Capture orders from e-mail automatically
Orders arrive as free text, a PDF or a photo. Onshop reads the e-mail, drafts the order and asks the customer when something is unclear.
Monday morning in the order mailbox
Twelve new e-mails have arrived by eight o’clock. One regular customer writes: “Same as last week, but 40 instead of 30 boxes of wall plugs 8 mm this time, plus 5 rolls of tape.” Another e-mail contains a PDF from a customer’s inventory system with seven lines using their own product names. A third is a photo of a handwritten list taken at a building site. Mixed in are an invoice, a newsletter, and an inquiry about delivery times.
Someone in sales support opens these messages one by one. They look up the customer, find the correct product for every line, and type the quantity and unit into your system. For the first e-mail, they must first determine what “same as last week” actually means. With the third, they struggle to decide if a digit is a 4 or a 9. This work is not complex, but it consumes time every day and is prone to mistakes. A swapped item number or a missed line often only becomes apparent when the customer complains about the shipment.
What Onshop does with that e-mail
You link the mailbox where your orders arrive via IMAP. Onshop checks it regularly and reads every incoming e-mail, including the text and attachments such as PDF files, photos, and scans. First, the software determines if the e-mail is actually an order. The newsletter, the invoice, and the delivery question are not turned into orders.
Every identified order becomes a draft. Onshop identifies the customer using the sender’s address, the company domain, or the name mentioned in the e-mail. It matches the lines to your products using the item number or the exact product name from your catalogue. Units like “Stk.”, “Karton”, or “VPE” are recognised and mapped to the units in your catalogue.
The original e-mail appears next to the draft. Every line is assigned a number, and the corresponding number appears on the highlighted section of the text or attachment where the data was found. You no longer have to search for the source of a figure; you see it immediately.
When something is missing, Onshop asks
Some orders are incomplete. A customer might write “tape as usual” when you stock three different types. Another might order “2 pallets” of a product you only sell by the box. In some cases, the quantity is missing entirely.
Onshop identifies these gaps and lists them on the order. With one click, it creates a draft reply to the customer asking these specific questions. You review the draft, make any necessary changes, and send it from Onshop through your own mailbox. When the customer replies, Onshop links the response to the order and updates the draft by filling in the quantity, selecting the tape, or adding a line. The customer does not need to resend the entire order, and you do not have to compare the answer with the original e-mail.
You have the final say
Onshop never confirms an order automatically. The draft remains open until someone on your team reviews it. You fix any errors, add missing details, and then confirm. Lines that Onshop could not match to a product are marked “not matched” so they are easy to spot.
This division of labour is intentional. AI processes free text quickly and usually correctly, but not always. The person responsible for the delivery must remain in control. Onshop shifts the workload from manual typing to checking rather than removing you from the process.
Your system stays as it is
Onshop is not a replacement for your CRM, your inventory system, or the Excel lists you currently use. You upload your customers and products as a CSV file and export confirmed orders as a CSV, which can be formatted for Excel. Direct transfer of orders to your ERP or inventory system is planned.
Many orders arrive as attachments rather than plain text. You can read more about how Onshop processes PDF files, scans, and photos at Orders from PDF, scans and photos.
