The day the dispatch spreadsheet stops scaling

Somewhere in most small carriers there is a workbook. It has a name like Planering_v4.xlsx, a tab per week, colour fills that mean something to two people, and it has planned every load the company has ever moved. It works. That is the part software vendors usually get wrong about spreadsheets — they treat them as evidence of disorganisation, when in a young transport operation the spreadsheet is normally the correct decision, arrived at for good reasons.

This article is about the point where it stops being correct. That point arrives later than vendors claim and earlier than most operators notice, and it is not marked by fleet size. It is marked by four specific symptoms.

Why the spreadsheet wins for so long

A dispatch day has a natural shape in a grid: rows are loads, columns are days or vehicles, and the whole week fits on one screen. A planner reads that faster than any list view, because the layout is theirs — built around how this company actually works, not around how a product designer imagined a generic carrier works.

It is also honest about cost. Changing the model takes seconds: a new column, a new tab, a rule that exists nowhere except in the planner’s head and the conditional formatting. No vendor, no ticket, no release. When a customer asks for something structurally odd — goods held for a week, a part-load split across two lanes, a price agreed on a phone call — the spreadsheet accepts it without argument. Systems that refuse such things are not more correct. They are less flexible, and sometimes that trade is worth making and sometimes it is not.

So the crossover is not about sophistication. It is about how many people and machines need to read and write the same facts at the same time.

Symptom one: two people in the same file

Microsoft does support several people editing one workbook simultaneously, but the conditions are worth knowing. Co-authoring requires the file to be in the modern .xlsx format and stored on OneDrive, OneDrive for Business or a SharePoint Online library — an on-premises SharePoint site does not support it. Certain contents switch it off, including OLE objects, ink objects and the legacy Shared Workbook feature. On a plain network share, the second person to open the file gets it read-only.

The visible cost is queuing: the traffic planner waits while someone in accounts has the file open. The invisible cost is worse. When the lock is inconvenient enough, people work around it — a copy on a desktop, a printout annotated by hand, a message that never reaches the file at all. Now there are two versions of the truth and no reliable way to tell which is newer.

The test: how often does somebody in the office ask are you in the file? If the answer is daily, you crossed this one a while ago.

Symptom two: the sheet stops at the office door

The workbook lives on the office network and the driver lives in a cab. The link between them is a phone call, which means every status change costs two people’s attention and arrives only when both are free to speak. Proof of delivery follows on paper: a signed note that reaches the office when the vehicle does, which on a multi-day tour can be Friday for a Monday delivery.

That gap has a direct financial edge, because most carriers cannot invoice what they cannot evidence, and the evidence is in a folder in the cab. We went through that specific delay in what paper PODs cost in invoicing time.

The test: count the days between delivery and invoice, and count the calls per day that exist only to ask a driver where they are.

Symptom three: nobody can reconstruct Tuesday

Overwrite a cell and the previous value is gone. OneDrive and SharePoint keep file versions, so a lost afternoon is usually recoverable — but recovering it means opening old copies and comparing them by eye. That is not the same as being able to answer who changed this delivery window, when, and what did it say before?

Most weeks that does not matter. It matters on the day a customer states the booked window was 08:00, or a damage claim turns on which driver loaded which pallet, or a waiting-time charge depends on when the vehicle actually arrived. In those conversations, an operation that cannot show its own record tends to concede for goodwill. That concession is a real cost, and it never appears in a software comparison because it gets filed under sales rather than IT.

The test: the last three disputes — settled with evidence, or absorbed?

Symptom four: the customer is outside the file

The spreadsheet is internal by nature. Customers therefore get their information the only other way available: they ring, and a dispatcher stops planning to read a row aloud. Larger customers formalise this into a demand for a status email each morning, which is the same interruption with a schedule attached.

None of that is unreasonable behaviour from the customer. They have their own production line, their own end customer, and no window into yours. But the cost lands entirely inside your dispatcher’s day. The arithmetic is in what a customer portal actually removes.

The fifth symptom, which is about people

Eventually the workbook becomes a person. One planner knows why column M is formatted the way it is, which tab is stale, and which of the two copies is the current one. When that person is on holiday the operation runs at reduced capacity, and everybody knows it. The company has accumulated a single-person dependency without ever deciding to.

Software does not remove key-person risk. It moves it from an undocumented file into a documented configuration, which is a smaller and more transferable version of the same problem — worth being clear-eyed about rather than selling as a cure.

What is not a reason to switch

Being small is not one. Neither is a competitor mentioning their TMS, nor a consultant calling the spreadsheet unprofessional. One vehicle, twenty loads a week, one planner who is also the owner: a spreadsheet is faster, cheaper and more adaptable than anything you could buy, and switching would cost you speed you cannot spare.

The crossover is genuinely not a fleet size. A three-vehicle operation running multi-drop distribution with customer-held stock can meet all four symptoms in its first year. A fifteen-vehicle operation running the same contracted shuttle every day, with one planner and one invoice a month, may never meet them at all. Judge the symptoms, not the truck count.

Put a number on it first

Before shortlisting anything, spend one ordinary week measuring: minutes spent retyping the same booking into a second place, status calls received, days between delivery and invoice, times the file was locked, disputes conceded for lack of evidence. Multiply by the weeks you work. That figure is what you would be buying against — and if it is smaller than a subscription, the correct answer is to keep the spreadsheet and look again in a year. That is a legitimate outcome, not a failure of nerve.

If the figure does justify a change, the questions that separate one system from another are in our checklist for choosing a TMS. And the workbook does not need to be thrown away afterwards: for a one-off analysis, a scenario, or a tender calculation, a spreadsheet fed by an export is still the best tool there is.

Worth knowing before you commit to anything: navichain has a free tier — 0 kr a month, one vehicle, 25 operations a month, two back-office users and 25 business partners, with every feature included and the driver app at two seats per vehicle. No card is required to start, so you can plan a real week in parallel with the spreadsheet and judge for yourself whether the four symptoms go away. The plans above it are on the pricing page, and the platform page shows what is inside them.