First Place Supply, a janitorial and cleaning supplies distributor, ran an early pilot of Dynition's purchase order automation in parallel with its normal manual process. Measured across the parallel run, 95% of POs were processed with zero rep touch, order entry went from about 15 minutes per PO to 20 seconds, each rep got roughly 6 hours a week back, and order errors fell from about 3% to under 0.5%.
Who is First Place Supply?
First Place Supply is a distributor in the janitorial and cleaning supplies (jan-san) category. Its wholesale customers send in purchase orders, and its reps turn those POs into fulfilled orders, the same inbound order flow that most distributors run every day.
The problem: manual purchase order entry
Manual PO entry looks the same at almost every distributor. A rep opens the purchase order, reads each line, matches part numbers and shorthand to the catalog, applies the customer's contract pricing by hand, checks stock, and types the order into the store or ERP. It takes minutes per order, it competes with selling time, and a single mistyped quantity or price ships the wrong product. Once rep time and corrections are counted, entering a PO by hand can run roughly $40 to $80 per order.
How the pilot ran
The pilot was a parallel run. Incoming purchase orders were forwarded to Dynition, which read each PO, matched every line to the catalog, applied the buyer's contract pricing, checked inventory, and produced a finished order for a rep to review, while the team's existing manual process kept running as normal. That made the comparison direct: the same real POs, processed both ways.
These are results from one early pilot, not a large-scale study. We publish them because they are the numbers we actually measured.
The results
The headline change is touch: 95% of purchase orders went from inbox to a clean, priced order in the store with zero rep touch, and the time from inbox to a live order fell from about 15 minutes to 20 seconds. Across a week of order volume, that returned roughly 6 hours per rep, time that went back into customers instead of data entry.
Accuracy moved the same direction. Manual entry ran at roughly a 3% order error rate; during the pilot, errors on processed orders fell below 0.5%, and every pricing, address, and payment-term discrepancy was caught before the order was created rather than after it shipped.
What changed for the team
The reps' job shifted from typing orders to approving them. Clean, fully matched POs came through ready to confirm in one click, and only the genuine exceptions, such as an unfamiliar part number or a price that did not match the contract, asked for a person's judgment. Nothing about the storefront, the catalog, or the customers' ordering habits had to change: buyers kept emailing POs exactly as before.