Before the Demo
Sales demonstrations are designed to show AI tools at their best. They typically use carefully selected examples, ideal inputs, and favourable conditions. Before you watch a demo, prepare your own questions so you can evaluate the tool against your actual needs rather than the vendor's preferred narrative.
The questions in this guide are organized by category. You do not need to ask all of them in every situation — use the ones most relevant to your organization and the specific tool you are evaluating.
Pricing Model Questions
Understanding the full pricing model before committing is essential. See also Understanding AI Costs: A Plain-Language Breakdown for a detailed explanation of AI pricing models.
Questions to ask
- Is pricing per user, per usage, or a flat fee? How does each model work in practice?
- What is included in the base price, and what requires a higher tier?
- Are there usage limits? What happens if we exceed them?
- How have prices changed over the past two years? What notice is given before price increases?
- Are there setup fees, onboarding fees, or professional services fees not included in the subscription?
- What is the minimum contract term? What are the cancellation terms?
- Are there discounts for annual payment versus monthly? What are the trade-offs?
Data Usage Questions
Understanding what the vendor does with your data is one of the most important areas to investigate.
Questions to ask
- Is the data we provide used to train or improve your AI models?
- Can we opt out of having our data used for training? What is the process?
- How long is our data retained after we cancel our subscription?
- Can we request deletion of our data? What is the process and timeline?
- Do you share our data with third parties? Under what circumstances?
- What data does the tool require access to in order to function?
- Is there a data processing agreement available for us to review?
Privacy Questions
For Canadian organizations, privacy compliance is a legal obligation, not just a best practice.
Questions to ask
- Where are your servers located? Is data processed or stored outside Canada?
- Are you compliant with Canadian privacy legislation (PIPEDA, provincial laws)?
- Do you have a privacy policy that covers how you handle customer data?
- What security measures protect our data?
- Have you had any data breaches? How were they handled?
- What is your process for notifying customers of a security incident?
- Do you have SOC 2 or equivalent security certifications?
Vendor Lock-In Questions
Once you build workflows around a specific AI tool, switching becomes costly. Understand your exit options before you commit.
Questions to ask
- In what format can we export our data if we decide to leave?
- Are there any proprietary formats or structures that would make migration difficult?
- What integrations or workflows would we need to rebuild if we switched tools?
- What is the process for cancelling our subscription?
- What happens to our data after cancellation? When is it deleted?
- Has your company been acquired or merged in the past? What happened to customer contracts?
Integration Questions
AI tools rarely work in isolation. Understanding how a tool integrates with your existing systems is essential for realistic implementation planning.
Questions to ask
- What systems does this tool integrate with natively?
- What integrations require additional configuration or cost?
- What technical expertise is required to set up integrations?
- What happens to our integrations when you release updates?
- Is there an API available? What are the API usage limits and costs?
- What is the typical implementation timeline for an organization of our size?
Support Questions
The quality of support matters most when something goes wrong. Understand what you are entitled to before you need it.
Questions to ask
- What support channels are available (email, phone, chat)?
- What are the support hours? Is support available in our time zone?
- What is the typical response time for support requests?
- Is there a dedicated account manager, or is support handled by a general team?
- What training resources are available for our staff?
- Is there a user community or knowledge base we can access?
- What is your uptime guarantee, and what compensation applies if it is not met?
Red Flags To Watch For
Some vendor behaviours during the sales process are worth noting carefully.
- Vague answers about data usage. If a vendor cannot clearly explain what they do with your data, assume the answer is unfavourable.
- Pressure to sign quickly. Legitimate vendors do not need you to commit before you have had time to review terms carefully.
- Reluctance to provide a data processing agreement. This is a standard document for any vendor handling personal data. Reluctance to provide one is a meaningful signal.
- Pricing that changes significantly between the demo and the contract. Understand exactly what you are buying before signing.
- No clear answer on data deletion. You should be able to get a clear, written commitment on what happens to your data when you leave.
- Testimonials but no references. Ask to speak with current customers in a similar industry or of similar size. A vendor confident in their product will facilitate this.
Key Takeaways
- Prepare your questions before the demo — do not let the vendor set the entire agenda.
- Understand the full pricing model, including what happens if usage grows beyond initial estimates.
- Know exactly what the vendor does with your data and whether you can opt out of training data use.
- Understand your exit options before committing — data export, migration complexity, and cancellation terms.
- For Canadian organizations, data residency and privacy compliance are non-negotiable considerations.
- Ask difficult questions before signing contracts. A vendor who cannot answer them clearly is telling you something important.