Practical guide

Sales order automation for distributors

Wholesalers and other B2B businesses often receive orders as email text and PDF attachments. Sales order automation turns those details into structured drafts. In FRB, a staff member reviews and approves the result before export for further processing.

Reduce manual order entry while keeping control

The work rarely consists of reading a SKU alone. A person must identify the customer, interpret product names, check quantities, spot missing information and enter the result in the right format. Different names for the same product increase the need for review.

Automated order entry requires more than text recognition. In FRB, automation covers collection, structuring and product matching; the order still needs approval, and an export file is not the same as an order entered into your ERP.

A controlled workflow

1. The order is collected. Orders are collected from a connected inbox or order folder. Email files, PDF, CSV, Excel and clear images can also be uploaded.
2. Products and quantities are structured. Known SKUs, product names and saved customer terms are matched to the product catalogue.
3. Exceptions are held for review. Unknown products, unusually high quantities, possible duplicates and unreadable documents are flagged for review.
4. A person approves. The original, sender, all lines and any warnings are reviewed.
5. The order is exported. Approved orders can be downloaded as CSV, Excel or FRB XML for an agreed next step.

Why FRB Collect does not replace your ERP

Your ERP, inventory and accounting systems remain responsible for downstream processing. FRB Collect sits before those systems and turns unstructured input into a clear order draft. The pilot does not send orders onward automatically and is not an invoicing, inventory or accounting system.

When automation should stop

Automation should stop when the answer is unclear. Two products may look alike, a quantity may be unusually high, or the same order may have been sent more than once. In these cases, a visible review task is safer than an automatic decision.

FRB Collect uses a product catalogue, customer-specific name mappings and consistent checks. When an unknown product is corrected and its name saved, the same wording can be matched next time.

Which documents fit the pilot?

The pilot supports plain-text and HTML email, text-based PDFs, clearly printed scanned PDFs, CSV orders, single-sheet Excel (.xlsx) files without formulas, clear PNG/JPG images and CSV product catalogs. Scanned PDFs use local text recognition, limited to 20 scanned pages per PDF and a bounded processing time. Product codes and quantities must be checked against the original. Unclear scans, handwriting, Word attachments and Excel files with formulas or multiple sheets require manual handling. Unusual tables may also need clarification.

Before launch, a limited set of documents is tested. The aim is to measure review needs, exceptions and export quality in the company’s real workflow, without promising automation that has not yet been proven.

Explore the workflow