AI leasing is moving the needle on NOI
QuadReal Property Group hit a wall with slow, legacy accounting software for its residential portfolio. After a 2023 pilot, the Vancouver firm rolled out Funnel Leasing to centralize and automate leasing across 10,000+ Canadian rental suites. Funnel's virtual leasing agent cut response times for tour requests from days to minutes. The result: a 33% lift in tour-to-lease conversion, shorter vacancy windows, and higher rental income.
Aggregation helped too. Instead of managing one unit at a time, staff work a portfolio view and prospects deal with a single point of contact. Funnel's CEO says another customer, Camden Property Trust, is realizing US$4-5 million in annualized savings through leaner admin and better occupancy. The pattern is clear: faster follow-up and fewer handoffs translate to fewer empty days.
Adoption is climbing, but it's still early
According to Statistics Canada, 18.6% of real estate rental and leasing companies planned to deploy AI software in Q2 2025, up from just under 11% in 2024. That's measured progress, not a stampede. Source: Statistics Canada.
Proptech startups keep multiplying: more than 590 in 2025, up from about 530 in 2024, with most building AI and machine learning products. Many target leasing, property operations, and building automation, per a recent report from Proptech Collective.
Where AI pays off now
- Leasing and CRM: instant responses, lead-to-tour booking, and pre-qualification handled by virtual agents.
- Portfolio routing: one agent across many buildings, less context switching, fewer handoffs.
- Rent collection and admin: automated reminders, document handling, and data entry.
- Market analytics: demand forecasts that inform pricing, incentives, and marketing spend.
- Smart building controls: systems that auto-adjust to internal and external conditions to save energy.
Energy and OPEX: measurable wins
Yardi remains the dominant property management suite in Canada, but specialized tools are gaining ground. Large owners are also using AI to trim emissions, improve comfort, and tighten security across office, retail, residential, and industrial.
Cadillac Fairview began deploying algorithms across its office towers in 2023, drawing on thousands of sensors to tune HVAC, predict maintenance and cleaning, and alert security to crowd formations or potential weapons via video analysis. On a portfolio topping 30 million square feet, the company reports 4-6% annual energy savings on a budget north of $110 million, with savings showing up within 30 months. Maintenance work orders are down 10-15%, while software licensing typically runs about $0.02 per square foot per year. Labor isn't necessarily reduced; it's redeployed to higher-value work.
Costs, constraints, and the funding climate
AI adoption in Canadian commercial real estate is still at an early stage. The blockers: compute costs, training, and governance. Tools exist, but they're not cheap-so fit-to-need matters.
Proptech funding in Canada hit $450 million in 2025, down from $800 million in 2024 (skewed by one large deal). Tighter capital is forcing sharper prioritization. As one investor put it, teams are picking the few areas where AI can have the biggest impact-and testing fast to learn what sticks.
A 90-day playbook to test, learn, and scale
- Define the business case: pick 1-3 buildings or a region, and lock KPIs (response time, tour-to-lease, vacancy days, occupancy, cost per lease).
- Map the leasing funnel: inquiry to signed lease. Identify every handoff and data gap.
- Centralize lead intake: route all channels (ILS, website, phone) into one CRM.
- Pilot a virtual leasing agent: automate inquiry responses, tour scheduling, FAQs, and pre-qualification. Keep humans for edge cases.
- Move to portfolio-level routing: one agent or team handles multiple assets; reduce back-and-forth and duplicate work.
- Standardize follow-ups: SLAs in minutes, not hours. A/B test scripts and cadence for 30 days.
- Tighten operations data: integrate pricing, concessions, and unit availability to avoid mismatches.
- Run a building automation pilot: baseline energy, apply algorithms for temperature and occupancy, and monitor drift weekly.
- Vendor checklist: open APIs, SOC 2/ISO security, audit logs, role-based access, transparent model behavior, and a clear exit plan.
- Compliance and privacy: document consent, data retention, and model training boundaries.
- Review at day 30/60/90: compare KPIs, calculate savings versus software cost per square foot, and decide on scale-up.
Metrics that matter
- Lead response time (minutes)
- Tour scheduled rate and tour-to-lease conversion (%)
- Vacancy days and effective rent
- Occupancy and economic occupancy
- Admin hours per lease and cost per lease
- Energy intensity (kWh/sq. ft.) and utility spend variance
- Maintenance work orders per 100,000 sq. ft.
- Software cost ($/unit or $/sq. ft.) versus quantified savings
What this means for owners and operators
The play is simple: compress response times, remove handoffs, and let software handle repetition. Start narrow, measure hard, and reassign people to work that lifts occupancy and experience. The payback shows up in fewer empty days, steadier rent rolls, and lower OPEX.
For practical how-tos and tooling examples across leasing and building operations, explore AI for Real Estate & Construction.
Your membership also unlocks:
