.png)
RioCan posted a blended leasing spread of 25.8% in Q1 2026, a company record, with new leasing spreads hitting 58.5%. Primaris renewed leases at a 7.4% spread in Q2 2026 with 88% retention. SmartCentres closed FY2025 at 98.6% occupancy, calling it industry-leading, with rent growth on lease extensions of 8.4% excluding anchors (RioCan Q1 2026 Report to Unitholders; Primaris Q2 2026 disclosure; SmartCentres FY2025 results).
Those are genuinely strong numbers. Here's what's missing from every one of those disclosures: not one of these REITs credits predictive analytics, AI, or data-driven modelling for producing them.
That's not a criticism. It's the honest starting point for this conversation. The technology to bring real predictive rigour to lease strategy exists and is being adopted at the brokerage level. The REITs setting the performance benchmarks in Canada right now haven't yet said, in any public disclosure, that this is how they're getting there. That gap between what's possible and what's confirmed is exactly where the opportunity sits for whichever operator moves first.
How Lease Decisions Still Get Made
Deloitte's 2025 Commercial Real Estate Outlook, surveying more than 880 C-suite CRE executives, found 76% of firms describe their AI adoption as still in the research or pilot stage. Only 14% report having well-structured data collection and governance in place. That's consistent with Deloitte's prior year survey, where 61% said their core technology infrastructure still relies on legacy systems.
A recent legal industry analysis put it plainly: retail leasing decisions have historically been "as much art as science, a blend of broker relationships, market experience, and static census data" (Hinckley Allen, 2026). Lease terms have also compressed structurally, with typical retail leases now running five to six years with three-year renewal cycles (MRI Software, via ICSC Exchange), meaning the rent-reset decision comes up more often and the cost of getting it wrong on instinct compounds faster.
Where the Technology Actually Is
The three largest global brokerages are all on record investing directly in this. CBRE's CEO confirmed the firm uses its Ellis AI platform to manage transaction data and analyze contracts. JLL's CEO pointed to a decade-long proprietary data and technology build through JLL Spark. Cushman & Wakefield's CEO was direct: "AI will create winners and losers" (CoStar News, February 2026). Proptech investment backs up the scale of that shift: $16.7 billion was invested in real estate technology in the year to early 2026, a roughly 68% year-over-year jump, with $1.7 billion invested in January 2026 alone compared to $615 million in January of the prior year (Center for Real Estate Technology and Innovation, cited by CoStar News, February 2026).
On the analytics side, Green Street, the industry's leading independent REIT research firm, integrated Placer.ai's foot traffic data directly into its analyst platform for institutional clients in March 2026. That's a genuine third-party signal that location intelligence data is becoming a standard input to how the capital markets themselves evaluate retail real estate, regardless of whether any specific landlord has adopted it yet for lease decisions.
The general practice of using data signals, sales trends, foot traffic decline, and payment patterns to flag tenant financial distress before it becomes a default is well established at the lending and research level. S&P Global Market Intelligence has described foot traffic data as usable to identify properties at elevated default risk.
CoStar and Trepp both offer credit and default-risk models built into their CRE data platforms. That's a mature, cross-corroborated category of practice. What's not established, in any public disclosure we could find, is a named Canadian REIT using this kind of model specifically to manage its own lease renewal or tenant retention strategy.
The Real Test Case: HBC and Co-Tenancy Risk
If predictive analytics were going to prove its value anywhere in the Canadian market recently, the Hudson's Bay collapse was the test.
HBC filed for CCAA creditor protection in March 2025 and fully liquidated its Canadian stores by year-end. The co-tenancy exposure across the industry was real and court-documented. Primaris disclosed 35 co-tenancy clauses naming HBC across its portfolio, covering roughly $11.6 million in annual gross rental revenue, about 1.4% of its annualized minimum rent (Primaris HBC Exposure Update, March 2025). RioCan recognized a $208.8 million valuation loss tied to its HBC joint venture and later sought receivership over that JV. Cadillac Fairview, which had both landlord and lender exposure to HBC, including a $200 million loan extended in 2023, litigated over a forced lease assignment and sued HBC's U.S. arm for more than $75 million in related losses (Torys LLP; Financial Post; Globe and Mail, 2025).
Primaris's own language in its HBC disclosure is instructive: the company said it had "been preparing for this day for years," drawing on experience from the earlier Zellers, Target, and Sears exits. That's contingency planning built on institutional memory and pattern recognition across prior anchor failures, which is genuinely valuable. It is not, based on anything Primaris has disclosed publicly, a stated predictive-analytics methodology.
Placer.ai markets a Void Analysis tool explicitly built around co-tenancy and cannibalization risk modelling. Whether it or a comparable tool was used by any of the REITs navigating the HBC fallout isn't something we could confirm from public sources. The technology category exists. Its application to this specific crisis, at least in disclosed form, doesn't yet have a documented Canadian case study attached to it.
What This Means for Lease Strategy Going Forward
The market data is unambiguous: Canadian retail real estate is performing well right now. CBRE's Q2 2026 cap rate report found retail among the sectors leading cap rate compression nationally. Colliers describes retail fundamentals as steady. Four of the five major Canadian retail REITs are posting occupancy in the 97 to 99% range with positive leasing spreads across the board.
The honest opportunity is this: the operators posting those numbers are doing it with tools that are, by Deloitte's own survey data, still mostly in pilot stage industry-wide. The brokerages are investing heavily in AI for their own transaction and lease-drafting workflows. Green Street is integrating location intelligence data into how it evaluates REITs. None of that has yet translated into a Canadian REIT publicly stating that predictive modelling drives its renewal decisions, its rent-setting, or its tenant risk management.
For a mid-market Canadian operator, that's not a reason to wait. It's a reason to move before the gap closes. The properties that build genuine predictive capability into lease strategy now- tenant performance forecasting, foot traffic-informed rent setting, early distress signals ahead of the next anchor disruption- aren't just improving internal processes. They're building a capability that the market leaders haven't yet claimed publicly, which means the competitive window is still open.
Interested in what a data-informed lease strategy could look like at your property?
