplier email in plain English, and list every date, quantity and commitment in it: [paste].”
• Prepare a supplier review. “Here are a supplier’s on-time delivery, quality and price figures for the last four quarters: [paste]. Write an agenda for a quarterly review and three questions to ask.”

For the figures to bring to that review, see our supplier scorecard guide.

Purchasing and procurement

• Write a request for quotation. “Write an RFQ for [item], [annual quantity], [specification summary], delivered to [location]. Ask for unit price at [quantity breaks], lead time, minimum order quantity, payment terms and how long the quote is valid.”
• Compare quotes. “Here are three quotes as a table: [paste]. Compare them on total cost per unit, including freight and duty, not only on price. List what is missing from each quote.”
• Prepare a negotiation. “I buy [quantity] a year of [item] from [supplier] at [price]. List the arguments the supplier is likely to make, my alternatives, and three things I can offer besides volume.”
• Summarize contract terms. “Summarize the delivery, price adjustment, liability and termination terms in this contract excerpt, and list anything a buyer should question: [paste]. This is not legal advice.”
• Group spend. “Group these purchase line descriptions into spend categories and suggest a name for each category: [paste].”

Before you compare quotes, work out the landed cost of each one, and check whether a minimum order is costing you more than it saves with EOQ vs MOQ. There is more on buyers’ use cases in AI for purchasing and procurement.

Inventory and planning

• Explain safety stock to a new planner. “Explain the safety stock formula using this example: average demand [x] a day, standard deviation [y], lead time [z] days, 95% service level. Show each step.”
• Check reorder points. “Here are average daily demand, lead time and safety stock for these items: [paste]. Calculate each reorder point, show the formula, and flag any item whose inputs look inconsistent.”
• Summarize the shortage list. “Here is this week’s shortage list: [paste]. Write a five-line summary for the operations meeting: what is short, why, the customer impact and what we are doing.”
• Plan for dead stock. “These items have had no sales in 12 months: [paste item, quantity, cost]. Suggest options for each (return, discount, bundle, scrap) and the questions to ask before deciding.”
• Split items into ABC classes. “Write the spreadsheet formulas to rank these items by annual value and give each an A, B or C class: [paste column names].”
• Explain a change in turnover. “Our inventory turnover went from [x] to [y] this year. List the likely reasons and the reports I should pull to find out which one it is.”

To check the answers, use the safety stock formula, the reorder point formula, our guide to finding dead stock and ABC/XYZ classification.

Demand planning and S&OP

• Draft a demand review agenda. “Draft a 30-minute monthly demand review agenda for a [size] manufacturer with [n] product families. Include the inputs, the owners and the decisions.”
• Explain a forecast miss. “Last month we forecast [x] units of [family] and sold [y]. Write questions for sales, marketing and customer service that would explain the gap.”
• Measure forecast accuracy. “Explain MAPE, WAPE and bias using these numbers: [paste forecast and actual]. Calculate each one and show the steps.”
• Turn meeting notes into decisions. “Summarize these S&OP meeting notes into decisions, owners and dates, and list anything left undecided: [paste].”
• Write up a scenario. “Our largest customer may order 30% more next quarter. Write a one-page scenario for the S&OP meeting: the questions to answer about materials, capacity, inventory and cash.”

See how the meeting itself runs in S&OP for small manufacturers, and check the accuracy figures against the forecast accuracy formula.

Data and reporting

• Clean supplier names. “Here is a supplier list with inconsistent names: [paste]. Group the names that are probably the same supplier and suggest one standard name for each.”
• Write a spreadsheet formula. “Write an Excel formula that returns the latest receipt date for each item from a table with the columns Item, Date and Quantity.”
• Check your item master. “Here are 20 rows of our item master: [paste]. List the problems: missing values, units that look wrong, and lead times that look unrealistic.”
• Write the monthly report. “Turn these KPIs into a three-paragraph report for the owner, in plain language, with one recommendation: [paste].”

If the item master check turns up a lot, start with a supply chain data readiness review.

When a prompt is not enough

ChatGPT drafts, explains and summarizes. It does not see your live data, it does not keep a plan up to date, and it can be confidently wrong. When you need forecasts, safety stock and reorder points for every item, worked out from your own sales and purchase history, that is a job for planning software. The Statadventure Optimization Lab does it in the browser, and its Ask AI add-on answers questions about your own plan. See AI inventory management software for how the two fit together.

Frequently asked questions

Is ChatGPT useful for supply chain work?

Yes, for writing, explaining and summarizing: supplier emails, meeting notes, first drafts of procedures and explanations of formulas. It is not a planning system. It cannot see your ERP, and it can state wrong numbers with confidence.

Can I paste our data into ChatGPT?

Only what your company policy and your contracts allow, and only in an account with data controls your IT team has approved. Leave out customer names, prices and personal data unless you are sure.

What makes a good supply chain prompt?

It names the role, the task, the context, the data and the format you want back. “You are a buyer. Draft a reply to this supplier asking for a confirmed ship date, in under 120 words” works far better than “write an email to a supplier”.