A landscaping company owner in the Fraser Valley spent $6,000 on an AI consulting engagement last year. The deliverable was a 38 page strategy document recommending a customer relationship management overhaul, a new website with chatbot integration, and a machine learning model to predict seasonal demand. The document was professionally formatted. It referenced industry benchmarks. It looked impressive. None of it was built. The owner read it twice, put it in a drawer, and went back to running the business exactly the same way. Six thousand dollars for a document that described a future state nobody was equipped to deliver. That outcome is avoidable if you know what questions to ask before signing.
The first question: do they build or just advise?
The AI consulting market includes two very different types of firms. Strategy consultants assess your business, identify opportunities, and produce a report or roadmap. Implementation consultants (or consultants who work with implementation partners) actually build the system you will use. Some do both. Many do only the first. The distinction matters because a strategy document without a build path is an expense, not an investment. Before the first meeting, ask directly: 'At the end of this engagement, will I have a working system or a document describing one?' The answer tells you everything about what you are buying. For context on what AI and automation actually mean in a small business setting, that article breaks down the terminology.
A good consultant can do both. The assessment informs the build. But if the engagement ends at the assessment with no clear path to implementation, the business owner is left holding a wishlist and no instructions.
Fixed pricing versus open ended billing
Enterprise consulting firms bill by the hour or by the day. A project scoped at 'four to six weeks' with a day rate of $1,500 can land anywhere between $30,000 and $45,000. The scope creeps. The timeline extends. The invoices keep arriving. Small businesses cannot absorb that risk. A consultant who serves small businesses should offer fixed pricing tied to a defined scope. You should know the total cost before work begins, not after. The scoping call exists to determine what the build requires, and the price is quoted from that. If the consultant cannot quote a fixed price after understanding your business, the engagement is not well enough defined to proceed. A breakdown of what AI consulting typically costs for a small business is here.
Who owns the finished system?
This is the question most business owners forget to ask. Some consultants build on proprietary platforms. The automation works, but it runs on the consultant's infrastructure. If you stop paying the monthly fee, you lose access. If the consultant closes the business, your system disappears. Others build on platforms the client owns: Make.com, Zapier, n8n, or native platform integrations. The client controls the account. The client can modify the system. The client can hire someone else to maintain it. Ownership is not a minor detail. It determines whether the automation is an asset you own or a subscription you rent.
Ask: 'When the engagement ends, do I own the system? Can I access and modify it without your involvement? Can I take it to another consultant if I choose to?' If the answer to any of these is no, understand what you are committing to before signing.
Red flags in the first conversation
The consultant leads with technology rather than business outcomes. If the first meeting is about platforms, AI models, and technical architecture rather than about your business, your processes, and your pain points, the engagement is being driven by the consultant's interests, not yours.
The consultant cannot name a small business client. Enterprise experience does not translate directly. A consultant who has delivered large scale digital transformation for a bank or a telecom has valuable expertise, but the constraints of a 10 person business are fundamentally different. Limited budget. No IT department. The owner is the decision maker, the project sponsor, and often the end user. Ask for examples of work done at your scale.
The consultant recommends enterprise software. Salesforce, SAP, Microsoft Dynamics, or any platform with a five figure annual licence is not built for a business doing $500,000 to $2 million in revenue. If the recommendation does not match your budget and team size, the consultant is applying a framework they know rather than a solution that fits. For Canadian small businesses specifically, the right platforms are affordable, accessible, and do not require a dedicated administrator.
The consultant is vague about timelines. A well scoped small business automation project has a defined start date, milestone dates, and a delivery date. Two to four weeks for most single automations. If the consultant says 'it depends' without being able to give a range after understanding your business, the scoping is incomplete.
Green flags worth looking for
The consultant asks about your existing tools before suggesting new ones. A good engagement connects to what you already use (your CRM, your accounting platform, your scheduling system) rather than replacing everything. They ask: 'Walk me through how this process works today.' They want to see the real workflow before proposing a change. If they recommend starting with the area of highest impact rather than the most technically interesting problem, that is a strong signal. A framework for deciding what to automate first applies here: the best consultants follow a similar logic.
The consultant explains the ongoing maintenance picture honestly. Every automation requires some maintenance. Platforms update. APIs change. Business processes evolve. A consultant who says 'set it and forget it' is either oversimplifying or planning to sell you a support contract later. A good consultant explains what maintenance looks like, how much time it takes, and whether you can handle it yourself or will need ongoing support.
The consultant is willing to start small. A $50,000 digital transformation programme is not the right first engagement with a new consultant. A single automation that solves one real problem, delivered in two weeks for a known price, tells you everything you need to know about how the consultant works. If they deliver well, expand the scope. If they do not, the cost of the lesson is manageable.
Five questions to ask before signing
One. Will I have a working system at the end of this engagement, or a strategy document? Two. What is the total fixed price, and what is included? Three. Do I own the system, and can I access it independently after the engagement? Four. Can you show me an example of a similar build you have done for a business my size? Five. What does ongoing maintenance look like, and what will it cost? These five questions eliminate most of the risk in choosing a consultant. A consultant who answers all five clearly and directly is worth a scoping conversation.
Common questions about hiring an AI consultant
Can I build automations myself without a consultant?
Yes, for simple automations. Platforms like Make.com and Zapier are designed for non technical users. Where a consultant adds value is in identifying the right process to automate, designing the workflow to handle exceptions, and building something robust enough to run unattended. If you have the time and interest to learn the platform, starting with one automation yourself is a reasonable approach. If your time is better spent on revenue generating work, a consultant gets you there faster. The AI readiness assessment helps determine which approach fits your situation.
How long should a small business automation project take?
A single automation (one workflow, one trigger, one output) typically takes one to two weeks from scoping to delivery. A more complex engagement involving multiple workflows, integrations, and training takes three to four weeks. Anything quoted at more than six weeks for a small business project should be examined closely. Either the scope is too broad for one engagement, or the timeline is padded.
What should I expect to pay?
For a single automation build, $1,750 to $3,500 CAD is a typical range depending on complexity. Multi workflow engagements with training and rollout run $4,750 to $8,750 CAD. Full pricing detail is here. Any consultant should be able to quote a fixed price after a scoping conversation. If they cannot, the scope is not clear enough to proceed.
What if the automation breaks after the engagement ends?
Ask about the support period included in the engagement. Most consultants include a window (typically 30 days) after delivery where issues are resolved at no additional cost. Beyond that, ask whether ongoing support is available and at what cost. The most important factor is ownership: if you own the system and it runs on a platform you control, you can always hire someone else to fix it. Understanding what an AI agent actually is and how it operates helps you maintain confidence in the system after handover. More about Highridge's approach and work is on the services page.