The sixty-second answer
Move off a spreadsheet when a second person starts editing it, when a row in it carries an obligation you cannot afford to miss, or when nobody can say which copy is current. File size is not the trigger. Shared editing and real consequence are the two signals that matter.
Spreadsheets are not the problem
It is worth saying at the start, because most articles on this subject are written by people selling the replacement: the spreadsheet is one of the best tools ever built for small business. It is flexible, it costs almost nothing, it needs no training, and it does exactly what you tell it. A very large share of Canadian businesses run on one, and given that 1.08 million of the country's 1.10 million employer businesses are small businesses [1], that is not a fringe practice. It is the default.
The question is not whether spreadsheets are good. It is which jobs they stop being good at, and how to notice the moment they stopped.
The three failure modes
Spreadsheets fail in three specific ways. Everything else people complain about is a symptom of one of these.
Concurrent editing. A spreadsheet assumes one author. The moment two people need to change it, you are either coordinating by message — "are you in the file?" — or you are relying on a cloud editor's merge behaviour and hoping. Neither scales past about three people, and both fail quietly rather than loudly.
Silent errors. A formula range that stops one row short of the data. A sort applied to one column instead of the whole block. A paste that landed in the wrong cell. None of these produce an error message. They produce a number that looks fine, and the mistake surfaces later, usually in front of a customer or an accountant.
No memory. A spreadsheet does not remember who changed a value, when, or why. Version history in a cloud editor helps a little, but it records file versions, not business events. When you need to know why a customer's status changed in March, the file cannot tell you.
If none of those three are hurting yet, you do not have a spreadsheet problem, no matter how many rows you have.
The signals that it is time
A second regular editor. This is the clearest one. One owner, one file, no problem. Two staff plus the owner, all editing the same file during the same week, and you have crossed the line whether or not anything has broken yet.
The file has copies. If there is a version in someone's downloads folder, a version attached to an email from last month, and a version on the shared drive, the system of record has already failed. You just have not found out which decisions were made from the wrong copy.
A row means an obligation. When a line in the file represents a promise — an appointment, a renewal, a deposit held, a follow-up committed to — the cost of a missed row is no longer inconvenience. It is a broken commitment to a customer.
You are re-typing. If the same customer is entered into the spreadsheet, then into an invoice, then into an email tool, the spreadsheet has become a manual integration layer staffed by a human. That is the most expensive thing on this list, and it never appears on any statement.
Someone asks a question the file cannot answer. "When did we last talk to them?" "Who approved that discount?" "How many of these did we do in July?" A spreadsheet holds state, not history. When the questions become historical, the tool is wrong.
Where the tax and privacy edges are
Two areas turn a spreadsheet inconvenience into an actual exposure.
The first is sales tax. The Canada Revenue Agency treats a business that does not exceed the small supplier threshold over four consecutive calendar quarters as a small supplier that does not have to register, with voluntary registration available; a business that is not a small supplier and makes taxable supplies in Canada must register [3]. That is a rolling four-quarter test, which means it has to be watched continuously rather than checked at year end. A spreadsheet can do it — but only if someone remembers to update it, and the whole point of the threshold is that it arrives during a busy stretch when nobody is updating anything.
Rate handling is the second half of the same problem. The rate you charge depends on the place of supply, and the CRA's own worked example has a Vancouver store charging 13 per cent HST on a delivery to Toronto because the place of supply is Ontario [4]. If you sell across provinces, a spreadsheet column of hand-entered rates is a standing invitation to a silent error of exactly the kind described above.
The second area is personal information. If a file holds customer details, PIPEDA obligations travel with it. Schedule 1 requires knowledge and consent for collection, use and disclosure (Principle 3), gives individuals the right on request to be informed of the existence and use of their information, to access it, and to challenge its accuracy and have it amended (Principle 9), and requires that information no longer required be destroyed, erased or made anonymous under guidelines the organisation has actually developed (clause 4.5.3) [2]. Every one of those is a task you have to perform against every copy of the file. Six copies means six searches and six deletions, and the ones in an inbox are the ones you will miss. Our guide to PIPEDA requirements for small businesses works through the obligations in order.
The same copy problem hits your mailing list. Canada's Anti-Spam Legislation prohibits sending a commercial electronic message without express or implied consent and requires prescribed identifying and contact information in the message [5]. If consent and unsubscribes are tracked in a spreadsheet that is exported to an email tool periodically, the two drift apart, and the drift always runs in the direction of mailing someone who asked you to stop.
What to move, and in what order
Move the file with the most editors and the most consequence first. In most small businesses that is the customer list or the schedule — not the financial model, which usually has one author and is genuinely well suited to a spreadsheet.
Move one thing. Let it run for a month. Then move the next. A simultaneous migration of every file is how a business ends up running two systems badly and reverting to neither. Our walkthrough on switching business software without downtime covers the sequencing, and if the file in question is your customer list, whether a small business needs a CRM is the more specific question to answer first.
When you evaluate a replacement, apply one piece of Canadian consumer-protection discipline to the sales conversation: under the Competition Act, a performance or efficacy claim must be based on an adequate and proper test, and the proof of that test lies on the person making the representation [6]. If a vendor tells you their system will save a specific amount of time, asking what that figure is based on is entirely fair.
Finally, retire the old file properly. Archive one dated, read-only copy and delete the working versions. A spreadsheet left editable next to a new system is the single most common reason migrations quietly reverse.
Where we sit
MapleWorkSuite is modular, which matters specifically for this problem: you can replace the one spreadsheet that has outgrown itself without committing to replace the rest. Switch on a customer record or a scheduling app, move that file, and leave your financial model exactly where it is.
We will not tell you that spreadsheets are obsolete, because they are not. We use them. What we will say is that a spreadsheet is a poor system of record the moment more than one person depends on it being right, and that most owners recognise that moment months after it happened. If you can name a file that three people edit and that would cause a real problem if it were wrong tomorrow, you already have your answer — and our guide to what to automate first is the natural next step.