Real shops. Real racks. Real results.

The situation

A portal-frame fabricator running two plants. Bills of materials lived in one estimator's spreadsheet, stock lived in the memory of whoever had been in the yard longest.

The challenge

Nobody trusted the stock figure, so every job was counted twice before it was committed. Purchase orders were raised from a phone call and rekeyed into Xero at month end. A missing bolt set could hold a frame on the floor for a week.

The outcome

Both plants barcoded and counted in over a fortnight. Bills of materials moved out of the spreadsheet and into formula-driven parts lists, so a change of span reflows the cut list. Purchase orders now post straight into Xero as they are approved.

The situation

Every order is a variant. Two estimators priced each enquiry by hand from a master workbook, then rebuilt the parts list once the order landed.

The challenge

Quotes took two to three days, by which time material costs had moved. The same product got built from a slightly different parts list depending on who priced it, and shortages only surfaced when the picker reached an empty bin.

The outcome

The configurator builds the bill of materials as the customer talks. Pricing runs from live cost, exchange rate and margin profile. On approval every line is routed pick, buy or make and the pickable stock is reserved, so a shortfall is a conversation on day one rather than a surprise on despatch day.

The situation

Three warehouses, each ordering independently against a min/max set years ago and never revisited.

The challenge

Cash was tied up in racks of stock nobody had asked for in eighteen months, while the fast movers ran out roughly once a fortnight. Nobody could see what the other two sites were holding.

The outcome

Demand-driven buffers replaced the static min/max, sized per part and per warehouse from real consumption and recalculated on the schedule they picked. Buffer alerts flag the reds before the line goes short, and inter-site transfers move stock instead of ordering more of it.

The situation

Twelve on the bench, work planned on a whiteboard, job sheets printed each morning and annotated in biro through the day.

The challenge

What had actually been made was typed up a fortnight later, if at all. Scrap was never recorded, so the costed job never matched the real one, and the office had no idea what was finished until someone walked the shop.

The outcome

Work orders are released, started and completed on a phone at the bench, with scrap captured as it happens. The office sees progress live, and the costed job finally matches what the workshop actually did.

The situation

A subcontract machine shop where the drawing office, the stores and the buyer each had their own name for the same component.

The challenge

Duplicate parts were being created weekly, each with its own stock figure and its own supplier. Nobody could answer how many of a component were genuinely in the building.

The outcome

New parts now go through the Design Queue instead of straight into the catalogue: the requester states the reason and the near-matches, a designer answers in the thread, creates it properly, and closes the request against the real part number. Duplicate creation effectively stopped.

The situation

Kit is fabricated in the yard, palletised, and fitted on site by the firm's own crews.

The challenge

Pallets were made up from a printed list and checked by eye. A wrong or missing item was discovered a hundred miles away, on the day, with a crew standing around waiting for it.

The outcome

Picking is grouped pallet by pallet and gated by a scan: the app will not accept a line until the location scans and the quantity matches what was reserved. Crews clock on in the same app, so the hours and the kit that went out sit against the same job.