Selling models and market structure
Twenty Buyers
6 min read
Twenty multiple listing services, four percent of the national total, serve half of all subscribers in the United States and generate roughly 49% of the sector's revenue.
That is the number worth sitting with, and it is not the one that gets quoted. The quoted number is 484, the count of MLSs operating on 31 December 2025. Below 500 for the first time, down from 514 a year earlier, and down 43% across a decade. Thirty disappeared in twelve months.
Counting them describes the tail. The head consolidated long ago, and the head is who buys.
Why the buyer matters more than the count
Most software an agent opens was not bought by that agent. RPR arrives with association membership. A CMA tool is frequently covered by a board site licence. Presentation software is bought by the brokerage.
When an organisation buys on behalf of its members, that organisation writes the specification. It evaluates adoption, brand consistency, support load, and how the thing demonstrates to a room of people who did not ask for it. Reasonable criteria for a board answering to thousands of members with one budget, and none of them is whether a listing priced well. That happens nine months later, at a kitchen table nobody in the procurement meeting will ever sit at.
So the specification for a large share of American real estate software is written by roughly twenty organisations, falling about 6% a year. Not a complaint about any of them. A question about how much variety a market produces with twenty customers.
What twenty customers does to a roadmap
Take the vendor's side. Functionally there are twenty accounts that decide whether you exist, the roadmap follows renewal, and renewal is measured on adoption and support load. A feature that helps one agent defend one price does not appear in that evaluation. It is invisible at purchase and surfaces months later at a listing that did or did not sell, by which time nobody attributes the outcome to the software. What gets built is what a board can see in a demonstration.
This is not a failure of imagination among vendors. It is what happens to any category where the buyer and the user are different people and the buyers can be counted on your fingers, which is the argument at the level of who signs. The count is what that argument was missing.
What a merger does to the subscribers underneath it
Thirty MLSs went in 2025, each disappearance meaning a group of subscribers inheriting a technology package they did not choose, with different tools and different data conventions, on a timetable set by a merger rather than by their own business. Local Realtor associations fell the same year, from 1,014 to 991, and twelve associations now serve 20% of all Realtors.
Set that against the agent population. NAR membership sits at approximately 1.48 million, down 2.2% from 1.52 million, still above the 1.2 million budgeted for 2026. So the number of agents is broadly holding while the number of organisations buying on their behalf falls quickly. That gap is the story. The industry is not shrinking. The number of people writing the specification is.
The case for consolidation, which is stronger than the case against
Almost everything about this is good for agents, and a piece making only the argument above would be dishonest by omission. Fewer MLSs means fewer artificial boundaries between markets that already function as one, better and more consistent data standards, and lower cost per subscriber, because the fixed costs of legal, compliance and technology spread across more people. Smaller organisations have struggled with those costs for years, and a merger is often the responsible answer rather than a defeat.
MLS leaders themselves expect the trend to continue, and they also do not expect it to end in a single national system. Both of those things can be true.
The cost is variety. Twenty buyers produce a narrower range of products than two hundred did, and that cost does not land evenly. It lands on the agent who needs something the twenty did not think to specify.
| Measure | Value |
|---|---|
| MLSs operating, 31 December 2025 | 484, below 500 for the first time |
| Subscriber concentration | 20 MLSs, 4% of the total, serve half of all subscribers |
| Revenue concentration | Top 20 generate approximately 49% of sector revenue |
| Association concentration | 12 associations serve 20% of all Realtors |
| NAR membership | Approximately 1.48 million, down 2.2%, above the 1.2 million budgeted for 2026 |
All figures from T3 Sixty's 2026 Organized Real Estate Indices, reported 4 March 2026. T3 Sixty is the only organisation tracking these consistently since 2018, so this is one source rather than a consensus, and it should be read that way.
How many MLSs are there in the United States?
484 as of 31 December 2025, below 500 for the first time, down from 514 a year earlier and down 43% across a decade. The concentration matters more than the count: twenty serve half of all subscribers.
Why are MLSs merging?
Fixed costs, mostly. Legal, compliance and technology obligations have risen faster than smaller organisations can carry them, and combining spreads those costs while removing boundaries between markets that already function as one. Local associations declined the same year, from 1,014 to 991. MLS leaders expect the trend to continue and do not expect a single national system.
Who decides which software real estate agents use?
Frequently not the agent. Associations, MLSs and brokerages buy on behalf of members, and the buyer writes the specification. With twenty MLSs serving half of all subscribers, much of that specification comes from a small group evaluating on adoption and support load rather than transaction outcomes.
What this means if you run an office
The annual technology package from your board is not a negotiation. No line item to remove, no alternative supplier to compare against, because the board bought one of each on behalf of every member. The negotiation you imagine you are having took place elsewhere, among people serving a membership rather than a business, and they were right to weigh it the way they did.
What you can do is know which of your tools were chosen by someone whose criteria were not yours, and be deliberate about the gaps that leaves. Nobody runs that exercise, because the software arrives without an invoice attached and free things do not get audited.
One disclosure, since this is a piece about who sells to whom. CMAflow sells both ways, direct to agents and through boards, on fees the board sets. So everything above is a structure we are inside rather than observing, and a vendor selling through twenty buyers carries the same incentives as any other. The direct channel does not make us different.
This article is general information and analysis, not financial, lending, or appraisal advice. Verify any home value with a licensed professional before acting.
The Independent Agent
Substack | Spotify | CMAflow FAQ | YouTube | Free CMA | Home valuation | Insights | Blog