The outbound plan almost always starts as a shared sheet, and it genuinely works until a few hundred orders a day. The cure is not forcing people back into ERP forms — it is putting the spreadsheet inside the system.
What are two hours of dispatch worth?
All it needs is daily volume and the dispatch time left after the change. Labour, hired-in vehicles and reconciliation follow from those — every one of them editable.
This is administrative labour, not how much earlier the truck rolls. The part actually on the vehicle's critical path is far smaller — the source estimates about 45 minutes a shift.
Assumptions — click to see and change every number
People and labour
Fleet
Errors and reconciliation
Implementation
This is an estimate, not a promise. 'Less hired-in transport' is the softest line and only holds if you are actually paying for hired-in vehicles — if not, set it to zero and read the total again.
In a distribution business, the outbound plan almost always starts life as a shared Google Sheet. And it works — visual, flexible, instant. That is why it survives so long, not because anyone is avoiding the ERP.
The trouble only shows past a few hundred orders a day. And the usual cure is the wrong one: push the dispatchers back into the ERP's data-entry forms. This piece quantifies the other cure — keep the spreadsheet experience and put it inside the ERP.
1. Where the shared file breaks
Not because spreadsheets are bad. Because of three things a spreadsheet cannot have.
- Nobody owns a row. Several teams have it open, a
VLOOKUPhangs, and one dragged row or deleted cell stalls the whole day's dispatching. Nothing errors. - Copy-paste eats the first two to three hours of the shift. Filter by route, copy into each driver's tab, print orders one by one, then add them up by hand to produce the picker's list and the driver's manifest.
- The dispatch numbers and the stock numbers are two different truths. Goods assigned to a driver that the ERP has not deducted; or stock that ran out while the sheet keeps allocating it. The result is failed deliveries, and then the cost of delivering again.
The third is a copy of what the first piece called fragmented data. Here it does not sit between departments — it sits between a file and the system.
2. What the improvement actually is
Not abandoning the spreadsheet. Moving it inside the system.
- Keep the grid. Keyboard shortcuts, column filters, drag-fill, multi-row paste. No retraining, and more importantly no reason to drift back to the old file.
- Batch and route automatically. By area, by route, by vehicle capacity — the part people do slowly and inconsistently, and a machine does quickly and the same way every time.
- Print once per trip. One consolidated pick list so the warehouse walks the floor once, and one manifest carrying barcodes, the stop list, the COD to collect and a signature line.
- Approving a trip is a posting, not an announcement. Press approve and stock is reserved, the issue is created, and the COD lands on the right person's account. That is the real difference from a spreadsheet.
3. Four numbers, and which one to trust
A. Dispatch and warehouse hours
The firmest, because it rests only on time you can put a stopwatch on. Measure one shift and you have your baseline.
B. Less hired-in transport, by leaving the yard earlier
Leaving earlier dodges the peak and fits in more stops. But this is the softest line, and it only holds if you are genuinely paying for hired-in vehicles. If you are not, set it to zero and read the total again — what remains is still large.
One confusion worth naming: freeing 12 hours of administrative labour a day does not mean the truck rolls 12 hours earlier. Most of that work sits off the vehicle's critical path. The part that actually moves departure is much smaller — the source estimates about 45 minutes a shift.
C. Fewer wrong deliveries
Most of these errors are not carelessness. They start as one dragged row in a file four people have open at once.
D. Less COD reconciliation
The manifest locks in what each driver owes, so the end-of-day addition disappears. The smallest of the four, and the easiest to verify inside a week.
4. A mid-sized example
15 trucks and 20 motorbikes, 600 orders a day, two dispatch shifts.
| Line | Per month | Per year |
|---|---|---|
| Dispatch and warehouse hours | 23.4m | 280.8m |
| Less hired-in transport | 15.0m | 180.0m |
| Fewer wrong deliveries | 3.0m | 36.0m |
| Less COD reconciliation | 1.8m | 21.8m |
| Total, against 80m of implementation | 43.2m | pays back in ~1.85 months |
The estimator above reproduces this table: three of the four lines match exactly, the total is within 0.3%, and payback lands on 1.85 months.
5. What to take away
People cling to the Google Sheet because it is convenient, not because they are resisting the ERP. This is where a great many projects misread the motive and pick the wrong cure. If the new system is slower to type into than the old tool, people go back to the old tool, whatever the policy says. The only way to beat a spreadsheet is to put the spreadsheet experience inside the system — and then add the things a spreadsheet cannot do.
And as with the piece on templates, what makes this worth doing early is not that it is the biggest, but that the people using it feel the benefit in the first shift.
A process only survives if it is faster than the shortcut around it. Taking away a convenient tool without handing back a more convenient one is not an implementation. It is a bet on discipline.
The figures illustrate a mid-sized distributor and are not taken from any client's system. Replace them with your own in the estimator above.