.png)
Retail personalization leaders grow revenue 10 percentage points faster than laggards. By the end of the decade, an estimated $570 billion in incremental growth will accrue specifically to the retailers who get this right (BCG, November 2024). Personalization leaders across sectors generate 40% more revenue from the activities they personalize than average performers, and 71% of consumers now expect it as standard (McKinsey, 2023).
Here's the part that should concern every Canadian shopping centre marketing director: we could not find a single publicly documented AI shopper segmentation program at a major Canadian retail REIT. Not Cadillac Fairview. Not Oxford Properties. Not RioCan, First Capital, or Ivanhoé Cambridge. The opportunity is sitting wide open.
Why Most Properties Are Still Treating Every Shopper the Same
The maturity gap in this industry isn't hypothetical, and it isn't evenly distributed. ICSC's February 2026 survey of retail real estate landlords' technology stacks found a genuine spectrum within a single industry. At one end, Cushman & Wakefield's Evan Walke described many smaller owners still running their operations on spreadsheets. At the other, Brixmor's Chief Information Officer Helane Stein described roughly 100 integrated proptech data feeds flowing into a single Salesforce CRM across the company's 63 million square feet of retail space. NewMark Merrill's CEO Sandy Sigal described a stack combining Wi-Fi data, smart camera analytics, and sentiment analysis run through an AI layer across the company's 13 million square feet (ICSC, February 2026).
Same industry. Wildly different starting points. Most Canadian properties are closer to the spreadsheet end than the integrated end.
CBRE's 2026 Canadian retail market outlook is direct about what that means competitively: "Leveraging AI for data-driven insights on consumers and their shopping habits will separate market leaders" (CBRE Canada, 2026). That's a forward-looking statement about who wins the next five years, not a description of common practice today.
What Segmentation Actually Requires
Real shopper segmentation isn't a marketing tactic layered on top of existing operations. It's built from data a property already owns, or should own, if it's collecting it properly: loyalty program activity, Wi-Fi opt-in data, mobile app usage, and event registration data.
The behavioural shift driving this is well documented. Third-party data reliability is declining across the industry, and first-party collection is now the central strategic asset (Deloitte Digital, Salesforce, Epsilon, 2024-2025). Deloitte Canada was explicit about what that means for retail specifically: "The value of third-party data and cookies is depleting. It's critical to obtain first-party data from customers at every interaction" (Deloitte Canada, October 2025).
The gap between demand and supply on this is stark. Over 70% of Canadian customers expect their loyalty programs to feel personalized. Only 45% of brands actually deliver it (Deloitte Digital, cited by Deloitte Canada, October 2025). Over 90% of Canadians are enrolled in at least one loyalty program (R3 Marketing, 2024, cited by Deloitte Canada). The data collection infrastructure is already in shoppers' hands. Most properties simply aren't building anything meaningful from it.
The Loyalty Data Case
Bond Brand Loyalty, a Toronto-based firm that has been running the industry's most comprehensive loyalty research for over a decade, surveyed more than 20,000 consumers across 360-plus programs for its 2024 report. The segmentation case it makes is specific: members who receive frequent, relevant communications spend 8% more of their category spend at that brand. Top-tier loyalty members show 66% satisfaction versus 44% for base-tier members, and 43% intent to increase spending versus 18% for base-tier members (Bond Brand Loyalty, 2024).
That's not a hypothetical benefit of segmentation. That's the measured difference between treating your loyalty base as one undifferentiated group and treating your highest-value members differently. BCG's research adds a sharper data point: the average consumer belongs to 19 loyalty programs but actively engages with only 9 (BCG, November 2024). The programs that survive that attention competition are the ones that feel like they're speaking to the individual, not broadcasting to the crowd.
What's Actually Being Built, and What's Still Unproven
Simon Property Group launched Simon+ in November 2025, its unified loyalty platform spanning more than 500 participating retailers across the portfolio (Simon Property Group, November 2025). Unibail-Rodamco-Westfield's Immersive Experiential Display network, deployed across 10 U.S. flagship centres, uses privacy-respectful sensor data to build what the company describes as 600 distinct shopper profiles per property (URW, April 2025, independently reported by two industry trade outlets).
Worth being direct about what these programs represent and what they don't. Both are real, confirmed deployments. Neither company has published an independently audited measurement of the actual lift these segmentation efforts produce. The infrastructure is being built. The proof of return, at least in public form, isn't there yet. That's a reasonable stage for the industry to be at. It also means the properties building this capability now are moving ahead of a competitive standard that hasn't fully formed.
The Canadian Regulatory Layer
Any Canadian property building segmentation from Wi-Fi, mobile, or behavioural data needs to build it correctly from the start, particularly if any property sits in Quebec. Law 25 explicitly defines profiling as the collection or use of personal data to assess preferences, interests, or behaviour, which is precisely what shopper segmentation does. Identification, geolocation, and profiling technologies cannot be enabled by default under Quebec law. Automated decisions require disclosure and a right to human review. Consent must be sought separately for each distinct purpose (Commission d'accès à l'information du Québec, official guidance).
PIPEDA, notably, has no equivalent statutory provision on automated decision-making or profiling outside Quebec, though the Privacy Commissioner of Canada has pushed for reform in that direction. That's a real regulatory asymmetry for any national operator to plan around, not just a Quebec-specific footnote.
Where to Actually Start
The properties with the clearest advantage right now aren't the ones with the most sophisticated AI. They're the ones that have made a first-party data collection point, a loyalty program, an app, a Wi-Fi opt-in, actually work, and then built even basic behavioural segmentation on top of it. The bar in Canadian retail property right now is genuinely low. Building past it doesn't require Simon-scale infrastructure. It requires deciding that shoppers aren't one group and building the smallest possible system to prove it.
Interested in what a first-party segmentation strategy could look like at your property?
