No one has published a return-on-investment number for this. What's published instead is a genuinely mixed picture worth reading honestly.
Short answer
No Canadian study publishes a return-on-investment figure for small-business AI use, so there is no dollar answer to give. What Statistics Canada's size breakdown does show is that Canada's smallest businesses adopt AI at nearly the national average rate, face some barriers less often than larger firms, and invest far less in training around it — a mixed picture, not a verdict either way.
Contrary to an assumption that AI is mainly a large-company tool, Statistics Canada found: “Among businesses with 1 to 4 employees, about one-fifth (19.9%) reported using AI in producing goods or delivering services in the last 12 months” — close to the 19.2% national average, with data analytics (32.5%) and text analytics (29.4%) the most common applications. At the same time, 41.4% of businesses with 1 to 4 employees say AI “is not relevant to the business”, against 21.3% of businesses with 100 or more employees — so small businesses split more sharply between using AI in specific areas and not seeing a fit for it at all.
Smaller businesses report some barriers less often: cybersecurity or privacy concerns limited AI use for 11.6% of businesses with 1 to 4 employees, compared with 30.0% of businesses with 100 or more employees, per the same StatCan release — though that gap may partly reflect lower awareness of the risk rather than lower actual exposure to it. What is unambiguous is the training gap: only 24.0% of AI-using businesses with 1 to 4 employees trained existing employees and 15.6% trained executives, against 68.1% and 51.7% among businesses with 100 or more employees; only 8.2% hired AI-skilled staff and 10.7% used external consultants or vendors, against 32.8% and 30.2% for larger firms. In practice, a small business adopting AI is far more likely to be doing so without formal training or outside help than a large one is.
With no published return-on-investment number for Canadian small businesses, the honest way to weigh this is structural rather than numeric: worth-it depends on whether the task you would hand to AI is one of the ones small businesses are already using it for successfully — data analytics and text work, per the figures above — and whether you can build in the same “someone checks the output” step that a larger, better-resourced team gets through formal training. See how AI is actually being used in Canadian small business for more on that fit, and whether you need AI to stay competitive for the sector-level version of this same question.
The training gap is real. Closing it doesn't require a large-company budget to start.
