The assessment costs $999, takes ten business days, and ends with a written document plus a 60 minute walkthrough. If you then go ahead with any automation project with me, the whole $999 comes off it. Not a percentage and not a credit note. The entire fee.
Canadian consultancies that publish rate cards quote $1,500 to $5,000 for a readiness assessment, with several opening at $2,000 or $3,500. This one is priced under the floor of that band on purpose, because its job is to be an easy decision rather than a considered purchase.
What is in the document
- An inventory of every repeated task in the business that a machine could plausibly do, gathered from watching the work rather than from a questionnaire.
- The three worth doing first, each with an hours-per-week figure and a dollars-per-year figure attached to it.
- The specific tools each one needs, named, with their real subscription costs in Canadian dollars.
- A build estimate for each, in the same ranges published on the workflow automation and AI agents pages.
- The ones I recommend you do not do, and why. This is usually the most useful section.
- Where personal information enters each process, because that decides which tools are eligible before anybody discusses features.
It is written so another supplier can quote from it. That is deliberate. An assessment whose conclusions only make sense if the author does the work is a sales document, and you would be right to discount it.
The ten days, step by step
| Stage | What happens | Your time |
|---|
| Day 1 | Kickoff call. Headcount, systems, the busiest hour of the week, and what breaks during it. | 45 minutes |
| Days 2 to 4 | Process capture. I watch or record how three to five recurring jobs are actually done, not how the manual says they are done. | 2 to 3 hours total, spread out |
| Days 4 to 6 | Systems inventory. Every piece of software you pay for, what it holds, and what it can and cannot talk to. | Read-only access or a screen share |
| Days 6 to 8 | Costing. Hours per week per task, loaded hourly cost, and the tool subscriptions each candidate build would add. | None |
| Days 8 to 10 | Writing, then a 60 minute walkthrough with whoever will actually decide. | 60 minutes |
Why the diagnosis is the hard part
Statistics Canada asked businesses that were not adopting AI why not. Cost was fourth at 10.6% and cybersecurity third at 13.4%. The leading answer, from 40.0% of them, was that AI is not relevant to what they do.
That answer is usually the product of a bad mental image. The owner pictures a chatbot on the website, because that is what the advertising shows, and correctly concludes it would not move the needle. In the same building somebody is retyping the same order into two systems 40 times a week, which is 30 hours a month, and nobody has counted it because it has always been done that way.
Counting it is most of the work. Once the hours are on paper the decision usually makes itself, in either direction.
When I will tell you not to buy this
- Under about five people, where the repeated work is real but does not yet total enough hours to pay back a build.
- When you already know exactly what you want built. Then you want a quote, not an assessment, and that is a free call.
- When the actual problem is that nobody has written down how the work is done. That has to happen first, and doing it often fixes half the problem without any software.
- When you have already bought a tool and nobody uses it. That is an adoption problem, and AI policy and training is the cheaper answer.
Funding
The Ontario Centre of Innovation runs a Digital Modernization and Adoption Plan worth up to $15,000 toward exactly this kind of planning work, open to Ontario for-profit SMEs with one to 499 employees, on a first come first served basis while funds last. The application is yours to run and I will not promise an approval I do not control. Tell me on the first call if you are applying and I will structure the assessment around what the application asks for. There is more detail in the note on the DMAP grant and Ontario AI funding.