The sixty-second answer
Most Canadian small businesses need about six software subscriptions: email and identity, accounting, a customer record, scheduling, a website, and payments. Anything beyond that should hold data nothing else holds, or do a job a customer would notice if it stopped. Most businesses carry considerably more than six.
Why the question gets asked at all
Nobody sets out to collect software. The list grows the way most business overhead grows: one problem at a time, each solved on the day it hurt, none of them ever reviewed together. Two years later there is a scheduling tool nobody opens, a design product bought for one brochure, an email platform that still holds a list of customers, and a project tracker that lost its argument with a shared spreadsheet.
This is not a large-company problem being handed down. As of December 2024 there were 1.10 million employer businesses in Canada, and 1.08 million of them — 98.2 per cent — were small businesses [1]. The overwhelming majority of software buyers in this country are owner-operators making purchasing decisions between other jobs, with no procurement process and no one whose task it is to notice that a charge has recurred forty times.
So the honest version of the question is not "how many is correct" but "how many can one person keep in their head". That number is small.
The six categories almost every business needs
Email and identity. A business address on your own domain, plus whatever accounts hang off it. This is the foundation, because it is how customers reach you and how every other tool identifies you.
Accounting. Bookkeeping, invoicing, and enough reporting to answer a question from your accountant without a weekend of reconstruction. You need this before you think you do — partly because the Canada Revenue Agency's small supplier test turns on revenue over four consecutive calendar quarters, and you cannot see that threshold approaching if your books are a shoebox [4].
A customer record. One place where a customer's name, contact details, history, and current status live. It might be a dedicated system, or it might be a well-disciplined feature of something else. What matters is that there is exactly one of it.
Scheduling. Anything time-based — appointments, jobs, shifts, deliveries — needs a calendar that other people can see and, ideally, book into without a phone call.
A website. Not a brochure; a working page that states who you are, where you operate, and what you do, in text a machine can read. This is increasingly the only version of your business an AI assistant will ever see.
Payments. Getting paid, and knowing who has not paid. Frequently attached to accounting, occasionally separate.
Six. Some businesses run this on three subscriptions because one system covers several categories honestly. Some run it on nine because each category has fragmented. Both can be fine. What is not fine is not knowing which you are.
The four tests for a seventh tool
When someone proposes adding something, put it through four questions in order:
Does it hold data nothing else holds? A tool that stores a unique record — signed documents, inspection photos, payroll history — earns its place on that basis alone. A tool that stores a second copy of your customer list is a liability, not an asset.
Would a customer notice if it disappeared? If the answer is no, the tool is internal convenience. Internal convenience is allowed, but it should be cheap and it should be the first thing cut when you review.
Is more than one person using it? Single-user tools quietly become single points of failure. When that person is away or leaves, the data is effectively gone even though the subscription is still being paid.
Can something you already pay for do this acceptably? Not perfectly. Acceptably. The cost of a second-best feature inside a system you already run is almost always lower than the cost of a best-in-class feature in a system that does not talk to anything.
A proposal that survives all four is a real need. Most do not survive the fourth.
Count data locations, not subscriptions
The subscription count is a proxy for the number that actually matters: how many separate places hold your customers' personal information. That is the number that determines how hard your life gets when something goes wrong.
PIPEDA's Schedule 1 requires that the purposes for collecting personal information be identified before or at the time of collection, that collection and use rest on the individual's knowledge and consent, and — the clause people forget — that information no longer required be destroyed, erased or made anonymous under guidelines and procedures the organisation has actually developed [2]. That last obligation is trivial with one customer record and close to impossible with six. When a customer asks what you hold about them, or asks you to delete it, every additional system is another search, another export, and another chance to miss something.
The same logic applies to your mailing list. Canada's Anti-Spam Legislation prohibits sending a commercial electronic message without consent, express or implied, and requires each message to carry prescribed identifying and contact information [5]. If consent was captured in one system and the sending happens from another, you need to be certain the unsubscribes flow back. Two lists that drift apart is the single most common way an otherwise careful business ends up sending mail it had no right to send. Our guide to PIPEDA requirements for small businesses goes through the obligations in order.
Reading a subscription for what it really costs
The monthly licence line is the part everyone looks at and the least interesting part of the total. Three costs sit underneath it.
Re-entry. Every system that does not share a contact list forces a human to type the same customer in twice. This cost is invisible because it never appears on a statement, and it is usually the largest of the three.
Switching friction. Ask, before buying, how your data comes out. A vendor that exports cleanly is a vendor you can leave, which changes the negotiation permanently. The domain-name system offers a useful comparison: under ICANN's Transfer Policy, registrants must be able to move their registrations between registrars, denials are permitted only in enumerated circumstances, and a transfer lock must be removed or a removal method made accessible within five calendar days [6]. Business software has no equivalent rule, so you have to write the equivalent into your own buying process.
Fees that appear later. Onboarding charges, per-seat minimums, mandatory support tiers. Canadian marketing rules take this seriously: under subsection 74.01(1.3) of the Competition Act, advertising a price that is not attainable because fixed obligatory charges have been excluded is itself a false or misleading representation, with a carve-out only for charges imposed under an Act of Parliament or a provincial legislature [3]. Tax on top is legitimate. A compulsory setup fee that surfaces at signature is not.
How to actually do the review
Pull twelve months of card and bank statements and list every recurring software charge. Beside each, write the category it belongs to and the name of the person who uses it. You will find three things: charges you cannot identify, two tools in one category, and at least one product nobody has opened this year.
Cancel in the order of least risk. Single-user, no unique data, no customer-visible output goes first. Anything holding customer records goes last, and only after you have exported the data and confirmed the export actually opens. Our software subscription audit walkthrough covers the sequence in more detail, and the suite versus best-of-breed comparison covers what to do with the overlaps you find.
Where we sit
MapleWorkSuite is built as a set of separate apps on one Canadian account rather than a single bundled product, which means the count is yours to control. You switch on what you need and nothing else appears on the bill. Contacts are shared across whichever apps you have enabled, so adding a second app does not create a second customer list to reconcile.
That design is a genuine trade-off and we will say so plainly: a specialist tool built for one industry will often beat a general one on depth. Where the suite wins is on the count — fewer logins, one contact list, one place to answer a privacy request. If your problem is that a specific job is done badly, buy the specialist. If your problem is that the same customer exists in five systems and none of them agree, that is the problem we were built for.
Either way, the right number of subscriptions is the number you can list from memory, explain the purpose of, and cancel cleanly. If you cannot do all three for a tool, you have found your starting point.