You Cannot Build an Advantage on a Cracked Foundation

September 30, 2026

The conversation in multifamily this year is about artificial intelligence. Operators want to know what AI will do for revenue management, what it will do for budgeting and what it will do for resident experience. These are the right questions. But they are being asked in the wrong order.

Before any of those questions can be answered, an earlier one has to be settled. What is the AI being built on?

The answer for most multifamily operators without a dedicated data foundation is the same answer it was five years ago. A patchwork of property management systems, spreadsheets, exported reports and homegrown databases that talk to each other inconsistently or not at all. The dashboards on top of that stack can be beautiful. The decisions made from them are still being made on quicksand.

The algorithm is rarely the problem. The data feeding it is.

The data is telling on us

The research on AI project outcomes over the last two years has converged on a single conclusion. According to Gartner, 85 percent of AI projects fail to deliver on their intended outcomes because of poor data quality or a lack of relevant data (Gartner, 2025). RAND Corporation research found that over 80 percent of AI projects fail to reach production deployment, twice the failure rate of non-AI technology projects (RAND Corporation, 2024).

The pattern is so consistent that Gartner now predicts at least 30 percent of generative AI projects will be abandoned after proof of concept by the end of 2025, citing poor data quality, inadequate risk controls, escalating costs or unclear business value (Gartner, 2024).

Informatica’s 2025 CDO Insights survey put a finer point on it. Asked to name the single biggest obstacle to AI success, 43 percent of organizations named data quality and readiness (Informatica, 2025). It was the most cited barrier on the list.

These are not multifamily-specific numbers. They describe the average across industries. The multifamily picture is worse, not better. The industry runs on systems that were designed to manage individual properties rather than to feed enterprise analytics. Definitions vary between portfolios. The same metric calculated two different ways shows up in two different reports and one of those numbers is wrong.

What this looks like operationally

If a revenue manager cannot trust the unit count in last night’s report, they cannot trust the occupancy number derived from it. If they cannot trust occupancy, they cannot trust the pricing recommendation built on it. If they cannot trust the pricing recommendation, they override it. If they override it, the model never learns. If the model never learns, the AI investment that was supposed to make pricing smarter just made it more expensive.

This is not a hypothetical chain. It is the operational reality for those whose data is inconsistent and unstructured. The cost shows up everywhere. Gartner research estimates that poor data quality costs organizations an average of 12.9 million dollars per year (Gartner, 2020). For a midsize multifamily operator, that is the loaded cost of a mid-sized capital project. It is sitting on the table every year, unaccounted for, because no one is invoicing for bad data.

Falling behind is not what it used to look like. It used to mean being slow. Now it means making confident decisions on the wrong information.

The work that must be done for success

Cleaning, structuring and governing data is the least glamorous work in multifamily technology. It is invisible when done correctly and catastrophic when not. This invisible work is easy to lose sight of until it’s causing havoc.

Which is why most operators underinvest in it and overinvest in the things that sit on top of it. The AI initiative gets the budget. The BI dashboard gets the headcount. The data foundation underneath both of them gets a slide titled “data integration” with a vague timeline.

The result is predictable. The AI initiative stalls. The BI dashboard contradicts itself. The team responsible for both spends, by some industry estimates, 80 percent of its time on data preparation rather than analysis (Pragmatic Institute, 2024).

What we are calling the question

There are two ways to look at the situation. The first is that multifamily operators are behind on AI and need to catch up. The second is that multifamily operators are about to make expensive mistakes if they catch up before they fix what is underneath.

REBA exists because the second framing is the right one, and it is the framing we built the company around. Since 2019, we have been building that foundation for multifamily operators managing 2.1M+ units across the country. Clean, structured, governed and enriched multifamily data is not a nice-to-have. It is the foundation on which every meaningful AI or BI capability must be built. Operators who skip the foundation will spend the next two years discovering what the broader market is already discovering. The model was never the problem.

Over the next six weeks, we are going to lay out in detail what a real multifamily data foundation looks like and how it changes when operators put one in place. We will publish the math. We will publish the comparisons. We will publish what our customers are doing with it. This is the first of four posts.

Want the diagnostic that goes with this post? Take the AI Readiness Checklist.

It takes four minutes and tells you exactly where your data foundation breaks down before it breaks downstream.

Sources cited

• Gartner. AI projects fail due to poor data quality or lack of relevant data (85 percent figure). 2025.
• Gartner. At least 30 percent of generative AI projects abandoned after proof of concept by end of 2025. 2024.
• Gartner. Poor data quality costs organizations 12.9 million dollars annually on average. Magic Quadrant for Data Quality Solutions, 2020.
• RAND Corporation. Over 80 percent of AI projects fail to reach production deployment. 2024.
• Informatica. CDO Insights 2025. 43 percent of organizations cite data quality and readiness as the top obstacle to AI success. 2025.
• Pragmatic Institute. Overcoming the 80/20 Rule in Data Science. 2024.