Preface · a hypothesis, not a diagnosis
Everything here is built from public signals: investor filings, press releases, LinkedIn org data, job postings, and a recruiter screening call. It represents my initial product hypothesis about where the company is and what the Data Products function is being built to do. The goal is to show how I think before we speak, and to refine this thesis against what the team has learned internally.
01 · Context
In four years, the company built a real technology foundation inside one of the most complex operating environments in commercial real estate. An LLM-powered document extraction pipeline. An Azure Data Lake with raw, curated, and consumer tiers. A GIS intelligence layer led by a former leading-CRE-data-vendor executive. An occupier analytics platform certified to ISO/IEC 27001. The a managed-data-operations acquisition for institutional data operations. A global global license to a leading CRE data and analytics suite. Internally shipped platforms including an occupier analytics platform, a query layer, and a leasing platform demonstrated to 500 brokers. They are operational. The CIO has carried both architecture vision and product translation since 2021 — an enormous amount of surface area for one leader. the Data Products lead was given the mandate in January 2026 to build the Data Products function from the ground up.
02 · Market framing
JLL relaunched Azara in November 2024 with a conversational AI layer powered by JLL Falcon, reporting 200,000+ weekly LLM prompts internally. CBRE's Ellis AI supports self-service data querying and won Forrester's 2025 Technology Strategy Impact Award. The next occupier analytics interface will be a conversation with the data, not another dashboard. The winner makes information easier to interrogate, trust, and act on. The company has the substrate to compete — and needs the product roadmap to do it before JLL and CBRE deepen client switching costs.
03 · Executive thesis
The company has already done the hard part. A pattern has emerged in the org signals: the platform build has reached the point where product leadership is what drives the next phase of value. The leverage gap is the constraint. The function exists to take the CIO's architecture and turn it into measurable products across valuation, capital markets, and occupier solutions, then scale those products globally instead of rebuilding them market by market. The data is there. The opportunity is a product leader who can take an AI extraction pipeline, a the enriched property graph, a GIS layer, and a managed data operations capability and turn them into daily workflows that move the recurring revenue the company's investor relations team calls its most durable business segment. Conviction has to replace motion.
04 · Root causes
- 01
The data platform has no workflow owner
The CIO has described the AI extraction pipeline publicly as a solid information foundation. A senior leader owns the GIS capability. A separate tool supports V&A workflows. An occupier analytics platform supports that workflow. No named PM owns the translation of those capabilities into daily workflows for underwriters, appraisers, brokers, capital markets teams, and clients. Infrastructure does not create revenue by existing. Revenue comes when users trust a workflow enough to change behavior.
- 02
The occupier analytics platform is behind the AI interface curve
JLL Azara has conversational AI. CBRE Ellis won Forrester's 2025 award. the company has an occupier analytics platform and query layer, but public signals still point to dashboarding and scenario planning rather than natural-language interaction. The category is moving. Every quarter the platform stays in dashboard mode is a quarter where competitors deepen client switching costs.
- 03
International M&A has created a replication problem
Similar products are built out of headquarters, then moved to additional international markets with local adaptation. Replication works until scale turns every local variation into product debt. A configurable platform with localization layers changes the equation: core workflows, shared infrastructure, reusable components, market-specific rules where needed.
05 · Operating problems
- 01
Recurring-revenue commitment without a product roadmap to back it
The company has publicly committed to recurring businesses generating 40%+ of total revenue. Management Services reached $1.24B in 2025, growing 12% YoY. A product roadmap is what makes that commitment credible quarter over quarter.
- 02
global CRE data and analytics licensed but underused
The global CRE data and analytics license signed in March 2026 is already paid for. a 54M+ property database with ML-based entity resolution is sitting inside the company's stack. The window to turn that intelligence into workflow value before it becomes shelfware is open now.
- 03
Debt origination is the right initial wedge but unscoped
The cached job description named debt origination automation and a 20% cost reduction target as the initial anchor use case. High-volume, document-heavy, measurable baseline, close to revenue. A working first product here establishes that the Data Products function can ship a workflow, instrument the result, and turn a manual process into operating leverage.
06 · Organizational readiness
The CIO has been carrying both architecture vision and product translation since 2021. The platform is ready for a dedicated product layer to take that translation work and run with it. The Data Products lead's mandate — kind of role that gets created when leadership is serious about making a platform bet pay off — exists to free the CIO to direct rather than carry the delivery layer himself.
07 · Product leadership mandate
- Own the workflow translation layer — assign each capability (data lake, GIS, the licensed property data, AI extraction) a specific user, decision, metric, and revenue outcome sequenced against the public 40% recurring-revenue commitment.
- Define the AI interface roadmap for the occupier analytics platform and its query layer — what questions occupiers should be able to ask in natural language, what data can be safely surfaced today, what needs governance or retrieval design.
- End international rebuilds with a configurable platform model — Spain, Germany, Asia receive configurable variants of the same product. International M&A becomes platform leverage.
- Connect Data Products to the capital markets workflow — debt origination automation as the anchor use case with a 20% cost reduction target.
- Establish instrumentation as a product discipline — search paths, abandoned workflows, field-level confidence, data quality gaps, user corrections, and decision outcomes all feed the roadmap.
08 · The first 90 days
Days 1–30 · Map
Map the architecture to the revenue line
Shadow underwriting, valuation, capital markets, and occupier teams at headquarters. Find what they trust, where they leave the system, where manual work still hides, and where the same data gets recreated in different forms. Output: a gap map presented to the Data Products lead and the CIO by Day 30 as the basis for what gets built first.
Days 31–60 · Scope
Sequence debt origination and scope the the AI layer on the occupier analytics platform
Define the debt origination automation product spec: which documents, which fields, what accuracy threshold triggers the 20% cost reduction, what needs human review, what the AI pipeline already supports. Alongside, scope the first conversational layer for the occupier analytics platform — what could ship in the next 90 days that a client or advisor would actually use. Bring both to the platform sponsor as parallel-track proposals.
Days 61–90 · Ship
Ship the first proof point and present the platform model
Deliver a working version of the AI-assisted debt origination workflow to the capital markets team. Measure time savings against baseline. Track corrections. Find the trust breaks. Alongside the prototype, deliver the global product architecture: how international variants get configured from a central platform, what stays global, what becomes local, and what governance model keeps both velocity and consistency intact.
09 · Metrics to watch
| Metric | Target |
|---|---|
Debt origination time savings Named in the JD as the anchor outcome. The first proof that Data Products can ship, measure, learn, and scale. | 20% cost reduction at the workflow level within 12 months |
Recurring-revenue mix The public IR commitment becomes credible quarter over quarter only with a data-product roadmap behind it. | Maintain progress toward 40%+ of total revenue |
International platform reuse rate Stops product debt compounding with each acquisition. | Configurable variants replace rebuilds in the next two market launches |
Occupier-platform NL-interaction adoption Closes the gap with JLL Azara and CBRE Ellis before client switching costs deepen further. | First conversational layer in production with named-user usage within 6 months |
10 · Risks & mitigations
AI theater instead of decision support
MitigationChase questions that get answered, not features that look impressive. Confidence scoring, source citation, and human escalation built in from the first wedge.
Debt origination over-scopes the first release
MitigationFirst useful wedge is bounded: specific documents, specific fields, specific accuracy threshold. Ship a working prototype before scoping the second.
International configurability slips to next quarter forever
MitigationPresent the global product architecture at Day 90 alongside the debt prototype. Without an explicit decision, replication continues by default.
11 · Why now
The AI interface gap is widening now — JLL Azara and CBRE Ellis are training the market to expect conversational access to real estate intelligence. The global CRE data and analytics license signed in March 2026 is already paid for; the window to turn it into workflow value before it becomes shelfware is open now. And The company has publicly committed to recurring businesses generating 40%+ of total revenue — a product roadmap is what makes that commitment credible quarter over quarter.