Nov 11AI vendor breach exposes Notion AI chat transcripts.Why it matters

AI prompts for fulfillment and shipping teams.

For the team that turns orders into packages, whether you ship in-house, use a 3PL, or both. 10 copy-paste prompts for Claude, ChatGPT, or any AI assistant, written from patterns we see in client work.

How to use these

Fill in the brackets.

Each prompt has [bracketed] spots. Replace them with your real numbers and context. The more specific you are, the better the answer.

Use a project.

In Claude or ChatGPT, put your policies, product list, and chart of accounts in a project once, so every prompt starts with your context.

Leave customer data out.

AI tools and vendors leak regularly (see the latest incidents). Remove customer names, addresses, emails, and messages before pasting. Card numbers and passwords never go into a chat. Amazon buyer data, including personalization text, has extra rules.

Check the math.

AI is good at structure and explanation and can still get arithmetic wrong. Spot-check any number you will act on.

Shipping cost

Find where you overpay on shipping.

Here is a month of shipments:

[paste order, weight, dimensions, zone, service, carrier, cost charged by carrier, and shipping charged to the customer]

Find: orders where a cheaper service would have met the delivery promise, dimensional weight surprises, boxes that are too big, and orders where we charged the customer less than it cost us. Total the savings for each fix.

Decide what to ask before changing shipping software.

We ship about [orders per day] from [locations] with [carriers]. Our current setup: [software and what frustrates us].

List what we should know about our own shipping before talking to vendors or an implementation partner: volumes, order types, rules we follow by hand, and integrations we rely on. Then list the questions to ask each vendor, and what answers should worry us.

Speed and accuracy

Find why orders ship late.

Here are last month's late orders with order time, channel, items, warehouse, when they were picked, packed, and shipped:

[paste data]

Group the late orders by cause (stock, cutoff time, batch size, staffing, address issues, system sync), count each, and suggest the fix with the biggest effect.

Cut mis-picks and wrong items.

Here are our mis-pick and wrong-item complaints for the last [period]:

[paste order, item ordered, item sent, picker if known, bin location, and similar-looking items]

Find patterns (similar SKUs, bin locations, variants, specific shifts), and suggest changes to bins, labels, scanning, or pick lists. Rank by how many errors each would prevent.

Special orders

Estimate what personalization mistakes cost you.

We sell personalized products. Each month we ship about [number] personalized orders, and about [number] go out wrong. Each mistake costs about [remake, reship, and refund cost] and [time].

Estimate the monthly and yearly cost, including lost reviews and repeat customers. Do not ask me for customer names, addresses, or personalization text; counts are enough. Then list where in our process mistakes usually start.

Plan for partial shipments and backorders.

Some orders ship in more than one package because [reason]. Our systems: [store, inventory, shipping, accounting].

Explain how a partial shipment should flow through each system so tracking, inventory, and invoices stay correct. List the places it commonly goes out of sync and a check for each.

Returns

Make sense of your returns.

Here are last quarter's returns:

[paste order, product, return reason, condition, refund amount, and channel]

Summarize return rate by product and channel, the top reasons, and what returns cost us including shipping and lost product. Suggest three product, listing, or packaging changes that would reduce returns.

Write a returns policy customers can actually follow.

Our returns rules: [paste current policy and the exceptions staff actually make].

Rewrite it in plain language, under 200 words, covering who can return, the window, condition, who pays shipping, refund timing, and exchanges. Then list the exceptions so support handles them the same way every time.

3PLs and warehouses

Check your 3PL invoice.

Here is our 3PL invoice and our own order and receiving data for the same period:

[paste both]

Check pick fees, storage, receiving, packaging, and special charges against our contract rates: [paste rates]. List every charge that does not match, with the amount, and draft a dispute email.

Decide whether to keep fulfillment in-house.

We ship [volume] orders a month in-house with [team size] and [space]. A 3PL quoted: [paste quote].

Compare full in-house cost (labor, space, materials, software, errors) against the 3PL, including setup and integration. Say what we would gain and lose in speed, control, and customization, and recommend one option with the volume at which the answer changes.

Make them yours

These prompts work best once your AI knows your business. Put your product list, policies, and chart of accounts in a project once, and every prompt starts from your context. When a prompt becomes something you run every week, that is usually the sign it should become a workflow that runs on its own.

That is what we build. See how we work with fulfillment and shipping teams →

Audited every year.

Deloitte and Amazon SP-API audits in 2026. PwC and Amazon SP-API in 2025. Shadow AI was involved in 43% of the breaches IBM studied for its 2026 report; we keep your data out of it. And if something we run goes wrong, we own the fix.

See the latest AI incidentsSee our audits