MLS data, broker incentives and the durability of the cooperative 

Thinking big about residential real estate success requires a big-picture perspective. Industry Decoded features industry experts who can enrich your understanding of issues affecting the industry as a whole.

The views expressed in this column are solely those of the author.

Real estate's MLS industry faces two critical, connected decisions: how to establish the value of its data with current and new customers, and how to preserve the contributions that make that data worth buying. Both begin with the same question: What does each stakeholder put into the cooperative, and what does it receive in return?

Better data agreements can generate revenue that rewards contribution. Returning some of that value to brokers and agents through rebates, credits or lower costs could encourage timely, complete listing submission, strengthening the cooperation that sustains the data compilation.

Artificial intelligence creates an opportunity to expand that revenue through AI model training and ongoing access to current listings. MLSs can proactively establish those relationships while helping the world's largest learning engines work with accurate real estate information.

The value of listing data depends heavily on its comprehensiveness. A fragmented feed has limited value to a customer seeking a view of the whole marketplace. Even the inventory of the largest brokerage in a market is far less valuable than the near-complete market coverage that most MLSs have historically possessed. MLSs deliver their highest value when they create marketplace benefits that none of the brokers could create on their own.

Maintaining comprehensive inventory is essential to the MLS' value proposition to consumers and professionals. Establishing the right rules and incentives will be critical to sustaining it.

Establish the commercial relationship

The New York Times didn't have to sue Amazon to establish its monetized AI licensing relationship. That's because The Times' existing lawsuit against OpenAI and Microsoft had already made its position clear: Valuable proprietary content isn't free training material, and it would take action to enforce its rights.

All future negotiations and litigation will begin from this position. All publishers win because one publisher put a stake in the ground. MLSs should be pursuing their own New York Times moment.

Unlock MLS and the Austin Board of Realtors provide a look into the possibilities. Unlock's new policy framework for MLS participation distinguishes brokerages according to their business activities and uses of MLS data, moving beyond a single participant definition. Data feed access and commercial licensing terms reflect those distinctions.

This approach offers a way to reconsider the economics of relationships with partners that don't operate direct brokerage services. It also points to how AI training could become an approved use under separately negotiated licensing terms. They've already trained on MLS data. The question is when the MLSs and brokers begin receiving financial benefits for it, and the AI gold rush taking place right now can't be ignored.

The initial revenue doesn't have to be tremendous. The strategic objective is to establish a paid, authorized route to MLS data while major technology companies are choosing their long-term sources of truth for real estate data. MLSs should be those sources.

Connect the revenue to the contribution

The other side of the equation is sustaining the inventory that makes the cooperative and its data valuable. The elephant in the room is whether the growing prominence of siloed pre-marketing and private marketplace practices will materially diminish the shared marketplace.

Fragmentation can weaken the value of MLS participation and the data product offered to customers. With the core mission of the MLS being improving brokerage services that serve consumers at a higher level, a certain scale of inventory fragmentation becomes an existential issue. Better licensing agreements could increase the durability of the cooperative's value, especially when their returns help sustain the inventory behind them.

Shared rules create cooperation, but benefits can also encourage listing submission. The comprehensiveness of MLS markets I've worked in as a broker in Washington and California benefited greatly from rules like mandatory submission for all participants, ensuring baseline cooperation on listing availability.

Additional financial incentives can create some of the same outcomes in markets with these kinds of rules and in those where they don't exist or aren't enforced. Multiple organizations have established listing-contribution payments and data feed or syndication revenue rebates: CRMLS, NorthstarMLS, Bright, MARIS, NTREIS and MetroTex are high-profile examples.

Tie incentives to the quality and quantity of contributions 

What if those models, combined with a participant distinction framework, also recognized how completely a brokerage contributes? If brokerages can be distinguished by their independent decisions about using the MLS's data output, could they also be distinguished by their independent decisions about data input?

An MLS might distinguish a brokerage submitting its full eligible inventory, or all of its publicly marketed listings, from one contributing only a portion while benefiting from the broader collection. More complete contributors could receive greater rebates or pay reduced fees for services.

Volume, timeliness and completeness of listing submission are different measures that an MLS would need to consider. Submitting all 20 of a broker's eligible listings is different from submitting 200 of 400, and each MLS would have to decide where to place the incentives. Clear distinctions and legal review would be critical to ensure these business decisions could be applied transparently and consistently.

Brokers and sellers could retain their independent marketing choices. The cooperative's corresponding commercial terms would recognize the amount of value that those choices contribute. It's a "Yes, and..." model. Availability for broker cooperation would remain distinct from authorization for public display or AI training.

Could those benefits also reach the agents creating the listings, through arrangements established with their brokerages? That could make the connection between contribution and financial return visible to the people creating the data and directly guiding seller's decisions. Expanding licensing to AI uses could increase the revenue available.

Progress without unanimous agreement

Today, HouseCanary's role connecting participating MLSs with Google offers a quizzical example of how MLS relationships with data consumers from outside the industry have taken shape over time. With congratulations to HouseCanary for having successfully established such an ambitious initiative with Google, it's surely a head-scratcher to an outsider as to why this is the only visible industry data conduit to one of the world's most successful companies.

The arrangement is emblematic of the industry's difficulty turning collective strength into coordinated commercial leverage. Nimble companies advance while shared industry initiatives encounter competitive, political and personal friction. An industry that can't align on a uniform coming-soon approach should be realistic about reaching consensus on AI licensing. Luckily, common infrastructure can support independent decisions without requiring everyone to adopt the same business strategy.

RESO has standards in development for communicating data governance instructions in machine-readable form. These would allow licensing constraints, such as AI training and retention permissions, to travel directly in the data feed while establishing a paper trail. MLSs need not solve every enforcement problem before preserving evidence of what was supplied and authorized. That record can support future compliance. 

Distribution infrastructure already exists through the most up-to-date RESO Web API providers like Bridge, Cotality, FBS, MLS Aligned, MLS Grid and RealtyFeed. And the transaction doesn't necessarily need to remain within the traditional MLS/IDX/API structure. The Global Data Exchange connects MLSs, brokerages and non-broker entities through independent data agreements. It's another example of a framework that could support licensing and monetization with global publishers and AI model companies.

Project NexusRE, being developed by NorthstarMLS and REcore, will provide visibility into downstream usage, permissions and monetization. Its anticipated contribution-based credits connect governance directly to the financial benefits for brokers.

Each MLS would choose its own providers and design its own classifications and licenses for its data consumers. Getting the first implementation in place will demonstrate the capability to the rest of the industry and be key to seizing this timely opportunity.

Build a reference agreement to show the pathway

Which MLS will establish a publicly documented agreement with a frontier-model AI company?

Some are likely already contemplating or negotiating in private. The first public agreement won't set pricing or licensing terms for the entire industry, but it can demonstrate a general approach others can learn from.

This isn't a thought exercise. I reached out to a senior director I know at OpenAI this weekend, and I'd invite the industry's technology leaders, who likely already have relationships with leading AI organizations, to begin discussing this topic with them. In the spirit of bolstering the comprehensive pro-consumer marketplace, it's important that all of the MLS' partners shine light on the opportunity and identify the decision-makers who we'll need to negotiate with. These conversations can begin independently as fact-finding missions, rather than waiting for industrywide consensus on the shape of the final agreements. 

Revenue returned to brokers through contribution-based benefits would encourage more complete, timely participation. Stronger participation would improve both the cooperative marketplace and the data offered to customers. The pool of data buyers and their financial contributions would grow. That is a virtuous business cycle for brokers and their customers. 

Who will connect the marketplace data's value to a more durable cooperative? The clock is ticking.

Sam DeBord is CEO of the Real Estate Standards Organization (RESO) and the author of Working with Real Estate Data, organized real estate's premier technology course for onboarding and professional development. He has served as President's Liaison for MLS and Data Management with NAR, and he has also worked as a consultant, managing broker, NAR REACH mentor, and board director for association, MLS and venture fund organizations.