Vertical Playbooks · established evidence

Real Estate Search After the Commission Settlement: How Buyers Now Find Agents

Last reviewed 2026-07-20. Written by Chandranshu Kumar, Founder, Raveneye Global. · 10 min read

For a generation, the buyer-agent commission sat in a field on the multiple listing service, quietly shaping which agents got shown and how they competed. The National Association of Realtors' $418 million antitrust settlement removed that field: sellers' agents can no longer advertise buyer-agent compensation on the MLS, and buyers now sign a written compensation agreement before they tour a home. When a number that once organized the market disappears, buyers do not stop choosing; they choose on what they can still see. This piece traces where the agent-selection signal actually moved, using the settlement record and the National Association of Realtors' own buyer surveys. The short answer, and the reason how buyers find a real estate agent is now a visibility question, is that selection has shifted to what is legible in public: reviews, reputation, the local map pack, and increasingly the answer an engine returns when a relocating buyer asks who the best agent in a neighborhood is.

The signal that disappeared from the MLS

On November 26, 2024, a federal court granted final approval to the National Association of Realtors' $418 million antitrust settlement, and the practice changes it required took effect across the industry. Two of those changes matter for how a buyer now selects an agent. First, sellers' agents can no longer post an offer of buyer-agent compensation on the multiple listing service, so the number that used to sit invisibly behind every listing is gone from the shared database. Second, a buyer's agent must now sign a written compensation agreement with the buyer before touring a home, which moves the commission conversation out of an opaque field and into an explicit, negotiated, buyer-facing disclosure.

The economic effect showed up quickly. Redfin data cited across settlement-tracking summaries recorded buyer-agent commissions easing from 2.61 percent to 2.55 percent within a single quarter of the rule taking effect. The precise figure matters less than the direction: a compensation signal that was once standardized and hidden became variable and visible, negotiated deal by deal.

This is a structural change to how the market coordinates, not a marketing story. For decades the commission field did quiet work: it was one of the levers that shaped which agents a cooperating brokerage would surface and how agents competed for buyer representation. Remove that lever and the question of who a buyer selects, and on what basis, does not go away. It relocates to whatever signals remain in view.

Search was always the first step, and it never replaced the agent

It is tempting to assume that if buyers start online, the internet has already disintermediated the agent. The National Association of Realtors' own longitudinal data says otherwise, and the two facts sit side by side.

Across the 2020 to 2025 trend, roughly 41 to 47 percent of buyers say that looking online for properties was their first step in the home-buying process. Yet in the same surveys, about 88 percent of buyers still purchase their home through an agent or broker. The online search does not replace the agent relationship; it precedes and directs it. The buyer forms an impression, builds a shortlist, and decides whom to trust before the first conversation, and they do that reading from public surfaces.

That is the precise seam this article is about. The search is where selection now begins, but the agent is still the intermediary the buyer ultimately hires. What the commission settlement changed is not whether the agent matters; it is which visible signal the buyer leans on to choose one agent over another once the standardized compensation cue is no longer in the picture.

What replaces a hidden number: visible reputation

When a coordinating signal is removed, buyers substitute the strongest signal still available to them. In a high-consideration, infrequent purchase like a home, that substitute is reputation, expressed as reviews and social proof, because it is the one quality cue a buyer can actually read before any commitment.

The economics of that substitution are well documented. Michael Luca's study of Yelp found that a one-star increase in rating produced a 5 to 9 percent increase in restaurant revenue, and, importantly, that the effect was concentrated in independent businesses rather than chains, where buyers already hold quality priors from the brand. An agent is the definition of an independent business: a personal brand with no chain to borrow trust from, exactly the case where a legible reputation signal carries the most weight.

For agents specifically, review depth compounds this. A buyer relocating to a new city cannot verify an agent's competence directly, so they read the aggregate: how many reviews, how recent, and what themes recur. The practical problem is that an agent's reputation is usually real but scattered across Google, Zillow, and Realtor.com, none of which reconcile with the others, which leaves each profile thin where a deep one would decide the contact.

The map pack decides the local shortlist

Reputation does not act in a vacuum; it is read inside the local results a buyer sees first. Aggregated studies of Google local search behavior report that searchers click the local three-pack far more often than they click standard organic or paid results, with one commonly cited figure putting three-pack clicks around 44 percent versus roughly 29 percent for organic. The magnitude here is secondary-sourced and should be treated as directional rather than exact, but the direction is consistent across studies: for a query like the best agent in a neighborhood, the map pack is the shortlist, and the agents inside it are the ones whose reviews get read at all.

The engine is becoming the introduction

A newer surface now sits above the map pack for a growing share of buyers: the generative answer. Instead of scrolling listings, a relocating buyer increasingly asks ChatGPT, Perplexity, Gemini, or reads Google's AI Overview, who the best agent in an area is, and reaches out to the names the engine returns.

The adoption curve is steep, though the precise slope deserves caution. BrightLocal's consumer survey reports that 45 percent of consumers had used an AI tool to find a local-business recommendation in the trailing year as of its 2026 edition, up from 6 percent a year earlier. That single-source jump is large enough to warrant independent verification, so we treat it as emerging rather than settled. The underlying adoption of chatbots, however, is on firmer ground: Pew Research Center's 2026 survey finds 49 percent of US adults now use chatbots, up from 23 percent in 2023, with 42 percent of those users turning to them specifically for information search.

For an agent, this creates a second selection gate stacked on the first. Being present in the map pack no longer guarantees being present in the AI answer, because the two surfaces draw on different signals and frequently disagree about who gets recommended. An agent absent from the generated answer is not ranking eleventh; they were never named in the introduction the buyer acted on.

The entity problem unique to agents

Real estate has a discovery constraint most local businesses do not: the agent is a personal brand attached to a brokerage the agent does not own, and agents change brokerages. The same person can read one way on the brokerage bio, another on Zillow, another on a personal IDX site, with a different headshot, license display, and history on each, and the day the brokerage affiliation changes, part of that record breaks.

This matters because a search or answer engine recommends confidently only when it can resolve a stable identity. When the signals about a single agent are fragmented across profiles that do not reference one another, the engine cannot tell whether it is looking at one agent or three, and it hedges toward a competitor whose identity is easier to read. This is analysis rather than a measured finding, but it follows directly from how entity resolution works: consistency across the profiles an engine reads is what lets it treat scattered mentions as one trustworthy person.

There is also a content constraint peculiar to the vertical. The neighborhood and market content that earns an agent citations must describe schools, commute, walkability, and amenities without steering, without demographic proxies, and within Fair Housing rules. The visibility craft that works here is the craft that respects that boundary, which is a reason generic, undifferentiated content is both a compliance risk and a weak signal at the same time.

Reading the shift

Two readings of this shift overreach. The first treats the commission settlement as having disintermediated the agent; the National Association of Realtors' own data shows 88 percent of buyers still transact through one. The second treats the AI-answer surface as already dominant; adoption is rising fast, but trust in what chatbots return is not keeping pace. Pew finds that only about 29 percent of US chatbot users trust the information they get a lot or some, and general trust in online reviews has drifted down from its mid-2010s peak toward roughly half of consumers today.

What the evidence supports is narrower and more useful than either extreme. The standardized, hidden compensation signal is gone, and the signals buyers substitute in its place are the visible ones: a deep and recent review record, a consistent identity across the profiles engines read, presence in the local map pack, and a chance at being named in the generated answer. Those are surfaces an agent can actually work on, and, unlike a commission rate, they compound rather than reset with each transaction.

The operational question that follows is simply where a given agent stands across those surfaces today, and that is a measurement question before it is a marketing one.

The evidence

Key findings, with their sources

The National Association of Realtors' $418M antitrust settlement received final court approval on Nov. 26, 2024; sellers' agents can no longer post buyer-agent compensation on the MLS, and buyer's agents must sign a written compensation agreement before touring a home.

established National Association of Realtors, "NAR Settlement FAQs", 2024 (final approval November 26, 2024).
Buyer-agent commissions eased from 2.61% to 2.55% within one quarter of the settlement rule taking effect.

established Redfin data cited via multiple settlement-tracking legal summaries, 2024.
Roughly 41 to 47 percent of buyers (2020 to 2025 trend) say looking online was their first step, yet about 88 percent still purchase through an agent or broker: search precedes and directs the agent relationship, it does not replace it.

established National Association of Realtors, 2025 Profile of Home Buyers and Sellers, Nov. 2025.
A one-star increase in Yelp rating produced a 5 to 9 percent increase in revenue, concentrated in independent businesses rather than chains, which already carry brand-based quality priors.

established Luca, M., "Reviews, Reputation, and Revenue: The Case of Yelp.com", Harvard Business School Working Paper 12-016, 2011 (rev. 2016).
Consumers click the Google local three-pack far more often than organic or paid results, one cited figure putting three-pack clicks near 44% versus about 29% organic.

emerging Aggregated Google local search behavior studies as reported by industry local-SEO research, 2025 (magnitude secondary-sourced).
45% of consumers had used an AI tool to find a local-business recommendation in the trailing year as of 2026, up from 6% a year earlier.

emerging BrightLocal, Local Consumer Review Survey, 2024 and 2026 editions.
49% of US adults now use chatbots (up from 23% in 2023), and 42% of those users use them for information search, but only about 29% trust the information a lot or some.

established Pew Research Center, "Americans and AI 2026: Chatbots, Smart Devices and Views on Impact", June 17, 2026.

Calibration

What is proven, what is promising, what is unproven

Evidence tier	Tactics	What the evidence says
established	The settlement mechanics and the search-precedes-agent pattern: build the visible reputation and identity that buyers now select on, because the commission cue is verifiably gone and the agent relationship verifiably remains.	NAR Settlement FAQs (2024); NAR 2025 Profile of Home Buyers and Sellers; Luca Yelp revenue study (2011).
emerging	Optimize for the AI-answer introduction and the map-pack shortlist as a fast-rising but not yet dominant surface; measure presence directly rather than assuming it.	BrightLocal Local Consumer Review Survey (2026) on the 6% to 45% AI-tool jump; aggregated local three-pack click studies (magnitude secondary-sourced).
contested	Do not overweight either the death-of-the-agent narrative or the AI-answer-dominates narrative; trust in chatbot output and in online reviews is itself in flux.	Pew Research 2026 chatbot-trust figures; multi-year decline in general review trust from its mid-2010s peak.

Reference

Glossary

NAR commission settlement: The National Association of Realtors' $418 million antitrust settlement, approved November 26, 2024, which ended MLS display of buyer-agent compensation and required written buyer representation agreements before touring.
Multiple listing service (MLS): The shared, broker-maintained database of properties for sale. It formerly carried an offer of buyer-agent compensation on each listing; that field can no longer be posted.
Buyer-agent commission: The compensation paid to the agent representing the buyer. Once standardized and advertised on the MLS, it is now negotiated and disclosed directly between buyer and agent.
Local three-pack: The block of three local business results shown with a map for a local-intent search. It is the buyer's shortlist and captures a disproportionate share of clicks.
Entity resolution: How a search or answer engine decides that scattered mentions of a name refer to one real person or business. Inconsistent profiles across sites make an agent harder to recommend confidently.

Straight answers

Frequently asked questions

How do buyers find a real estate agent now that commission is off the MLS?

They select on what stays visible. Because the standardized compensation cue is gone, buyers lean on public signals they can read before any commitment: the depth and recency of an agent's reviews, a consistent identity across Google, Zillow, and Realtor.com, presence in the local map pack, and increasingly the names an AI answer returns when they ask who the best agent in an area is.

Did the NAR settlement remove the need for a buyer's agent?

No. The settlement changed how buyer-agent compensation is discovered and negotiated, not whether buyers use an agent. In the National Association of Realtors' 2025 data, about 88 percent of buyers still purchase through an agent or broker, even though 41 to 47 percent begin by looking online. Search precedes and directs the agent relationship rather than replacing it.

Why do reviews matter more for agents after the settlement?

When a coordinating signal like the MLS commission field disappears, buyers substitute the strongest remaining quality cue, which for an infrequent, high-consideration purchase is reputation. The Yelp revenue research shows this reputation effect is concentrated in independent businesses with no brand to borrow trust from, which is exactly what an agent is: a personal brand read through its reviews.

Are buyers really using AI to choose an agent yet?

It is rising fast but not yet dominant, and the picture is still mixed. BrightLocal reports AI-tool use for local recommendations jumped to 45 percent in 2026 from 6 percent a year earlier, a large single-source swing worth verifying, while Pew finds only about 29 percent of chatbot users trust the output a lot or some. The prudent read is to measure whether an engine names you, not to assume it does or does not.

What is the single most useful thing an agent can do about this?

Find out where you actually stand across the surfaces buyers now select on, before scoping any work. That means a measured read of your reviews and sentiment, your identity consistency, your map-pack presence, and how often the AI answers name you against real buyer questions for your market. Measurement comes before marketing.

Provenance

Sources

National Association of Realtors, "NAR Settlement FAQs", 2024 (final approval November 26, 2024) (established)
National Association of Realtors, 2025 Profile of Home Buyers and Sellers, November 2025 (established)
Redfin buyer-agent commission data cited via settlement-tracking legal summaries, 2024 (established)
Luca, M., "Reviews, Reputation, and Revenue: The Case of Yelp.com", Harvard Business School Working Paper 12-016, 2011 (rev. 2016) (established)hbs.edu
BrightLocal, Local Consumer Review Survey, 2024 and 2026 editions (emerging on the 6% to 45% AI-tool jump)brightlocal.com
Pew Research Center, "Americans and AI 2026: Chatbots, Smart Devices and Views on Impact", June 17, 2026 (established)pewresearch.org
Aggregated Google local search behavior studies as reported by industry local-SEO research, 2025 (established direction, emerging on precise magnitude)

Every figure above is attributed to a real, dated source and tagged with its evidence tier. Where a claim could not be verified to a primary source, it is not stated as fact.