In this guide 6 sections
A spreadsheet suits one editor. It fails once several people edit stage status, parts go out for processing, and dispatch depends on partial quantities against one order. Switch when the sheet answers fewer questions than people ask the planner, and the planner becomes the only way to read it.
Where the spreadsheet works well
A workbook is cheap and quick. For ten or fifteen open orders, one sheet with order, part, quantity and due date is enough, and the owner can see everything at once. A plant machining brackets for two or three regular customers can run this way for years without much pain, because the owner remembers every drawing.
It also suits a plant still learning. You can add a column on Tuesday for a fault found on Monday, which is hard to do in rigid software that wants a change request first. Pivot tables and VLOOKUP formulas cover a surprising amount of ground, from late-order lists to weekly rejection counts per customer.
- Short order list
- One person owns the file
- Few stages per part
- Customers who accept a weekly status call
Where it starts to fail on orders and stages
The first crack is stage status. Cutting is done for 400, turning for 250, and plating has 120 back. One status cell cannot hold that split, so people add columns, then colours, then a second tab. Conditional formatting helps for a while, but a sheet coloured by five people for six months is read by none of them. Anodising, zinc plating and induction hardening each add their own wait to a part’s route, and each wait needs its own date on the sheet.
The second crack is editing. Two people open the same file, one saves over the other, and a finished stage vanishes, while copies named final, final2 and latest multiply on desktops.
The third is traceability. A customer asks which heat number went into a returned batch, and the sheet has no answer because nothing links heat number to order. A drawing revision change midway through a run has the same problem, since the sheet cannot show which pieces were made to the older print.
Outside processing and dispatch break it first
Job work sends parts out for plating, heat treatment or grinding and takes them back in lots, sometimes split across two trucks. A sheet can record that a lot left, but it rarely links the lot to the order, the vendor’s delay and the quantity short on return. Hardening cracks, plating peel and short counts are noted in emails instead of against the lot, and the emails are found only when a customer complains.
Dispatch has the same trouble. Partial deliveries need a balance per order, a packing list and a matching invoice, but by the third shipment the sums are done by hand. Errors reach the customer. Gate passes, delivery notes and weighbridge slips pile up beside the file, never linked to it, and a clerk matches them only when a customer disputes a delivery.
- Vendor, date sent, date due, date back
- Quantity sent against quantity returned
- Partial dispatch balance per order
- Packing list matched to invoice
Questions that tell you it is time to switch
Ask who can say where order 2174 stands. If only the planner can answer without phoning the floor, the sheet has become a person, and that person is the bottleneck. Try the same test with a customer’s inspection report, a rework order, a returned flange or a certificate sent to a buyer last month.
Count the weekly hours spent copying status between tabs, and count the customer calls that begin with a missing part. Watch both counts against order volume. If either keeps rising while volume holds steady, the sheet is the constraint, and month-end shows it first when invoicing waits for a manual match of delivery notes.
A sensible way to move over
Do not retype the history. Start with new orders only, in software that tracks stage status, outside lots and partial dispatch, and keep the old workbook read-only for closed orders. Trial it on a mix of turned shafts, milled housings and sheet brackets, so that cutting, turning, milling, plating and dispatch each get exercised at least once.
Run both for a month. Pick a single product family, list the five questions the planner hears most often, and test those first, including the dispatch balance for a partly shipped order. Include a vendor who processes parts for you, because outside lots are where software shows its value fastest, and key that vendor’s lot reference against each order.
Spreadsheet versus software test sheet
A comparison worksheet with rows for order status, stage quantities, outside processing lots, partial dispatch, rework and traceability, each row phrased as a question the planner hears. Three columns record how the current spreadsheet answers each, how long it takes, and whether it matches the floor, with a total at the bottom.
Open a blank worksheet to printQuestions owners ask
Can a spreadsheet handle outside processing?
For a few lots a month, yes. Once several vendors hold parts at once, tracking sent and returned quantities by hand becomes unreliable, and a short return is noticed only at dispatch.
Will software slow the planner down?
At first, a little. Entry takes discipline from the floor, the stores and the dispatch desk, and the first fortnight feels slower than the sheet did. The payoff is that stage status, dispatch balance and vendor delay are read from one place, without a phone call.
Should we keep the spreadsheet?
Keep it for closed orders and one-off analysis, and use software for live orders, where several people update the same record at once.
How long does a trial take?
One month on one product family is usually enough to see whether status questions get faster answers, if it includes one vendor lot and one partial dispatch.