This week's signal is a split. On one side, who gets to see a listing — and when — is being contested everywhere at once: New York's OneKey MLS joined the federal suit defending cooperation rules, the first big independent brokerage publicly cooled on the new Open Real Estate platform, and RealScout started stitching pre-market inventory from both sides of the divide into one agent-client search. On the other side, the pitch itself is being automated: an AI avatar can now present your listing video without you ever filming, and two website vendors launched conversational sites that qualify visitors before an agent ever dials. The distribution pipes are contested; the presentation layer is automated. For an agent, that means the week's real work is not buying a new tool — it is deciding, for each listing, which pipes it travels through and whether the automated pitch still sounds like you.
01
Make it shine
Tools that make every listing look irresistible.
2 stories
NewVideo & Visual
Pedra's Presenter puts an AI avatar of you in the listing video — no filming required
On October 28, Pedra launched Presenter, a feature that puts a realistic AI avatar of the agent into real-estate videos made from listing photos, with a voiceover in the agent's own cloned voice. The workflow, per Pedra's own release page: upload the listing photos plus one photo of yourself, choose Presenter on the scenes you want, and Pedra generates horizontal and vertical versions with subtitles and music. Each scene with a presenter costs 5 credits, whatever its length; no dollar price is stated on the release page, so do not convert credits to dollars without checking Pedra's pricing. Agents who would rather not appear can pick one of 9 AI presenters instead. Pedra says the feature works best with a front-facing, well-lit photo and in rooms with open space, such as the living room or kitchen. All quality claims ('realistic') are the vendor's own, not independently tested. Two caveats for US agents: Pedra is Barcelona-based and its listing-link import covers European portals, not US MLS systems — the tool itself works from uploaded photos — and a video that shows an avatar that looks and sounds like you presenting a property you never filmed raises an obvious disclosure question that Pedra's release does not answer.
Why it matters
This removes the most common blocker to presenter-led video: going to the shoot alone, having nobody to film you, or disliking being on camera. A solo agent can now ship a presenter Reel from photos alone, which is genuinely useful for vacant listings. But the disclosure question is the story, not a footnote: if a buyer assumes the agent filmed that walkthrough, the video has misled them before the showing. If you try it, label the avatar on the video itself — the trust you save is worth more than the polish you gain.
CoStar to buy aerial-imagery firm Vexcel for ~$730M — the imagery pipeline behind tomorrow's listings
On October 27, CoStar Group (owner of Homes.com, Apartments.com, and Matterport) announced a definitive agreement to acquire Vexcel Group, operator of a survey-grade aerial imagery program, for approximately $730 million in cash, funded from cash on hand and its revolving credit facility. The deal has not closed: it is expected to close in the first quarter of 2027, subject to regulatory approvals and customary closing conditions. Vexcel captures more than 60 million images a year across 45+ countries and 320 major metros, covering 99.6% of the US population, re-flying major US metros up to three times a year and disaster zones within 24 hours, at 7.5–15 cm resolution — roughly 16 times finer than typical satellite imagery, per CoStar's own comparison. Vexcel's trailing-12-month revenue exceeded $140 million, up over 30% year over year, at an approximately 20% adjusted EBITDA margin, all company figures. After closing, Vexcel keeps its brand and its CEO Erik Jorgensen stays on. CoStar's forward-looking ambition — extracting 'hundreds of data fields per property, across the approximately half a billion properties captured by Vexcel' with AI — is the company's own projection, flagged as such in its forward-looking-statements disclaimer. No new Homes.com listing-visual feature was announced.
Why it matters
No agent can use this today, and this issue will not pretend otherwise. It matters because CoStar now owns the capture pipeline — the aircraft, the UltraCam cameras, the processing — behind the imagery that feeds its marketplaces, on top of Matterport's indoor capture. The practical read: the richest property data will increasingly come from above and from AI extraction, not from what an agent types into a listing form. That raises the value of the one thing aerial imagery cannot capture — accurate, current, on-the-ground condition — and it is a reason to keep your own listing facts sharp rather than assuming the platform's data will fill the gaps.
Tools that put every listing in front of the right eyes.
4 stories
NewMarketing Stack
OneKey MLS joins the federal suit defending listing-sharing rules — the cooperation fight goes to New York
On October 27, New York's OneKey MLS joined California Regional MLS (CRMLS) as a plaintiff in its federal lawsuit against Compass in the Southern District of New York, asking the court to declare that rules requiring publicly marketed listings to be shared through the MLS are lawful and pro-competitive. OneKey says it holds more than one million active, pending, and historical listings across the Hudson Valley, Long Island, and New York City, and trade coverage puts its subscriber count at roughly 40,200. The move builds on OneKey's October 20 rejection of Compass's demands to dismantle cooperation policies. OneKey's stated position — that withholding publicly marketed inventory to preferred brokers and buyers 'harms consumers by restricting buyer access, undermines fair housing objectives, and potentially limits a seller's return' — is a litigant's argument, not a court finding; Compass's counter-position, as characterized in coverage, is 'seller choice.' No ruling has been reported, and the suit now also invokes New York's Donnelly Act alongside the federal Sherman Act and California's Cartwright Act.
Why it matters
This is the legal backdrop to every pre-market conversation you have with a seller. If cooperation rules hold, a publicly marketed listing must still reach the full MLS audience; if they fall, brokerage-only distribution gets easier to defend. Until a court rules, the honest seller script is the conditional one: here is the exposure you get through the MLS, here is what a pre-market-only route costs you in audience, and here is what we cannot yet promise about either. Do not sell either outcome as settled.
Howard Hanna, left out of Open Real Estate, says it is unlikely to join — the first crack in the 'everyone's invited' story
On October 29, Real Estate News reported that Howard Hanna Real Estate CEO Howard 'Hoby' Hanna IV said his firm was not consulted on Open Real Estate — the shared listing-data platform announced October 26 by AGNT Inc., HomeServices of America, Keller Williams, and Real REMAX Group, and covered in Issue #011 — and suggested Howard Hanna is unlikely to join because it has already invested millions in its own similar platform, Find It First. Find It First, announced in mid-2019 and built with Ocusell, lets anyone with a Howard Hanna website account see the brokerage's listings before wider distribution, alongside its HannaList pre-marketing program. Hanna also warned that inviting the whole industry to shape the platform risks stalling it like the defunct Upstream initiative: 'so many chefs in the kitchen.' His account of not being invited and his 'unlikely to join' signal are his own statements, reported by trade media — credible as his position, not as proof of Open Real Estate's internal process. Real Estate News also notes the platform's founders have not yet assembled an executive team, chosen a tech vendor, or finalized governance. Hanna said he expects 75 MLSs to be mapped within Find It First by next summer — his own projection, unverified.
Why it matters
This is the follow-up that keeps Issue #011's launch story honest: the new network is not yet industry-wide, a major independent already runs a competing pre-market pipe, and governance — not technology — may be the launch risk. Do not pitch Open Real Estate to sellers as a done deal or as universal coverage. The question to ask your broker now is sharper than 'are we joining?': if the biggest independents stay out, does the platform still deliver the audience your seller is paying for?
RealScout adds eXp Coming Soon listings — pre-market inventory from both camps lands in one agent-client search
On October 27, Inman reported that RealScout will integrate eXp Realty's Coming Soon listings into its collaborative home-search platform later this year — no firm launch date disclosed. Buyers working with an agent on RealScout will see eXp's Coming Soon listings in the search and collaboration tools they already use, labeled within the agent-client relationship and replaced by the active MLS listing once the preview period ends. The integration follows RealScout's agreement six weeks earlier to carry Zillow Preview listings, making it one of few platforms offering pre-market inventory from both sides of the divide. RealScout says more than 100,000 real estate professionals have access to its platform — a company figure. RealScout co-founder Andrew Flachner told Inman the eXp listings are not off-MLS: 'These aren't off-MLS listings. eXp's Coming Soon still go into the MLS,' but not every MLS distributes that status externally, so the integration closes that gap. Showings and offers in the early period remain governed by local MLS rules, which vary by market.
Why it matters
This is where the pre-market fight is actually being fought this week — not on a new portal, but inside the search tool buyers already share with their agent. If your buyer clients use RealScout, eXp Coming Soon inventory may soon appear in their feed, and Zillow Preview inventory is coming to the same place. Do not promise clients access until a launch date is confirmed — but do check whether your own buyers' search tool is quietly becoming the place where pre-market inventory aggregates, because that changes which listings your clients see first.
Roomvu opens Territory: one neighborhood per agent, built to be named by AI answers
On October 27, Roomvu opened Roomvu Territory, which applies its AI-visibility method to a single neighborhood per agent: a voice interview (12 sessions of 3–6 minutes) that builds the agent's 'Voice DNA,' 20 published neighborhood answers, a dedicated neighborhood domain, a bio of 500+ characters, a connected personal domain, 5+ connected channels including Google Business Profile, and sold listings from the past two years. Monthly reporting tracks whether AI assistants name the agent for that neighborhood. Roomvu says more than 1,400 agents have completed the core steps of the method and that 400,000+ professionals are signed up across North America — both company claims — and it explicitly does not guarantee rankings or AI citations. No price was stated in the release. The release leans on BrightLocal's Local Consumer Review Survey 2026 (see By the numbers) and on published research it cites second-hand: Ahrefs' finding that AI cited content 25.7% fresher than Google organic results across 17 million citations, and a KDD 2024 paper finding statistics, quotations, and citations raised AI-answer visibility by up to 40%. Roomvu CEO Sam Mehrbod is running the method on Vancouver's Point Grey neighborhood against untouched Kitsilano as a control since September 2026, with results promised but not yet published.
Why it matters
Strip out the product and the method is the story: depth in one neighborhood beats city-wide breadth, and the checklist — consistent Google Business Profile, bio, domain, fresh question-answering local content — is replicable without buying Roomvu. The honest test is free: ask ChatGPT or Gemini who knows your farm neighborhood, then check whether it names you. If it does not, the fix starts with the same checklist, not with a purchase. And note what Roomvu itself concedes — no vendor can guarantee an AI citation, so any vendor who promises you one is selling something Roomvu will not.
Turn attention into relationships, and relationships into commissions.
3 stories
NewMarketing Stack
Agent Image's SiteAI turns your website into a conversation — and hands you the visitor's brief
Agent Image (El Segundo, CA) announced SiteAI, a patent-pending conversational platform for agent, team, and brokerage websites: visitors ask questions in their own words and get answers built from the agent's website content and local knowledge — not as a chat window, but as a magazine-style page in the agent's brand, composed on the spot for the question asked. When a visitor is ready to connect, the agent receives an 'intelligence briefing' with budget range, neighborhood preferences, intent level, questions asked, and a suggested follow-up approach. Fair Housing guardrails are claimed as built in. Real Estate News covered the launch on October 26; the company release is dated October 27 — a one-day dateline difference, not a contradiction in substance. SiteAI is available now to all real estate agents, per the release; no price was disclosed. Agent Image president Jon Krabbe's positioning — 'Real estate agents don't have a lead problem. They have a lead quality problem' — and the claim that conversations arrive 'already warm before you dial' are vendor claims with no independent conversion evidence offered. The Fair Housing guardrails are asserted, not independently tested.
Why it matters
This attacks the exact leak in most agents' funnels: anonymous website traffic that never becomes a conversation. The visitor brief is the genuinely useful workflow change — your first call starts with budget, neighborhood, and intent instead of a cold 'how can I help?' But treat the 'warmth' as a hypothesis to test on your own traffic, not a result: run it on one page, compare the quality of the first calls it produces against your form leads, and check the answers it gives about your listings before a buyer ever sees one. An unreviewed AI answer on your own site is still your answer.
Delta Media Group's Intelligent Websites: the registration form is dead, the conversation is the lead form
On October 28, Delta Media Group (Canton, OH) introduced Intelligent Websites, the first product under its new Delta Connected Intelligence platform, designed to connect AI across its websites, CRM, and other tools. Visitors search conversationally instead of via menus and registration forms, and the site requests contact information naturally mid-conversation — 'conversational lead generation' rather than form-fill capture. No price was disclosed, and no AI model is named ('frontier AI' is the company's term). The follow-on phases — Intelligent DeltaNET for agent-facing tools and an Action Center for agent priorities — are announced as coming, not yet shipped, so the CRM payoff (conversation data flowing into follow-up tools) is promised, not delivered. Delta describes itself as 'America's largest family-owned real estate technology provider for brokerages,' partner to 80+ LeadingRE affiliates and 50+ top-ranked brokerages — vendor claims. An early-customer quote from Tara Siegel (VP Marketing, Lewith & Freeman Real Estate) is a testimonial, not a performance result.
Why it matters
Pair this with SiteAI and the week's theme is unmistakable: two established website vendors independently concluded the registration form is dead, within 48 hours of each other. The practical question is which stack you already pay for — Delta's pitch only pays off for brokerages on its platform, and only when the promised CRM connection ships. If your website vendor is not Delta or Agent Image, the takeaway still applies: audit your own form. How many visitors abandon it, and what would a conversation capture that the form loses? That audit costs nothing; the vendor switch should wait for the answer.
DealNav 3.0: click any lot on the map, get its county record, and ask ChatGPT about your pipeline
On October 27, DealNav launched DealNav 3.0, a map-based CRM for real-estate acquisitions and brokerage professionals. Click any lot on the map to open its county record — owner, mailing address, acreage, zoning, flood zone, year built, building size, last sale — and one more click turns it into a tracked deal with the owner linked as a contact. The release also connects DealNav directly to ChatGPT, Claude, and Perplexity, plus a public API on every plan, so a team can query its own pipeline ('Which owners in this submarket should we be following up with?'). List, Kanban, and pipeline views, Gmail/Outlook sync, and email/text reminders round it out. Pricing is disclosed: Basic $59/month (25 parcel lookups/month), Pro $99/month (unlimited lookups), both with a 14-day free trial; Enterprise custom. DealNav says users have mapped 175,000+ properties and contacts across all 50 states and that it is rated 4.7/5 on G2 — vendor claims, and the release is wire-only with no trade-media coverage found, so treat it as the weakest-sourced story in this issue.
Why it matters
This is the most concrete workflow story of the week — it collapses the spreadsheet plus county website plus generic CRM shuffle into one map — but it is a niche one: investor-facing agents, land and acquisition teams, and farm prospecting, not the typical residential listing agent. If that is you, the $59 entry point makes it cheap to test against your current prospecting stack for one farm area. If it is not you, skip it: the week's conversational-website stories are the ones that touch a standard residential funnel.
45% of consumers now ask AI for local recommendations — up from 6% a year earlier
BrightLocal's Local Consumer Review Survey 2026 (1,002 US adults, published February 11, 2026) is not new this week — but it became this week's evidence base: Roomvu's October 27 Territory launch leans on it, and it is the dataset behind the GEO pitch agents are hearing from every vendor right now. The headline figures: 45% of US consumers have used ChatGPT or other generative AI tools for local business recommendations, up from 6% a year earlier, making AI the third most-used source after Google and Facebook; 42% trust AI recommendations as much as traditional reviews; and 97% of consumers read reviews for local businesses, with Google's share as a review source slipping from 83% to 71%. Read the limits with the numbers: this is self-reported survey behavior — people who have tried AI for a recommendation at least once, not people who hired through one — and the 6%-to-45% comparison carries the usual survey-question and sample caveats. The direction, corroborated across multiple secondary analyses of the same survey, is what matters: discovery now starts in an AI answer, and the answer is checked against reviews before anyone calls.
45%
of US consumers have used ChatGPT or other generative AI tools for local business recommendations, up from 6% a year earlier
42%
trust AI recommendations as much as traditional reviews
97%
of consumers read reviews for local businesses — the check AI answers still get put through before anyone calls
1,002 adults
the US survey sample behind these figures (BrightLocal, February 2026) — self-reported behavior, not measured hiring
“The pipes are contested; the pitch is automated. Own the decision in between.”
This week, nobody agreed on how a listing should travel. An MLS went to federal court to defend sharing rules. A major independent declined, in effect, to join the new sharing platform. A search tool stitched pre-market inventory from rival camps into one feed. And while the industry argued about the pipes, vendors automated the pitch: an avatar that presents your home, websites that hold the first conversation for you. The pattern is that the two halves of the job are moving in opposite directions — distribution is fragmenting into competing arrangements, while presentation is consolidating into AI that speaks in your voice. The part neither side automates is the judgment in between: which pipes serve this seller, and whether the automated pitch still tells the truth about this home. That judgment is the agent's remaining monopoly, and it is also the part that gets skipped when the tooling feels complete.
For your next listing, write two lines before you touch a tool: (1) which pipes this listing travels through, and in what order — MLS, pre-market, portals, AI answers — and what audience each one costs or gains the seller; (2) one sentence the automated pitch must never say about this home. Then check the AI output against line two before it ships. The platforms will keep arguing about the pipes and automating the pitch. The decision in between is yours to keep.
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: Pedra's 'realistic' avatar quality, 5-credit pricing unit, and feature behavior are the company's own, from its release page, which states no dollar price; CoStar's Vexcel figures (60M+ images/year, 99.6% US population coverage, 7.5–15 cm resolution, ~16x satellite comparison, >$140M revenue, ~20% EBITDA margin) and its 'hundreds of data fields' ambition are company statements, and the acquisition has not closed. OneKey MLS's and CRMLS's characterizations of Compass's conduct and of cooperation rules' effects are litigants' arguments, not court findings; Compass's 'seller choice' position is as characterized in coverage. Howard Hanna CEO Hoby Hanna's account of not being consulted, his 'unlikely to join' signal, and his 75-MLS projection are his own statements via Real Estate News. RealScout's 100,000+ professional-access figure and its 'alongside the MLS' framing are company claims; its eXp integration is planned, not live, with no firm launch date. Roomvu's adoption figures (1,400+ agents, 400,000+ professionals, ~99% US listing coverage via 150 MLS connections), its cited Ahrefs and KDD 2024 research, and its Sarasota agent anecdote are company statements via its release, which itself disclaims any guarantee of rankings or AI citations. Agent Image's lead-quality claims and Fair Housing guardrails, and Delta Media Group's 'frontier AI,' scale, and partner claims, are vendor statements; Delta's CRM-connected phases are announced, not shipped. DealNav's 175,000-property figure and 4.7/5 G2 rating are vendor claims from a wire-only release. BrightLocal's survey figures are self-reported consumer behavior from a February 2026 survey of 1,002 US adults, re-cited in Roomvu's October 27 release — not a new in-window study.
