CRE firms have years of data so Why can’t anyone answer a simple question?
The promise of the CRE CRM was simple. Put relationships, transactions and intelligence into one system and nobody would have to guess what a client needed. When the broker asked a question, the answer would be sitting there in the data.
The reality has turned out rather differently. Relationship history may sit in one system. Deal and commission information may be somewhere else. Comparable transactions can live inside a specialist research tool that nobody opens often enough. And the most valuable context of all may be buried inside an email thread from six months ago.
Ask a broker what they know about a particular tenant and, remarkably often, they are answering from memory rather than from the data their organisation already possesses.
That is the central issue explored in a recent episode of Ascendix Technologies’ The AI Brief: CRE Intelligence, where CEO Wes Snow and CTO Todd Terry sit down with Rob Ward to examine why this problem has persisted for twenty years — and what AI has actually changed. The answer is potentially much bigger than another new piece of proptech.
The data was never necessarily the problem. Accessing and using it was. The 20-year CRM promise.
The traditional CRM was supposed to eliminate information silos. Instead, many CRE businesses have accumulated more technology, more databases and more information silos. The fundamental difficulty is that CRM systems depend upon people continually putting information into them.
Every relationship, conversation, meeting, transaction and piece of market intelligence creates another potential data-entry requirement. Someone has to open the system, find the appropriate record, complete the fields, categorise the information and save it. It is a hidden cost of using technology — what the Ascendix discussion describes as the data-entry tax.
And it is particularly problematic in an industry where the people generating the most valuable information are often the least interested in spending their time entering it into software. The result is predictable. The company owns the data, but the data is incomplete. Information exists, but nobody can easily find it. The CRM becomes a database that contains some of the corporate memory rather than all of it.
AI changes the equation
This is where AI potentially represents a fundamental change. AI does not necessarily need people to behave differently before it can become useful. Instead of asking a broker to stop what they are doing and complete another form, AI can potentially capture information from emails, documents, meetings and conversations and turn it into usable intelligence.
That changes the relationship between the individual and the CRM. Historically, the user had to understand the technology. They needed to know which system contained the information, where to look, which fields had been completed and which report might produce the answer. The emerging model is almost the reverse.
The user asks a question. The technology works out where the relevant information resides. Imagine a broker asking: How does this lease compare with comparable properties within five miles?
That question could require several systems, an internal database, external market information, research tools and perhaps a spreadsheet. An AI-enabled environment can potentially bring those sources together and provide a single response. The significance is not simply that the answer arrives faster. It is that the user no longer needs to understand the architecture of the technology before getting to the answer.
The data doesn’t have to be perfect
Perhaps the most interesting implication is that AI changes the traditional obsession with perfectly structured data. For years, technology projects have attempted to clean, standardise and centralise information before it can become useful. That has created enormous projects — and often enormous bills. And AI introduces another possibility.
Data can remain distributed. It can exist in different formats. Some of it can even be “dirty”. An intelligent layer can potentially interpret those different sources and bring them together when a question is asked. That does not mean poor data suddenly becomes good data. It means previously inaccessible or difficult-to-use information can become considerably more valuable.
For CRE businesses that have spent decades accumulating information, this is potentially significant. The technology investment has already been made. The data already exists. What may have been missing is the intelligent layer sitting above it.
The CRM isn’t dead
There is an important qualification. AI does not make the CRM or ERP irrelevant. In fact, the opposite may be true. The system of record remains critically important because it provides the authoritative foundation for customer, transaction and business information. What AI changes is the interface between the human being and that system.
Instead of forcing people to interact with software through forms, menus and fields, AI can become another interface — one capable of understanding natural language and retrieving, synthesising and potentially updating information across established systems. This means decades of technology investment do not necessarily have to be discarded.
They may finally become more useful. The opportunity for property technology companies could therefore be less about replacing the existing stack and more about making the existing stack accessible.
But AI creates a new problem
There is, however, a catch. Once AI makes it easier to obtain an answer, another question becomes increasingly important:
Where did that answer come from?
If an AI system tells a broker something about a tenant, a lease or a comparable transaction, can the broker identify the source?
Can the information be traced?
Is it current?
Is it authoritative?
Or has the AI simply generated an apparently convincing answer from information of uncertain provenance?
This is where the discussion moves beyond the familiar AI narrative.
The problem used to be getting information out of the systems.
The emerging problem is knowing whether the information AI has brought together can be trusted.
In commercial real estate, that distinction matters enormously.
The future of AI-driven property intelligence therefore cannot simply be about accessibility. It also has to involve provenance, traceability and appropriate human oversight.
The next proptech opportunity may already be inside the business
Perhaps the biggest message from the discussion is that the next generation of property technology may not require another giant system. It may require an intelligent layer across the systems businesses already own. Commercial real estate has spent twenty years building digital infrastructure. CRM systems, property databases, transaction platforms, research tools, email archives and other applications contain an enormous amount of corporate knowledge.
The problem has been that the knowledge is fragmented. AI potentially changes that equation. It can act as the interface between the human being asking the question and the collection of systems containing the answer. That could transform the economics of decades of technology investment. The test is surprisingly simple.
Count how many systems you or your team open every day just to answer ordinary business questions.
That number is the beginning of the problem statement. The question for CRE businesses may no longer be ‘Where should we put all our data?’It may be, Why can’t the technology we already own simply answer the question? That is a very different proptech opportunity. And it may be one of the most important consequences of AI for commercial real estate: not creating more data, but finally making the enormous amount of data the industry already possesses usable.
Watch/listen the full podcast now The AI Brief: CRE Intelligence
