How CRE Data and AI Can Improve Investment Decisions
Commercial real estate has always rewarded information asymmetry. The investor with better market visibility, cleaner property-level data, and faster analysis usually makes better decisions. What’s changing now is how that edge is built.
The old model depended on fragmented reports, broker anecdotes, spreadsheets, and manual underwriting. The new model combines structured CRE data, AI-assisted analysis, and human judgment. That shift doesn’t eliminate experience or relationships. It makes them more powerful.
In a recent discussion featuring James Nelson and Lonnie Hendry, Chief Product Officer at Trepp, one theme stood out: real estate professionals no longer have an excuse to rely only on instinct when the data is available to test assumptions in near real time.
For investors, operators, and technical teams building data-driven real estate workflows, this creates an important opportunity. The advantage is no longer just access to data. It’s the ability to connect reliable datasets, ask better questions, and turn answers into action.
sbb-itb-8058745
Key Takeaways
- Use data to challenge market narratives. Headlines about oversupply, insurance, or declining rents often miss local nuance.
- Property-level operating data is a major underwriting advantage. Rent, vacancy, expense line items, loan terms, and appraised values can reveal risk that broad market comps may miss.
- AI is most valuable when paired with trusted data sources. Large language models are useful, but output quality depends on the underlying data.
- Natural-language querying changes who can analyze CRE data. Users can increasingly ask plain-English questions instead of manually navigating dashboards or exporting spreadsheets.
- Job growth and population growth remain core market signals. They are not perfect, but they are foundational for long-term demand.
- Short-term supply shocks should be separated from long-term fundamentals. Oversupply can pressure rents temporarily without invalidating a market’s long-run attractiveness.
- Regulatory context matters, but rarely in isolation. Pro-development markets can face oversupply; highly regulated markets can still attract capital if demand remains durable.
- Human verification still matters. AI should accelerate analysis, not replace diligence.
- Career growth in CRE now requires both relationship skills and data fluency. The strongest professionals can interpret markets and communicate insight clearly.
The Real Shift: From Anecdotes to Evidence
CRE remains a relationship-driven business. Deals still happen between people who trust each other. But when it comes to underwriting, portfolio strategy, or market selection, relying too heavily on anecdotal evidence is expensive.
That matters because some of the most repeated views in the market are broad generalizations:
- "Insurance costs make this state uninvestable."
- "This market is overbuilt."
- "That city is too regulated."
- "Rents are going nowhere."
Those statements might contain some truth. But they’re often incomplete. What data does is force precision.
If insurance costs are rising, by how much? Over what time frame? For which asset type? In which submarket? Are those increases offset by rent growth, population inflows, or stronger occupancy? Without that context, investors are reacting to noise.
A useful line from the conversation captured this well: opinions are easy to debate; data is harder to ignore. For CRE teams trying to scale acquisitions or improve forecasting, that mindset is critical.
What Kind of CRE Data Actually Improves Decisions?
One of the most practical parts of the discussion was the distinction between general market commentary and underlying asset-level information.
According to Hendry, public securitized lending markets generate deep reporting on the collateral behind those loans. That can include:
- Ownership information
- Mortgage balance and payment status
- Fixed vs. floating rate structure
- Amortization profile
- In-place rent
- Vacancy
- Other income sources such as parking or laundry
- Expense line items
- Reserve assumptions
- Appraised value at origination
For investors, this is more than reference material. It can directly improve underwriting.
Why this matters in practice
Most acquisition teams don’t lose deals because they can’t build a model. They lose because the model is based on weak assumptions.
Access to line-item operating data helps answer questions like:
- Are payroll costs out of line with similar assets?
- Is the property under-reserved for future capital needs?
- Are management fees unusually low, masking future operating pressure?
- Does reported NOI reflect sustainable performance or temporary conditions?
- Is the implied cap rate from appraisal data disconnected from current market reality?
These are not theoretical questions. They affect pricing, debt sizing, and hold-period strategy.
For the strategic operator, better underlying data also means fewer wasted motions. If you can benchmark expenses or debt structure before pursuing an opportunity, you reduce time spent on assets that won’t survive diligence.
The Real Promise of AI in CRE Isn’t "Magic." It’s Interface
AI in real estate is often marketed as if it can replace expertise. That’s not what this discussion suggested, and that’s a good thing.
The more compelling point was simpler: AI changes the interface between the user and the dataset.
In practical terms, that means an analyst may no longer need to:
- Log into multiple systems
- Export datasets manually
- Normalize fields by hand
- Build every table from scratch
- Hunt through dashboards just to answer a narrow question
Instead, the workflow increasingly looks like this:
- Connect a trusted data source
- Ask a plain-English question
- Review structured results
- Refine the prompt
- Validate the output
- Use the answer in underwriting or reporting
The discussion referenced the use of an MCP server, described as an AI-native connection layer similar in function to an API. The significance here is less about one vendor implementation and more about the broader direction of the market:
CRE data is becoming easier to query conversationally.
That’s a major shift for technical and non-technical users alike.
Why this matters for BatchData-style audiences
For a technical architect, this points to a future where real estate data pipelines become easier to operationalize through machine-readable interfaces.
For a strategic operator, it means fewer bottlenecks between a question and an answer.
For a deal-maker, it means faster validation of opportunity before the market moves.
The key qualifier: AI only works well if the underlying data is clean and reliable. If the source is weak, the output may sound polished while still being wrong.
Better Questions Lead to Better Investment Decisions
The best use case described in the conversation was simple and powerful: asking a specific, location-based trend question in natural language.
For example:
- What happened to multifamily insurance costs in Miami over the last five years?
- How does Orlando compare?
- Can you show the output as a table?
- Can you layer in broader economic context?
That workflow matters because it converts a vague concern into an analyzable problem.
Instead of saying, "Florida seems risky", the investor can investigate:
- Asset-level operating pressure
- Trend duration
- Relative market comparison
- Implications for future cash flow
This is where AI becomes useful. Not because it "knows" real estate, but because it reduces friction between your question and the evidence.
Market Selection: The Two Signals That Still Matter Most
When the conversation turned to demand drivers, Hendry emphasized two fundamentals:
- Job growth
- Population growth
That’s not a flashy answer, but it’s a disciplined one.
Commercial real estate ultimately depends on people and economic activity. Jobs support income. Income supports rent. Population growth supports household formation, retail demand, service demand, and, depending on asset type, logistics and office absorption.
Why these two metrics still hold up
Even in an AI-heavy world, many market screens still collapse to a few durable questions:
- Are more people moving in than moving out?
- Are quality jobs being created?
- Are those jobs diversified?
- Is the migration temporary or durable?
This doesn’t mean every growth market is automatically attractive. It means these variables usually deserve disproportionate attention in the screening process.
A useful nuance: short term vs. long term
One of the better insights from the discussion was that supply imbalances are often shorter-lived than market narratives suggest.
A market can experience:
- Rising vacancy
- Concessions
- Flat or negative rent growth
…and still remain attractive over a longer horizon if its demand drivers are strong.
That’s especially relevant in Sun Belt multifamily markets. High supply can pressure performance in the near term, but the very fact that developers built there often reflects confidence in long-run demand.
For operators and investors, this creates a practical distinction:
- Short-term underwriting risk is not the same as
- Long-term market invalidation
Those are often conflated.
Regulation vs. Growth: Why Neither Side Tells the Whole Story
The discussion also explored a common investor tension: should capital favor high-growth, low-barrier markets or heavily regulated gateway markets with tougher supply constraints?
There is no universal answer, but the trade-off is important.
In pro-development markets
States with fewer barriers to construction can attract capital and new supply quickly. That supports long-run growth, but it can also create painful oversupply cycles.
Austin was cited as an example of a market where multifamily rents had fallen substantially from peak levels. That’s a warning sign for investors entering too late in the cycle or using aggressive assumptions.
In highly regulated markets
Cities with rent rules, political friction, or development hurdles may look less attractive on paper. But limited new supply and dense economic activity can keep those markets relevant for institutional capital.
The lesson is straightforward:
- Low regulation does not guarantee better returns
- High regulation does not eliminate investment opportunity
What matters is how those realities flow through to rent growth, expense growth, supply, liquidity, and debt performance.
For disciplined investors, regulation should be treated as a model input, not a headline conclusion.
Why Social Media and Alternative Signals Matter, With Caution
A less obvious but important theme was the role of social platforms in market intelligence.
That may sound unserious at first, but the point was not that posts replace diligence. It was that market participants often publish useful transaction clues in real time:
- Deal announcements
- Financing details
- Leasing updates
- Pricing chatter
- Market sentiment
For local operators and acquisitions teams, this can provide early directional signals before they show up in formal reports.
The caveat is equally important: these signals must be verified.
This is where a modern workflow can outperform legacy research habits. Instead of blindly trusting a social post or ignoring it altogether, teams can:
- Capture the signal
- Compare it to known property data
- Test it against internal comps
- Use AI to organize and summarize
- Apply human review before acting
This hybrid process is more realistic than either extreme. It neither worships AI nor dismisses it.
The Most Overlooked Skill in Modern CRE: Translating Insight
The interview closed with career advice, but it has direct relevance for operating teams too.
Knowing something is not the same as communicating it. That distinction matters in every CRE function:
- Analysts must explain assumptions
- Operators must justify budget decisions
- Developers must defend market bets
- Tech teams must explain what a data product enables
- Acquisition professionals must persuade investment committees
In other words, data literacy alone is not enough. The highest-value professionals can:
- Find the signal
- Interpret the signal
- Explain the implication clearly
That’s especially important now that AI makes surface-level analysis more accessible. If everyone can generate a chart or a summary, the differentiator becomes judgment and communication.
What This Means for Real Estate Teams Right Now
The practical takeaway from this discussion is not that every firm needs a massive in-house AI lab. It’s that most firms should already be improving three things:
1. Tighten your data foundation
Bad source data creates false confidence. Whether you’re evaluating properties, owners, market risk, or outreach strategy, the value of AI depends on the quality of the records underneath it.
For data-heavy teams, this means prioritizing:
- Clean structured fields
- Reliable refresh cadence
- Address and ownership normalization
- Transparent sourcing
- Compliance-aware workflows where contact data is involved
2. Reduce the distance between question and answer
If your team still needs multiple systems and manual exports to answer basic market questions, your analysis cycle is too slow.
That doesn’t just hurt productivity. It hurts timing. In competitive markets, latency is strategy.
3. Build human-in-the-loop workflows
Automation should accelerate diligence, not bypass it. The best model is usually:
- machine-assisted retrieval,
- machine-assisted summarization,
- human validation,
- business action.
That approach protects accuracy while improving speed.
Final Thought
Commercial real estate is entering a more data-native era, but not a less human one.
The investors and operators who win won’t be the ones who simply "use AI." They’ll be the ones who pair trusted data, faster analysis, and disciplined judgment. They’ll know how to separate headline noise from market truth. They’ll ask better questions. And they’ll make decisions with more precision than competitors still running on instinct alone.
In a market where bad assumptions are expensive, that may be the most durable edge available.
Source: "Data Insights with Lonnie Hendry" – James Nelson NYC, YouTube, Jun 25, 2026 – https://www.youtube.com/watch?v=7y969sOH1GI
