There’s a tax running through most businesses that never shows up in the accounts: the human time spent moving information out of documents and into systems. Invoices, quotes, contracts, shipping docs, customer forms. Somebody opens each one, reads it, and re-types it somewhere else.
Put a number on it
On a live build for a UK supply-chain operator, the pre-automation baseline was 30 minutes per document — conservatively — to read a shipping document from email and key it into their transport management system. Multiply your own weekly document count by half an hour, then by a loaded hourly rate. Most operators who do this sum for the first time stop us halfway through it.
Why it never gets fixed
Because it’s nobody’s job. The tax is spread across everyone in slices — ten minutes here, a document there — so no single person feels enough pain to fix it. It’s also invisible in reporting: no P&L line reads “re-typing PDFs.”
What the fix looks like
The system reads incoming files, extracts the fields that matter, validates them against your rules, and pushes them where they belong. The part that makes it trustworthy is the exception queue: anything the AI isn’t sure about goes to a human, with its uncertainty flagged. People review the 10% that needs judgement instead of processing 100% by hand.
For that supply-chain client, documents now flow from inbox to TMS without a human in the middle — and volume has doubled every month for three months without headcount moving. That build sits in our standard £4k–£8k range. The tax, left alone, costs more than that every year. Start with the sum above; if the number scares you, the audit is how you find out exactly where it’s leaking.