This week's signal is that the listing is escaping the portal — in three directions at once. Zillow put its listings, maps, and affordability math inside Claude (its third chatbot integration after ChatGPT and Gemini), so a home search can start and nearly finish in a chat window, with the listing display carrying Zillow's own MLS-approved attribution. In the courts, six MLSs representing more than 120,000 professionals jointly sued Compass in Manhattan to defend the rule that publicly marketed listings must flow through the shared database, and New York's OneKey MLS publicly refused Compass's September demand to loosen the same rule. And the agent's side of the screen is being rebuilt around it: Real Brokerage's Leo Pro turns listing photos into video tours and chases cold leads on a pay-per-task token model — $9.99 a month intro, 20,000 tokens use-or-lose — while a new startup launched to measure whether AI assistants recommend you at all. The listing is no longer a page; it's data that travels. Every story this week is about the terms of travel: who owns the pipes, who writes the rules, and who does the work when the channel is a conversation.
01
Make it shine
Tools that make every listing look irresistible.
2 stories
NewVideo & Visual
Planpoint launches Visit — interactive floor plans for resale listings, from $15/month
On October 21, Planpoint (the Montréal software company behind interactive floor plans for new residential developments, founded by Laurent Cardinal) announced Planpoint Visit, a tool for agents and photographers marketing resale homes. It links curated listing photos and notable room features to an interactive floor plan so buyers can explore room by room. The workflow: a free iPhone scanning app places viewpoints and publishes an interactive tour — Planpoint says a floor plan can be generated 'in minutes instead of days' — plus an AI interior-design tool that lets buyers redecorate rooms, direct tour links for listings on Zillow, Trulia, and Redfin, and Pro analytics tracking time spent, room-level attention, and where buyers drop off. The company says the whole build process can be completed in under 45 minutes. Pricing starts at $15 per month for five active property listings. Note the attribution boundary: the release's traction figures (24.7 million unit views, 15.3 million floor plan downloads, 1 million favorites) describe Planpoint's existing new-development product, Planpoint Viewer — not Visit. All feature, speed, and pricing claims are the company's own, issued through a vendor press release (EIN Presswire, via Disrupt PR), not independently verified.
Why it matters
Interactive floor plans have been a new-development luxury; Visit is a credible attempt to bring the format to resale at a price point ($15/month for five listings) most agents can test. The analytics angle is the real story: room-level attention data tells you which spaces sell the house, and that's feedback your next shoot can act on. Test it on one listing and compare its room-level drop-off data against the photos-only baseline before committing.
Higgsfield publishes an AI property-video guide — and warns agents to check the windows
On October 20, Alex Mashrabov's Higgsfield published a guide in its AI newsroom showing agents how to animate listing photos into short room-by-room clips with its Cinema Studio workflow. The notable part isn't the workflow — it's the warning, issued by the vendor itself: a generated clip 'can show a window, wall or view that the home does not have.' Higgsfield tells agents to compare every clip against the original images before using it. RuntimeWire's staff-written coverage adds the honest caveat: the guide 'offers no independent evidence on how reliably Cinema Studio preserves rooms and features.' This is vendor content covered by a tech-press outlet, not a product launch or feature update — treat it as a workflow story, and there is no pricing or independent product-claim verification.
Why it matters
This pairs directly with Issue #009's disclosure stories: even the AI video vendors are now writing the compliance guidance themselves. The 'compare every clip to the original' rule is the floor plan–era version of keeping unedited originals — build it into your video workflow, and keep the D.C./California disclosure logic in mind for anything that invents a window. When the vendor warns you, believe it.
Tools that put every listing in front of the right eyes.
3 stories
NewMarketing Stack
Zillow launches a Claude connector — home search inside the chatbot, with a path to tours, agents, and financing
On October 19, Zillow announced a connector bringing its listings, home value estimates, and affordability tools into Anthropic's Claude. Setup: open Claude, go to 'Customize,' select 'Connectors,' and connect Zillow; then ask in plain language ('Find three-bedroom homes for sale in Denver under $600,000 with a garage' or 'Show me apartments for rent in Seattle under $2,500 a month'). Listings appear with photos, prices, and interactive maps, with links back to Zillow — and in the same conversation, users can request a tour, connect with a real estate agent, or explore financing, which Zillow frames as keeping 'agents at the center of consumers' journeys.' Zillow says it sends no data to Anthropic; the display inside Claude is Zillow's own MLS-approved listing display with the same broker and MLS attribution as Zillow.com. This is Zillow's third chatbot integration — existing experiences in ChatGPT and a rentals-focused one in Gemini. All details are from Zillow's own announcement (PRNewswire release, October 19, 3:13 PM ET); no pricing was mentioned and whether it requires a paid Claude plan was not stated.
Why it matters
The portal is no longer a destination — it's becoming a service inside other people's conversations. The MLS-approved attribution detail matters: your listings will carry their broker/MLS identity into chatbot answers, which is good for accuracy and attribution. The agent's job in this world shifts upstream: make sure the underlying listing data (photos, facts, showing instructions) is pristine, because a chatbot won't ask clarifying questions the way a portal searcher does — it will just read what's there.
Six MLSs jointly sue Compass in federal court — asking judges to declare cooperation rules lawful
On October 21, six MLSs — Arizona Regional MLS (ARMLS), Bridge MLS, Canopy MLS, REcolorado, Hive MLS, and Doorify MLS, representing more than 120,000 real-estate professionals — filed a joint federal lawsuit against Compass in the U.S. District Court for the Southern District of New York. They ask the court to declare that their cooperation rules (requiring publicly marketed listings to be submitted to the MLS) comply with Section 1 of the Sherman Act and the state antitrust laws of Arizona, California, Colorado, and North Carolina, and to permanently enjoin Compass from suing them over those rules. The complaint alleges Compass has been 'campaigning to restrict access to residential listings so that consumers and other brokers must go through a Compass agent just to learn about a home for sale' and that 'Compass aggressively used the court system to effectuate that campaign' — all plaintiffs' claims, not adjudicated facts. A Compass spokesperson told RISMedia: 'Only the law and the homeowner should tell agents how to publicly market a home, and these MLSs are neither the law nor the homeowner,' also claiming MLSs serving 350,000-plus agents across 12 states already allow off-MLS public marketing (a Compass claim, unverified). Covered by RISMedia on October 22; Inman, HousingWire, and Real Estate News also reported the filing.
Why it matters
This is now a six-plaintiff federal war over the plumbing of distribution — and it consolidates the opposite side of Compass's own October 16 antitrust suit against CRMLS (Issue #009). Wherever it lands, it will define whether your listing is required to live in the shared database or can live only inside a brokerage's private channel. Keep selling the database either way: until the courts rule, your MLS cooperation rules — and their fine schedules — are the law of your local market.
OneKey MLS publicly rejects Compass's demand to loosen private-listing rules — the third MLS to do so
On October 20, OneKey MLS — which operates across metropolitan New York — published a defiant response to Compass's September demand letter giving it until October 6 to stop penalizing agents who publicly market office-exclusive listings off the MLS or face potential antitrust litigation. OneKey said it will not change its rules governing how publicly marketed listings are shared, arguing that publicly marketing homes while 'hiding' them from the MLS database 'allows a single brokerage to act as a gatekeeper, controlling who can see a listing and which competing agents can bring buyers to the table' — OneKey's characterization, in its statement. CEO Richard Haggerty: 'Consumers deserve the freedom to work with the professional they trust without being limited to one company's private inventory. That cooperative model creates equal opportunity, and those fundamental protections are worth defending.' OneKey also cited Bright MLS and Zillow studies finding that homes on the MLS typically sell for more than off-market listings, and raised fair-housing concerns, referencing NAACP President Derrick Johnson's warnings that private listings are a 'modern form of digital redlining.' Compass pushed back, calling the MLS arguments a way to make 'a collection of competitors the judge, jury and executioner for how agents compete.' OneKey is the third MLS to publicly rebuke Compass, after CRMLS and Austin's Unlock MLS. Reported by Real Estate News on October 20.
Why it matters
The lawsuits above are the legal track; this is the political one — and both sides are now citing the same evidence set (cooperation rules, price studies, fair housing) in public. For non-Compass agents, the practical move is the same as Issue #009's: know exactly where your MLS stands and what its fines look like, because this fight is being litigated nationally but enforced locally, house by house.
Turn attention into relationships, and relationships into commissions.
2 stories
NewMarketing Stack
Real Brokerage unveils Leo Pro — pay-per-task AI agents priced at $9.99/month intro, plus Referral Central
At its RISE conference in Las Vegas on October 20, The Real Brokerage unveiled Leo Pro, the next generation of its in-house AI platform, living inside reZEN, its proprietary operating system. The pitch: a library of more than three dozen AI agents — agents describe the task, and Leo Pro picks the capabilities and coordinates the work, checking in for approval before what Real calls 'important actions.' Capabilities include turning property photos into video tours, creating social media content, and customized promotional materials; AI Builder lets agents build custom agents without code; Leo Pro Voice adds a spoken interface; and AiRM connects agents' CRMs to Leo Pro for automated follow-up, combining CRM data with property and ownership records to flag contacts who may be ready to move. The pricing model is the notable departure: Leo Pro runs on Leo Tokens — agents pay per completed task rather than per seat. A Real spokesperson told Inman that under the monthly plan, agents get 20,000 tokens to use or lose (additional tokens can be purchased), with an intro price of $9.99 per month for agents who sign up the day of launch. The brokerage is using the model partly to hedge against runaway AI expenses from 'power users.' Separately, Real launched Referral Central, which moves agent-to-agent referrals inside reZEN — matching clients with colleagues using production data and service areas, sending the referral agreement in a few clicks, and tracking it in one place. Both are available only to Real Brokerage's 36,000-plus agents for now, not the wider Real REMAX Group; president Jenna Rozenblat tied Referral Central to the REMAX acquisition, which created a combined group of more than 180,000 agents worldwide. CTO Pritesh Damani's claim that 'at least 80% of real estate operational work will be automated' is a vendor-side aspiration, not data. Reported by Inman and Real Estate News on October 20.
Why it matters
Token pricing is the proptech industry's answer to 'show me the ROI on AI' — you pay when the work lands, and the vendor is on the hook when it doesn't. That's a healthier contract than a per-seat license you never open. Watch two things: whether the use-or-lose token bucket changes how agents actually behave (it rewards measured use, punishes experimentation), and whether Leo Pro's approval gates keep the human genuinely in the loop on anything client-facing.
Nick Bailey joins AuthoritySignal AI — a startup that scores whether AI assistants recommend you
On October 22, HousingWire reported that longtime industry executive Nick Bailey — nearly 30 years of real-estate experience including leadership roles at REMAX, Century 21, Zillow Group, and most recently T3 Sixty, where he was chief real estate officer — has joined the Las Vegas-based AuthoritySignal AI as a founding member to lead growth and strategy. The startup's premise: analyze the 'digital signals' around an agent, team, or brokerage and measure visibility across the markets and specialties they want to be known for, so professionals can see how AI tools interpret their expertise and authority — and strengthen the signals that make them more likely to surface when consumers ask AI-powered tools for recommendations. Bailey: 'Consumers are beginning to ask AI who they should hire, who knows a particular market and who they can trust.' The article notes this kind of intelligence could influence content strategy, PR, review management, and partnership decisions. No pricing, funding, or methodology details were disclosed. Two caution flags: the article was 'generated using HousingWire Automation and reviewed by a HousingWire editor before publication,' and all effectiveness claims are company-side and unverified.
Why it matters
Read this as the flipside of the Zillow-Claude story: home searches are moving into chatbots, and 'are you recommended by the AI' is becoming a measurable business metric — the GEO (generative engine optimization) conversation, now aimed at agents. Treat the scoring product with skepticism until there's a public methodology, but don't dismiss the question it asks: open ChatGPT or Claude and ask 'who's the best agent in my market' — the answer is already someone's client flow.
NAR's 2025 Profile of Home Buyers and Sellers (transactions completed July 2024–June 2025) keeps producing usable baselines. Among buyers who used the internet, photos led the website features rated 'very useful' at 81%, followed by detailed property information at 77%, floor plans at 57%, real-estate-agent contact information at 44%, and virtual tours at 38% — with the caveat that the question doesn't separate listing photos from other online photos, so the 81% is best read as a measure of photos in general. The funnel context: 46% of all buyers said their first step was looking online for properties, and 52% found the home they purchased online. The figures were surfaced and quoted in a real-estate analysis article published October 19, 2026, which links to the NAR summary materials hosted by association sites; the underlying survey is NAR's published profile.
81%
of online buyers rated photos 'very useful' — the top website feature in NAR's 2025 profile (transactions Jul 2024–Jun 2025)
57% vs 38%
rated floor plans 'very useful' vs 38% for virtual tours — both useful, but the foundation is still the gallery
46% online
of all buyers said their first step was looking online for properties; 52% found the home they purchased online
Caveat
the survey doesn't separate listing photos from other online photos — read 81% as photos in general, and all figures as NAR survey responses, not measured outcomes
“The listing is no longer a page — it's data that travels. Own the inputs, and the pipes will argue over you.”
Look at the week's mechanics: Zillow ships your listing into Claude with MLS attribution attached. Compass and the MLSs sue each other over whether the listing must live in the shared database. Leo Pro's AI agents turn your photos into videos and work your cold leads by the token. AuthoritySignal wants to score whether a chatbot recommends you. In every story, the valuable thing is the same: the inputs — clean listing data, good photos, a compliant workflow, a CRM that knows your contacts, a reputation the machines can read. Nobody in this week's news is fighting over who takes the photo. They're fighting over the pipes the photo flows through, and the rules that govern the flow. That's a better fight to be above: an agent whose inputs are pristine — accurate data, labeled media, working relationships — wins in the portal, in the chatbot, and in the private channel, because every channel is drinking from the same well. The MLSS fight decides the plumbing; your inputs decide whether the water is worth drinking.
This week, audit one input that feeds every channel: pull your most recent listing and check it the way a chatbot would read it — facts accurate, photos labeled, floor plan attached, showing instructions current, disclosures complete. Then ask ChatGPT or Claude to recommend an agent in your market and read the answer as a buyer would. The pipes are being renegotiated above you; the inputs are the part you control.
Editorial disclosure: NanaStack is independent. Nothing in this issue is sponsored. Vendor and company claims are reported from the cited sources and are not independently tested: Planpoint's feature, speed ('under 45 minutes'), and pricing claims come from its own vendor-issued press release, for which EIN Presswire disclaims accuracy; its traction figures (24.7M unit views, 15.3M floor plan downloads, 1M favorites) describe the existing Planpoint Viewer product, not Visit. Higgsfield's property-video guide is vendor newsroom content with no independent evidence on how reliably it preserves room features. Zillow's connector details ('no data sent to Anthropic,' MLS-approved display, same attribution) are company statements from its press release. The six-MLS complaint's characterizations of Compass's 'campaign' are plaintiffs' allegations in active litigation, not adjudicated facts; Compass's claims (MLSs serving 350,000+ agents across 12 states allow off-MLS marketing; MLSs are 'neither the law nor the homeowner') are company statements, not verified facts. OneKey MLS's 'gatekeeper' characterization and its cited price studies are the MLS's own claims. Real Brokerage's Leo Pro capabilities and the 80%-automation outlook are vendor-side statements; the token pricing (20,000 use-or-lose tokens, $9.99/mo intro) was confirmed by a company spokesperson to Inman. AuthoritySignal AI's effectiveness claims are company-side, undisclosed in pricing/methodology, and reported by an automation-generated HousingWire article reviewed by a human editor. The NAR statistics are survey responses from the 2025 Profile of Home Buyers and Sellers, not measured outcomes.
