In real estate, air rights are the value of the space above the ground, owned and sold separately from it. Stratus is the air rights of the portfolio: the data, media, and commerce layer above the physical floor, monetized apart from rent. Across 95+ properties, that layer is an audience no retail-media network can replicate.
Real estate already knows how to sell the space above the ground separately from the ground itself. That is what an air right is. Stratus is the air rights of the portfolio: the platform layer that monetizes audiences, data, and commerce above the physical floor, owned and shared apart from rent. Carving out that interest is necessary, but it is not the hard part. The hard part is the data infrastructure that turns physical-retail behavior into something an AI agent can act on, and that is something GGP can build and own, drawing on a pattern Brookfield already operates. It does not have to invent the approach: the reasoning-era methodology, entity resolution and a knowledge, context, and decision graph under versioned semantic governance, is what Brookfield runs in institutional capital practice today. The institutional knowledge that produced it is inside the organization, so GGP extends a pattern that works rather than inventing an unproven design. What it still has to supply is the one thing that pattern has never seen, the physical-retail consumer.
The sun is GGP's own data infrastructure, something GGP can build and own. The portfolio is the irreplaceable mass beneath it. Brookfield's Data Infrastructure is the proven blueprint to follow, drawn on by choice, not a dependency. The methodology is proven; the mass is GGP's. Click the core to open the data estate.
95+ assets, the consumer traffic they generate, the tenant mix that shapes it. The irreplaceable gravitational source, footfall no card network, AI lab, or competing landlord can synthesize. Every data layer is a recording of a physical visit.
GGP can build and own its data infrastructure: entity resolution, a knowledge, context, and decision graph, versioned semantic governance. The engine that turns the air rights into media, loyalty, attribution, and commerce. GGP-owned, not rented from anyone.
GGP does not have to invent the reasoning-era methodology. It is proven in institutional practice, the same pattern Brookfield is applying in Brookfield's Data Infrastructure, which GGP draws on and interoperates with by choice. Optionality and a head start, not a reliance.
"GGP owns the mass and can build the engine. The methodology is proven. No competitor has both."
If Stratus is the air rights of the portfolio, the commercial model is how those rights are valued and split. GGP's framework is explicit about the prize: move from a landlord collecting rent on the ground to a platform monetizing audiences, data, and commerce above it. The near-term revenue argument is the media business. The structure protects the ground-floor economics first, then gives Stratus the majority of the new value created above them.
Commercial model and revenue framing in this section draw on GGP's Stratus Intelligence Brief.
Retail media across the digital-screen and DOOH network, sponsorships, and events. The revenue that already exists at the property level, and the base the preferred return is set from.
Loyalty, attribution services, and data products. Higher-margin and identity-dependent: the streams that grow once enrollment and the data estate are in place.
Agentic commerce, fulfillment, and AI-enabled transaction fees. The Retail 4.0 revenue the governed data layer makes possible.
Each property keeps a preferred return equal to its actual 2026 advertising revenue, paid before any sharing. The ground-floor economics are fully protected, which is what makes the lender and JV-partner conversations straightforward.
Above the preferred-return line, the new air-rights value splits 25% to the property and 75% to Stratus. The property contributes traffic, location, and infrastructure; Stratus contributes technology, data, sales, attribution, loyalty, and commerce.
| Illustrative, per property | Amount |
|---|---|
| 2026 advertising revenue (the baseline) | $1.0M |
| Future total revenue generated | $5.0M |
| Step 1: preferred return to property | $1.0M |
| Incremental revenue to split | $4.0M |
| Step 2: property share (25%) | $1.0M |
| Step 2: Stratus share (75%) | $3.0M |
| Property receives | $2.0M |
| Stratus net (after ~$1.0M cost) | $2.0M |
Conceptual and subject to legal and tax review. Properties with large one-off buys are normalized before the split.
"Retail media is the fastest-growing line in advertising, and it was built online. Simon+ and Westfield are racing for the physical audience. The preferred-return structure lets GGP move now, without disturbing property or lender economics."
A competitive landscape is the market, not a self-assessment, so this is competitors only, in two frames that compete on different axes. Direct peers are landlords building the same retail-media + loyalty model. Adjacent platforms are horizontal players that compete for the same advertiser budgets or could disintermediate the transaction. Stratus's own positioning is stated separately, against the landscape rather than inside it.
| Platform | Identity & enrollment | Transaction layer | Data governance | AI / agentic readiness | Posture toward Stratus |
|---|---|---|---|---|---|
Simon+ Simon Property Group · Nov 2025 |
Consolidation of 3 legacy programs (Mall Insider, VIP Shopper Club, ShopSimon Rewards). 500+ retailers enrolled. Card-linking available. | Card-linked purchases + ShopSimon.com. Wi-Fi behavioral layer via Adentro. | Retrofitting Three metadata standards being consolidated. Consent inherited, not designed. |
Building First-party data for RMN targeting. No agentic layer described. |
Closest peer. Ahead on enrollment, but carrying metadata-consolidation debt from merging three legacy programs, debt a clean build avoids. |
Westfield Rise URW · US expansion Apr 2025 |
550M annual visits. 40M online consumers. GDPR-compliant but largely anonymized. IXD network for audience measurement. | Drive-to-store measurement via Rise+ Data. DOOH-led. No owned transaction layer. | GDPR-anchored Explicit but constraining for US enrichment strategies. |
Building AI-driven campaign optimization. No transaction-linked layer. |
Peer. Measurement-led with strong reach, but no owned transaction layer and so no commerce layer to extend into. |
| Platform | Category | What they own that Stratus does not | Posture / duality |
|---|---|---|---|
Mastercard / Visa Commerce Media · Oct 2025 |
Card networks | Cross-merchant observed transactions, 500M+ enrolled, 160B+ txns/yr, full wallet visibility across all categories. | Partner & competitor Has the transaction; lacks the physical context, property, tenant mix, visit pattern. The data-partnership conversation is also a competitive one. |
Retailer media networks Nike Digital Commerce · Nordstrom Media Network |
Brand-owned RMNs | Deep first-party loyalty, SKU history, member profiles, activity data, inside their own four walls. | Partner & competitor Clean-room partner and a competitor for ad budget. The same entity sits in Layer 3A (partner) and in this landscape (competitor), a duality to manage, not hide. |
Amazon Ads Endemic retail media at scale |
Retail-media benchmark | Transaction-rich audiences and the scale advertisers benchmark everyone against; an agentic-commerce frontrunner. | Benchmark Not a physical-mall competitor, but sets the budget bar and the agentic expectations Stratus is measured against. |
A standalone scan would mark Stratus "pre-build" across governance and AI-readiness, its weakest line. That reading is wrong. The decade-long, hard part (entity resolution, semantic governance, decision lineage, an agent-ready data layer) is a proven pattern, the reasoning-era methodology, which GGP can build and own on its own infrastructure, with the benefit of the Brookfield blueprint, not an unproven bet. Stratus is pre-build only on consumer enrollment: a real cold-start, but a narrow one. Every competitor below would have to invent both halves.
Stratus owns which property, which tenant mix, which visit pattern, which experience surrounded a purchase. The card networks see the transaction but not the context, and that context cannot be reconstructed from transaction data alone.
Stratus sees the full cross-tenant visit; each tenant loyalty program and RMN sees only its own transaction. The combined cross-tenant behavioral record is something no single retailer can assemble.
GGP can build and own its knowledge, context, and decision graph, with entity resolution and semantic governance, on the reasoning-era methodology Brookfield already operates in institutional capital practice. The blueprint is not a specification to be commissioned. It exists, it works, and the institutional knowledge that produced it is inside the organization. No retail-media competitor has that and the physical mass to build on.
Because the engine follows a proven pattern, the net-new work is the consumer enrollment and consent layer the institutional methodology never had to carry. That is where the JV and lender governance concentrates, and where Stratus earns its position.
Stratus accumulates data from sources with different ownership profiles, trust levels, and value trajectories. Treating them as a single "data asset" is the mistake that produces confident attribution errors and ungovernable datasets. A source's layer is determined by who owns the consumer consent and relationship, not by how valuable the consumer is or how the data is accessed. Everything else (identity resolution, observed vs. modeled, signal type, partner-or-competitor status) is an attribute, not a layer. That rule keeps the one thing Stratus actually owns, Layer 1A, distinct from everyone else's first-party data.
Co-enrollment promotes a Layer 3A partner's member into an owned Layer 1A identity, the only mechanism that moves data inward. The partner's underlying file stays in 3A.
This is the air-rights inventory: the data above the physical floor. Two altitudes, one system: at the system level, these resolve into GGP's own data infrastructure; here, at the data level, these are the layers that make up the rights. The sun in the diagram above is that engine; this estate is what feeds it.
The only data Stratus owns outright, the enrollment record, the consent, and the behavior Stratus directly observes. It is the only layer Stratus can use unilaterally, license, or train on. Partner loyalty programs (tenant, travel, payment) are not 1A, they live in Layer 3A. Co-enrollment is the one pathway that promotes a partner's member into an owned 1A identity.
| Source | Signal type | Value | Notes |
|---|---|---|---|
| Enrolled consumer profile | Opt-in identity + consent | Highest quality. Does not exist yet at scale. The relationship Stratus owns and controls. | |
| App / web behavior | Browse, save, wish-list, offers | Stratus-captured digital engagement. Closes the browse-vs-buy loop when tied to a visit. | |
| Receipt / card-link (Stratus) | Self-submitted or linked spend | Transaction evidence the consumer gives to Stratus directly. Distinct from a card network's view of the same purchase. | |
| Cross-visit history | Visits, dwell, cross-tenant path | Stratus-observed behavior of an enrolled consumer across the portfolio. The record no single tenant can assemble. |
| Source | Signal type | Value | Notes |
|---|---|---|---|
| Wi-Fi signals | Device presence, dwell, path | Anonymized without identity link. High value if linked to 1A. Low value standalone. | |
| Foot traffic sensors | Volume, flow, zone dwell | Passthrough data. Supports attribution but not personalization without identity resolution. | |
| POS passthrough | Transaction signals, basket-size proxy | Observed transaction, not modeled estimate. Ownership contested. Highest-value data in the building after enrolled tenant loyalty. |
| Source | Signal type | Value | Notes |
|---|---|---|---|
| Lease economics | Rent, concessions, term, renewal | Feeds attribution modeling and media pricing. Requires data sharing agreement. | |
| Tenant mix + performance | Sales/sqft, occupancy, category | Required for cross-tenant behavioral reasoning. Not currently accessible to Stratus. | |
| Brookfield portfolio intelligence | Cap rates, NOI, redevelopment pipeline | Investment operations layer. The consumer layer extends the same institutional reasoning methodology to the shopper. |
Four dimensions: consent (basis, timestamp, notice version), provenance (source system, transformation log), identity (resolution method, confidence score), freshness (last validated, decay rate). The value of Layers 1A, 1B, 1C, and 3 is a function of Layer 2 quality. A dataset without consent and provenance metadata is an accumulation, not an asset. This is the layer that determines whether the data is auditable and AI-ready, whether attribution disputes can be resolved, and whether a regulator's question can be answered.
Every external loyalty, transaction, and reservation program sits here, regardless of how premium the consumer is. Sephora, Nike, AAdvantage, Bonvoy, Mastercard, Resy, Ticketmaster are one class: identity-resolved data owned by the partner, reachable only in a clean room under the partner's consent. They differ only by signal type and whether the partner is also a competitor.
| Source | Signal type | Value | Competitor? | Notes |
|---|---|---|---|---|
| Tenant loyalty | SKU / member history | Some | Sephora, Nike, Nordstrom, Starbucks. Deep-dive in the Clean Rooms section. Property-specific coverage. Nike/Nordstrom also run RMNs. | |
| Travel loyalty | Origin, tier, stay frequency | No | AAdvantage, SkyMiles; Bonvoy, Hilton Honors. Highest-income out-of-market visitor signal. Lower-conflict partners. Cardlytics precedent. | |
| Payment networks | Observed transaction, wallet share | Yes | Visa / Mastercard. 160B+ txns/yr. Both data partner and competitor (Commerce Media), scope carefully. | |
| Dining / reservations | Reservation, occasion, spend tier | No | Resy carries an AmEx identity anchor (three-way clean room). OpenTable occasion tags self-reported; Tock prepaid = confirmed intent. Entry via GGP restaurant tenants. | |
| Events | Event identity, occasion intent | No | Ticketmaster / Live Nation. Identity-rich under a different consent architecture. Careful scoping to avoid consent-boundary violations. |
| Source | Signal type | Value | Notes |
|---|---|---|---|
| Placer.ai | Modeled foot traffic | Estimate, not observed count. Benchmarking and trade-area sizing only, never a primary attribution metric. | |
| Experian | Demographics | Trade-area characterization. Not identity-resolved. | |
| Spatial.ai | Social sentiment, lifestyle clustering | 72 consumer segments from geotagged social. Trade-area characterization. Not identity-resolved. | |
| Convention / CVB data | Forward demand, event calendar | Forward-looking demand signal Placer cannot generate. Out-of-market high-income surge events. CVB feeds / Exhibitor database. |
Reconciliation note: the layer is set by who owns the consent relationship. Airline, hotel, and tenant loyalty are all partner-owned (Layer 3A), not Stratus-owned (1A), their consumer value does not change who owns the record. Co-enrollment is the only pathway that moves a partner's member into owned 1A. Identity resolution, observed-vs-modeled, and competitor status are tags within a layer, not layers themselves.
Each Stratus data layer has a home in GGP's own data infrastructure, which follows the reasoning-era methodology proven in institutional practice. The machinery follows that proven pattern; the consumer consent dimension is the net-new piece the institutional methodology never had to carry.
| Stratus layer | What it is | Maps to (Brookfield's Data Infrastructure) | Built or to-build |
|---|---|---|---|
Layer 1A Enrolled identity |
Stratus-owned consumer entity + consent record | Knowledge-graph entities, resolved via entity resolution (layers 03–04) | To build The consumer-enrollment cold-start |
Layer 1C Portfolio institutional |
Tenant mix, lease economics, property performance | The context graph, who/how/when, extended from funds & properties to consumers (layer 06) | Proven pattern Resolved by the methodology |
Layer 2 Metadata |
Consent · provenance · identity · freshness | Append-only ingestion + semantic governance + decision lineage (layers 02, 05, 07) | Split Machinery proven; consent net-new |
Layers 3A / 3B Partner + modeled |
External sources via clean room / license | Additional sources resolved into the same canonical identity layer (layers 02, 03) | Per partner Clean-room agreements |
Agentic commerce Terminal state |
Agents acting on behavioral data | The reasoning + action surface, "agents are the design target" (layers 08–09) | Proven pattern Consumer domain is new |
Two of the five commitments behind that proven methodology (also guiding Brookfield's Data Infrastructure) are, word for word, this brief's warnings: identity is foundational, and if it is retrofitted every layer above absorbs the cost, and meaning is versioned. The methodology already enforces them; GGP adopts the discipline rather than relearning it.
This is the deep-dive on the highest-value Layer 3A partners. The most data-sophisticated retailers inside GGP properties, Sephora, Nike, Nordstrom, Starbucks, have first-party loyalty programs with behavioral depth that will dwarf Stratus's enrolled base for years. Travel, payment, dining, and event partners sit in the same 3A class; tenant loyalty is simply where the value concentrates. Mostly this is a clean-room opportunity, though Nike and Nordstrom run their own media networks, so a few of these partners are also competitors. The clean room is what lets a partner-and-competitor relationship still create value for both sides.
The two parties hold two different assets. The tenant owns the data: the SKU-level transactions, the purchase history, the depth inside its own four walls. Stratus owns the metadata: the cross-tenant visit context around that purchase, and the consent, provenance, and identity layer that makes any record trustworthy and usable. The data has one value, and it stays with the tenant. The metadata has another, and it is Stratus's. In a governed clean room the two combine and each is worth more, without either party giving up its asset. That is why the largest loyalty programs are Stratus's first partnerships, not its competitors.
| Tenant program | Data depth | GGP presence | Clean-room value | Constraint |
|---|---|---|---|---|
| Sephora Beauty Insider LVMH · 34M+ members |
SKU-level purchase history, beauty profile, skin-tone data, repurchase frequency, category affinity. | Select properties | Cross-visit profile: does the Insider also shop apparel, dine, attend events? Stratus sees the full visit; Sephora sees the transaction. | LVMH protective of member data. Clean-room architecture and a formal agreement required. |
| Nike Member 160M+ members globally |
Purchase history across digital and physical, activity data, size/fit profile, early access, event attendance. | Select properties | Cross-tenant visit data Stratus holds is invisible to Nike and valuable to any brand selling to the same consumer. | Operates Nike Digital Commerce, partner and competitor for advertiser budgets. |
| Nordstrom Nordy Club ~7M active members |
Multi-category purchase history, return behavior, alteration history, stylist interaction; verified high-value spend. | Anchor presence | Anchor partnership sets a high-value baseline for the whole property, the attribution story Stratus needs to sell to non-anchor retailers. | Operates Nordstrom Media Network. Formal data sharing and revenue model required before any clean-room discussion. |
| Starbucks Rewards 33M+ active US members |
Visit frequency, order history, time-of-day behavior, mobile-order adoption. A dwell-behavior proxy. | Wide presence | High frequency creates a repeating signal across the portfolio. A morning visit then retail shopping is a pattern Stratus can attribute and Starbucks cannot see beyond its transaction. | Protective; operates sophisticated data infrastructure. Needs clear value beyond the Starbucks visit. |
| The coverage constraint Structural reality for all partnerships |
Not all retailers are in all GGP properties. A Nordstrom clean-room partnership covers the subset of assets where Nordstrom operates, not the full portfolio. The identity layer must know which consumers visited which properties with which tenant mix. Audience segments built on tenant loyalty data are property-specific, not portfolio-wide. Selling a "GGP portfolio audience" that is actually a subset is a credibility risk if not disclosed in campaign terms. | |||
How Stratus collaborates with tenant loyalty programs, travel partners, and payment networks without either party exposing raw consumer data
Neither party exposes individual consumer records. Computation happens in a neutral environment neither party controls.
Every data use requires a documented consent record in Layer 2. Without the consent architecture, the clean room has no legal foundation.
What emerges is audience segments, attribution reports, and match rates. No individual consumer can be reverse-engineered.
"The clean room is not a concession to tenant data sovereignty. It is the architecture that lets two different assets meet. The tenant brings the data, its transaction depth. Stratus brings the metadata, the cross-tenant context and the governance. Neither can generate the combined insight alone."
Every revenue stream in the original framework begins with a physical visit to a GGP asset. That is the right starting point, not the terminal state. The data business becomes significantly more valuable when it can reach consumers before they arrive, identify them across digital and physical behavior, and reason across categories a single mall visit cannot capture.
Enrolling consumers before the visit, not only during it
Today, every retail media and loyalty platform is visit-anchored: a consumer enrolls when they arrive, or after they transact. The behavioral record begins at the point of enrollment, not at the beginning of the consumer's relationship with the trade area.
Stratus can build an enrollment layer that reaches consumers before they visit, a digital commerce destination (the ShopSimon analog for GGP), trade-area marketing, travel/hospitality co-enrollment, or a tenant co-enrollment flow. When they arrive, they are already known. The visit is attributed. The behavioral record is richer from day one.
The consumer who shops Nike.com and then visits Nike at a GGP property is a single behavioral record
Every major tenant operates digital channels that generate behavioral signals the tenant captures and Stratus cannot see. A consumer who browses Nordstrom.com for three weeks before visiting made a consideration journey no property-level data source captures.
A consent-based tenant SDK / pixel lets Stratus know a visiting consumer has prior digital engagement without accessing the raw record. The insight is not "this person browsed Nordstrom.com last Tuesday." It is "this visit is likely a conversion visit." The attribution model treats that visit differently.
Reasoning across consumer behavior, tenant performance, and property economics simultaneously
The reasoning-era methodology proven in institutional capital operations (knowledge graph, context graph, decision graph) treats a portfolio as a connected system, not a collection of independent assets: tenant performance, lease economics, market positioning, and capital allocation reasoned together.
GGP applies that methodology on its own infrastructure and adds the consumer behavioral layer, answering questions no retail media network or investment platform can answer separately: which tenant categories drive cross-shopping at which property types; which consumer segments are underserved by a specific asset's tenant mix; where to prioritize enrollment investment to maximize data-asset value, not just immediate media revenue.
AI agents making or assisting purchase decisions require exactly this class of data to do it reliably
Retail 4.0 in the Stratus framework means AI-enabled consumer engagement and agentic commerce. The framework names these as future revenue streams without describing the architecture they require. That architecture is the same one built for the earlier phases: identity-resolved, transaction-linked, cross-category behavioral data with clean provenance and a governed semantic layer.
An agent making a purchase recommendation needs to know what a consumer has bought, at what price points, in which categories, with what frequency, and under what contextual conditions. Generic behavioral data does not support reliable agentic reasoning; synthetic data cannot replicate it. The data asset Stratus builds in the earlier horizons is the foundation that makes this phase possible.
The companies that will win the agentic commerce layer in physical retail are not the ones with the most sophisticated AI. They are the ones with the most trustworthy behavioral data at the moment the agent needs to act. Trust here is not a qualitative judgment, it is a metadata property: consent basis documented, provenance intact, identity confidence scored, freshness validated. The metadata architecture is the agentic-commerce architecture.
The CDP, the identity-resolution vendor, the clean-room platform, the measurement stack: all downstream. They are constrained by upstream decisions that must be made before any vendor is engaged. And one decision comes before even those: whether Stratus builds and owns GGP's data infrastructure on a proven blueprint, or invents an unproven design from scratch.
Does Stratus build and own GGP's data infrastructure on the reasoning-era methodology Brookfield already operates in institutional capital practice (the proven pattern, not a specification to commission), interoperating with Brookfield where it adds value, or does it invent an unproven design from scratch? This is not the fourth decision on the list. It is the one that de-risks the other three: adopt the proven pattern, and identity, semantic governance, decision lineage, and agent-readiness follow a design that already works; the CDP question shrinks; most of the metadata architecture is set by the same commitments. The three decisions below are then about the one layer that pattern has never carried, the consumer.
The people who understand this pattern are already inside the institution. It belongs in the architecture specification from day one, not surfaced in a discovery engagement six months later, and it is the difference between inventing an unproven stack and building on a pattern GGP already owns.
Does the consumer's relationship live at the Stratus level or the property level? Platform-level enrollment means Stratus owns the consent, the identity record, and downstream rights, including whether airline/hotel co-enrollment becomes a Stratus-owned 1A asset. Property-level enrollment means inheriting a data asset subject to JV governance it did not design. This cannot be reversed after the CDP is selected.
The Stratus economic structure defines the revenue waterfall precisely: preferred return, 25/75 incremental split, property vs. Stratus economics. There is no equivalent framework for data. Who owns consumer data collected at the property level? What can Stratus use it for, under what consent basis, across which JV structures? The governance waterfall has to be designed with the same rigor as the economic one, anticipating clean-room partnerships and institutional data access.
The metadata architecture (consent, provenance, identity, freshness) has to be specified before the CDP is selected. The CDP will lock in the metadata architecture by default otherwise. This is not a technical detail. It is the decision that determines whether the data asset is trustworthy and auditable, or an accumulation that looks like an asset until someone asks a hard question.
Identity-resolved audiences for premium advertisers. Attribution that closes the loop between campaign exposure and in-store transaction. Measurement credibility Simon and Westfield are still building toward.
Clean room partnerships with anchor tenant loyalty programs. Pre-arrival enrollment through digital commerce, travel/hospitality, and tenant co-enrollment. Cross-channel attribution from digital impression to physical transaction.
Reasoning across consumer behavior, tenant performance, and property economics simultaneously, on GGP's own data infrastructure. A capability no competitor can assemble.
The data foundation for AI agents making or assisting purchase decisions in physical retail. The architecture that supports it is the same one that supports the media and partnership businesses.