You can create a shipping label from a screenshot by uploading the image to Billy, the AI assistant built into GoatLabels. Billy reads the order details visible in the picture — recipient name, street address, city, state, ZIP — and turns them into a draft label you review, rate-shop across USPS, UPS, FedEx, and DHL, and buy from a prepaid wallet. Nothing is purchased until you confirm the draft, and the whole flow works from a phone.

What can Billy actually read from an order screenshot?

Billy extracts whatever shipping-relevant text is visible in the image. In practice that means the recipient's name and full address, and — when the screenshot shows them — item names, quantities, and any notes the buyer typed ("it's the blue one," "gift, no receipt"). Order screenshots from Shopify admin, an Etsy order page, a marketplace conversation, an emailed invoice, or a plain Instagram DM all work, because Billy is reading the pixels, not integrating with the app the screenshot came from.

That last point matters more than it sounds. Platforms without a store connector — Whatnot, Mercari, Facebook Marketplace, a group chat — produce orders that never touch a CSV or an API. A screenshot is the universal export format every app already has. That's why the screenshot workflow is the backbone of AI shipping labels on GoatLabels, and why it's the standard answer for Instagram sellers whose entire order pipeline is a DM thread.

What Billy does not invent: package weight and dimensions that aren't in the image, the service you want, or payment. Those come from you — either typed in on the draft or pulled from package defaults you've saved. Billy drafts; you decide.

How do you go from screenshot to purchased label, step by step?

The full loop, assuming you already have a funded wallet:

Capture the order. Screenshot the DM, order page, or invoice on your phone or desktop. Crop if you like, but you don't have to.
Hand it to Billy. Upload the image in the Billy chat. You can add context in the same message — "8 oz, poly mailer, cheapest service" — and Billy folds it into the draft.
Review the draft. Billy shows the parsed recipient, address, and package details as an editable draft. The address is validated at quote time, with a residential/commercial badge so you can see what the carriers will see.
Compare rates. One quote returns an all-in price from each of the four carriers. Toggle insurance or a signature requirement per shipment and the fee is folded into the quote before you commit.
Buy and print. The label price comes out of your prepaid wallet — funded by card, Apple Pay, Google Pay, ACH, or crypto (BTC, ETH, USDT) — and you print from desktop or phone. No monthly minimum on the Free plan.

For a routine domestic order, the human part of this is about a minute: skim the draft, pick a rate, tap buy.

What should you always double-check before you buy?

Billy's extraction is a draft, not a verdict. The split between what the AI fills and what stays your job looks like this:

Field	Billy drafts it?	You confirm it?	Why it matters
Recipient name	Yes, from the image	Quick skim	Wrong name can complicate delivery and claims
Street address, city, state, ZIP	Yes, plus validation at quote time	Yes — especially unit numbers	Bad addresses cause returns and re-ship costs
Items / order contents	Yes, when visible	Skim for accuracy	Feeds customs descriptions on international orders
Package weight and dimensions	Only from your saved defaults or your note	Always	Wrong weight invites a carrier billing adjustment
Service choice	Suggested via rate comparison	Always — you pick the rate	Speed vs cost is a judgment call
Insurance / signature	No — per-shipment toggles	Your call each time	Fees fold into the quote only if you opt in
Payment	Never automatic	You tap buy	Nothing is purchased without your confirmation

The two fields worth slowing down for are weight and unit numbers. Weight, because carriers re-weigh packages in their networks and bill adjustments when the label understates it. Unit numbers, because a screenshot of a casual DM often just says "123 Main St" while the buyer lives in apartment 4B — and the validation badge can't conjure a unit the buyer never typed.

What happens when the screenshot is missing the weight or a unit number?

Billy doesn't guess silently. If the image has no weight, the draft uses your saved package defaults if you have them, or leaves the field flagged for you to fill. If an address element looks incomplete, quote-time validation surfaces it before money moves — an address that fails validation is a prompt to go back to the buyer, not a label that quietly ships to the wrong place.

The practical fix is usually one message back to the buyer: "Can you confirm the full address with any apartment or unit number?" Sellers who ship from DMs a lot keep a couple of these asks saved as snippets; there's a full set of copy-paste scripts in our guide to shipping labels for DM orders. Thirty seconds of confirmation beats a returned-to-sender parcel every time.

Which screenshots work best — and which reliably fail?

Screenshots that parse cleanly share three traits: the address is fully visible in one frame, the text is rendered by the app (not handwritten or photographed off paper at an angle), and there's exactly one order in the image.

Screenshots that cause trouble:

Cut-off addresses. The scroll position hides the ZIP or the second address line. Billy can only read what's in the frame — retake the shot with the whole block visible.
Multiple orders in one image. A screenshot of an order list forces the model to pick one. Screenshot the order detail page instead, one order per image.
Low-contrast or tiny text. Dark-mode apps at low brightness, heavy compression from re-forwarding, or a photo of a screen instead of a native screenshot all degrade extraction.
Buyer typos. Billy faithfully extracts what's written, including the buyer's mistakes. The validation step catches undeliverable addresses, but a valid wrong address (right street, wrong house number) only gets caught by you or the buyer.

When a screenshot fails, you're never stuck: paste the text of the conversation instead, or dictate the order by voice. All three inputs land in the same draft-review-buy flow — more on that in our overview of Billy, the AI shipping assistant.

Is a screenshot workflow safe for customer data?

The data in an order screenshot is the same data you'd type into any label form: a name and a shipping address, which end up printed on the outside of the box regardless of how the label was made. Using Billy doesn't broadcast it anywhere new — the image is used to draft your label inside your GoatLabels account.

Sensible hygiene still applies. Crop screenshots to the order details rather than uploading a full conversation if the thread contains payment details or personal chat. And keep your own records tidy: every label you buy lives in your shipment history with its address and tracking number, which is a better system of record than a camera roll full of DMs.

FAQ

Does Billy buy the label automatically after reading the screenshot? No. Billy produces a draft; you review it, pick a rate from the four-carrier comparison, and confirm the purchase yourself. Nothing is charged to your wallet until you do.

What if the screenshot is from a platform GoatLabels doesn't sync with? That's exactly the point of the feature. Store sync covers Shopify, Etsy, WooCommerce, and TikTok Shop; orders from everywhere else — Whatnot, Mercari, Instagram, group chats — come in via Billy's screenshot, paste, or voice input.

Can Billy read international orders from a screenshot? Yes. For international shipments Billy can also draft the customs description from the visible item details, but you supply the HS code and declared value yourself. Saved customs templates make repeat items fast.

Do I need a subscription to use screenshot drafting? No — Billy works on every plan, and the Free plan has no monthly minimum. The optional $40/mo Pro plan adds unlimited shipments, cheaper rates, and Billy on Telegram.

How accurate is the extraction? Clean, fully-visible screenshots parse reliably; the weak spots are cut-off frames and details the buyer never typed. The review step and quote-time address validation exist precisely so you catch the exceptions before money moves.