.png)
Simon Property Group's retail media network now generates $6.4 billion in annual revenue, built on Wi-Fi-based location data collected across more than 200 shopping centres and over 2 billion annual visits. In November 2025, the company layered Simon+, a unified loyalty platform spanning more than 500 participating retailers, on top of that same data infrastructure.
That's real, confirmed, large-scale infrastructure. What we couldn't find, anywhere in Simon's public disclosures or independent trade coverage, is a published number showing what that personalization infrastructure actually lifts in redemption, conversion, or revenue per campaign. The program is real. The proof of return, at least in public form, isn't.
That gap between infrastructure and evidence runs through almost this entire topic, and it's worth being honest about before your property invests in either direction.
What's Actually Proven About Personalization's Value
The most solid, independently corroborated finding in this space is older than most people assume, and still holds up: Epsilon's 2018 research, cited independently by both McKinsey and Deloitte in more recent analyses, found consumers are 80% more likely to make a purchase when brands offer personalized experiences. That's genuine cross-firm corroboration, which is rarer in this space than it should be.
Beyond that specific stat, most of the "3x returns" and "15% revenue lift" figures circulating in personalization marketing trace back to single proprietary research programs, BCG's own index, McKinsey's own analysis, repeated verbatim across dozens of secondary blogs. Repetition across the internet isn't the same as independent corroboration. We could not find a Gartner or Forrester study establishing a specific, verified redemption-rate gap between personalized and generic promotions.
What Canadian Shoppers Actually Say They Want
Here's where the picture gets genuinely encouraging for a property willing to build this properly. The Canadian Marketing Association's 2025 research found 73% of Canadians prefer relevant, personalized ads over generic ones, up 12 percentage points year over year, and 73% say they're comfortable trading their data for offers or discounts specifically.
That comfort has a condition attached. The Office of the Privacy Commissioner of Canada's own 2024-2025 public opinion research, surveying 1,500 Canadians, found 89% are concerned about privacy generally, and only 40% trust businesses to respect their privacy rights. Canadians want relevance. They don't automatically trust the businesses asking for the data required to deliver it. The properties that earn that trust, through clear disclosure and an obvious value exchange, get access to a genuinely receptive audience. The ones that don't are asking for data from a population that's already skeptical by a two-to-one margin.
Where AI-Driven Pricing Crosses a Line
Personalized promotions and personalized pricing are not the same thing, and the second one carries real, documented risk that the first doesn't.
The U.S. Federal Trade Commission's Surveillance Pricing Study, published in January 2025, and its follow-up enforcement policy statement in August 2026, put regulatory weight behind a consumer instinct that Gartner's own October 2024 survey of 303 consumers confirmed directly: 68% feel taken advantage of by dynamic or personalized pricing, and 80% say they trust brands with consistent, transparent pricing more. A joint investigation by Consumer Reports, Groundwork Collaborative, and Consumer Reports' own testing, published in December 2025 and independently corroborated by multiple news outlets, found price differences of up to 23% on identical grocery items depending on the shopper's profile, with an average 8.4% difference across full carts.
This is the line worth drawing clearly for any property or tenant considering AI-driven pricing: a personalized offer that gives a loyalty member an extra discount is functionally different, and far less risky, than a system that charges two different shoppers two different base prices for the same item based on what it infers about them. The first is a promotion. The second is what's currently drawing FTC scrutiny and state-level legal action in the US.
The Canadian legal picture on this is genuinely unsettled. A peer-reviewed analysis published in the Canadian Bar Review in 2024 found no Canadian statute or court decision currently addresses algorithmic personalized pricing directly, unlike the EU, which mandates disclosure. The author argues PIPEDA's existing consent and reasonable-purpose requirements likely already prohibit undisclosed personalized pricing, but that's her legal analysis, not a settled ruling. Whether Quebec's Law 25 disclosure requirement for automated decisions extends to personalized discounts and promotions specifically, as opposed to decisions like credit or insurance, is an open question with no CAI guidance we could find confirming it either way. Treat this as a live legal risk to watch, not a resolved compliance checklist.
What a Canadian Property Has Actually Built
Cadillac Fairview's LiVE by CF platform uses AI-powered product search, built with Adeptmind, across all 19 CF properties, letting shoppers search inventory across the mall's tenants from a single interface. It's a real, confirmed deployment. It's worth flagging honestly that the substantive public coverage of it dates to 2019 through 2022; if the program has evolved since, that update hasn't surfaced in current trade coverage we could verify.
That's the state of the Canadian market right now: one confirmed, if dated, product-search deployment at CF, and no current, corroborated example of a Canadian mall operator running AI-personalized promotions across its tenant base the way Simon is building toward with Simon+. The properties that build this capability now, honestly and with proper disclosure, aren't following an established Canadian playbook. They're building one.
What This Actually Means for a Property Considering This
The evidence supports three specific things clearly: personalization itself drives real purchase intent when it's genuinely relevant, not creepy. Canadian consumers are more receptive to it than the privacy headlines suggest, provided the value exchange is obvious, and the disclosure is clear. And personalized pricing, as distinct from personalized promotions, carries real and rising regulatory risk that a property or tenant should treat with real caution rather than adopting quietly.
What the evidence doesn't yet support is any specific claimed redemption or revenue lift from AI personalization tools, whether from a vendor or a competitor's case study. If someone shows you a number, ask where it comes from, the same way we'd ask you to ask about any of the vendor claims we've flagged across this whole research process.
Interested in what a properly disclosed, genuinely value-exchanged personalization program could look like at your property? BOOK A MEETING
