Why the standard keyword list stopped working
Search for "real estate keywords" and you will find the same artifact over and over: 150 phrases in ten numbered categories, no volumes, no difficulty, no indication which are winnable. They are all descended from each other.
That artifact made sense when a query returned ten blue links and your job was to have a page matching the phrase. It does not describe what happens now.
What happens now is fan-out. A modern search system takes one query, generates a set of related sub-queries from it, runs them in parallel, and fuses the separate result lists into a single answer. Google's AI Mode does it. Assistants do it before they answer. We covered the fusion mechanics — and why consensus across many lists beats dominance of one — in the piece on how AI mentions actually get made.
The consequence for keyword targeting is specific and under-appreciated: you are no longer competing for a phrase. You are competing for a cluster. A page that matches your seed term exactly but answers none of the sub-questions it implies loses to a page that covers the cluster and never uses your phrase verbatim.
Which is why a longer list is not the upgrade. The upgrade is knowing what each entry explodes into.
You can watch the fan-out yourself, which beats taking our word for it
Nobody outside these companies has the actual sub-query lists, and any article claiming otherwise is guessing. But the shape of a fan-out is directly observable, and it takes about two minutes:
- Open an AI assistant or Google's AI Mode.
- Ask it a seed query exactly as a client would — "should I sell my house in [your area] this year".
- Read what the answer addresses, not just what it says. It will cover things you didn't ask: current inventory, what comparable homes sold for, seasonality, whether to renovate first, what it costs to sell.
- Look at the sources it cites for each of those.
Every one of those addressed-but-unasked points is a sub-question the system generated. That list is your content brief. The sources it cited for each are your competition — and they are frequently different sources for different sub-questions, which is exactly the opening a local agent has.
Do this for the five queries that matter most in your market and you will have a better keyword strategy than any list can give you, because it will be about your market rather than the national average.
Two things this page deliberately does not include
No search volumes. Not because we couldn't put numbers in, but because for the keywords that matter to an agent they would be fiction. "Homes for sale in [city]" is tens of thousands of searches a month in Phoenix and barely double digits in a town of four thousand. A national average for a hyperlocal pattern describes no real market. Every list you have seen omits volumes for the same reason; most just don't say so. How to get real numbers for your market is further down, and it is not hard.
No consumer head terms. You will not find homes for sale in [city], houses for sale near me, or [city] real estate below. Zillow, Realtor.com, and Homes.com own those, your site will not outrank them, and the traffic is largely people browsing rather than hiring. Pursuing them is the single most common way agents waste a year. What you can win sits beside those terms, not on top of them.
How to read the map
Each cluster below has four parts:
- Seed — the pattern, with
[brackets]where your market goes. - Variants — the phrasings real people use for the same intent.
- The fan-out — the sub-questions that seed implies, which the answer will address whether or not you did.
- What the page needs — what has to exist for you to be part of that answer.
Treat it as a coverage audit, not a writing assignment. See the warning near the end about what happens if you turn each line into its own page.
Cluster 1 — Agent selection (the one that matters most)
Seed: realtor in [neighborhood] · best listing agent in [city]
Variants: real estate agent near me · who is the best realtor in [area] · top listing agents [city] · [niche] real estate agent [city] · realtor recommendations [area]
The fan-out: who is active in this specific area · how many homes have they sold recently · what do their past clients say · do they handle this price range · do they specialize in my situation · are they licensed and with which brokerage · how do I contact them
What the page needs: almost none of this lives on your website, and that is the point. This cluster is answered from corroborated facts spread across independent sources — your portal profiles for recent sales and price bands, Google and Yelp for reviews, LinkedIn for verified professional identity, your site for specialty and contact. All of it saying exactly the same thing.
This is the highest-value cluster in real estate search and the one keyword lists handle worst, because it cannot be won with a page. It is won by being resolvable as a person.
Cluster 2 — Seller intent
Seed: should I sell my house in [city] · how much is my home worth in [neighborhood]
Variants: is it a good time to sell in [area] · [neighborhood] home values · selling a house in [city] · how long does it take to sell in [area] · cost to sell a home in [state]
The fan-out: what are comparable homes selling for right now · how long are they sitting · is inventory rising or falling · what does it cost to sell here · should I renovate first · what happens to my taxes · do I sell before I buy
What the page needs: current, dated local numbers — recent sold comps, days on market, inventory direction — refreshed often enough to still be true. This is where a genuine monthly market read beats any national content, because nobody national can produce it and the sub-questions demand recency.
Cluster 3 — Buyer intent
Seed: buying a house in [city] · is [neighborhood] a good place to live
Variants: what is it like living in [area] · [neighborhood] vs [neighborhood] · best neighborhoods in [city] for [families/commuters/retirees] · moving to [city]
The fan-out: what do homes actually cost here · what are the schools like · how long is the commute · how safe is it · what is there to do · how are the taxes and HOA fees · is it appreciating · what is the catch nobody mentions
What the page needs: this is the neighborhood page cluster, and the fan-out explains why thin ones fail. A page with a stock photo and three sentences answers one sub-question badly. A page covering price reality, commute, schools, character, and the honest drawback is eligible for every branch of the fan-out at once. That is the entire argument for depth over volume — the build template is a separate post.
Note the last sub-question. Answering "what's the downside of living here" honestly is the fastest way to be the source that gets cited, because nearly nobody does it.
Cluster 4 — First-time buyers
Seed: first time home buyer [state] · how much house can I afford
Variants: first time buyer programs [state] · down payment assistance [city] · how much do I need to put down · credit score to buy a house · steps to buying your first home
The fan-out: what programs exist in my state and do I qualify · how much down payment do I really need · what is my monthly payment at today's rates · what are closing costs · what credit score do I need · how long does the whole process take · what do I do first
What the page needs: state-specific program detail with a last-verified date, since these programs change and stale information is worse than none. This cluster rewards precision and punishes generic advice, and it is heavily served by lenders — which means your angle is the local, current, agent's-eye version rather than another payment explainer.
Cluster 5 — Life-event niches
Seed: selling a house during divorce in [state] · selling an inherited house [state] · downsizing in [city]
Variants: probate home sale [state] · relocating to [city] · selling a house in a trust · moving for a job to [area] · senior downsizing [city]
The fan-out: what is the legal process in my state · who has to sign · how long does it take · what are the tax consequences · do I need court approval · how do we split proceeds · who do I need on my team besides an agent
What the page needs: genuine process knowledge for your state, and disciplined hedging — these clusters sit next to legal and tax advice, and the correct posture is explaining the process while routing to an attorney or CPA for the advice. Low competition, high intent, and a strong fit with picking one thing to be known for.
Cluster 6 — Process and cost questions
Seed: how much does it cost to sell a house in [state] · what does a realtor do
Variants: realtor fees [state] · who pays closing costs in [state] · do I need a realtor to sell · what is a buyer agency agreement · how do agent commissions work now
The fan-out: what is the actual percentage · who pays what after the settlement changes · what is negotiable · what do I get for it · what happens if I sell without an agent · what are the other closing costs · when do I pay
What the page needs: current, accurate, state-specific answers — this area changed with the commission settlement and a great deal of published content is now wrong. Being correct here is a real advantage, and the willingness to say plainly what is negotiable is what makes a source worth citing rather than worth skipping.
Cluster 7 — Investment and rental
Seed: investment property in [city] · is [city] a good place to invest in real estate
Variants: rental property [city] · cap rate [market] · short term rental rules [city] · best neighborhoods to invest [city] · [city] rental market
The fan-out: what do properties cost versus what do they rent for · what are the local short-term rental rules · what are property taxes · is the population growing · what is vacancy like · what are financing options for investors
What the page needs: real local numbers and, especially, current regulatory detail — short-term rental rules change constantly by municipality and outdated information here is actively harmful. Smaller audience than the buyer and seller clusters, more transaction value per person, and much thinner competition.
The hyperlocal modifier is a different animal
Most of the seeds above carry a [city] or [neighborhood] bracket, and that bracket behaves differently from the rest of the phrase.
It is the part with no national volume, the part you cannot look up in a generic list, and the part that decides whether the term is winnable. It is also the only part a large national competitor cannot fake, because answering it well requires having been there.
Two practical rules. First, be specific past the point of comfort. "Phoenix" is a portal term. A named subdivision, a school-attendance boundary, a 55+ community — those are terms with real intent and almost no competition, and they are the ones where a fan-out will reach for a local source because no national source has anything to say.
Second, pick few and go deep. Five neighborhoods covered properly beat thirty covered thinly, which is the same argument as specializing rather than covering everything and fails for the same reason when ignored: thin coverage produces evidence too weak to cite.
How to get real numbers for your own market
Since no list can give you volumes worth having, here is the method. It costs nothing.
- Google Search Console, if your site has any history. The Performance report shows the queries you already appear for, in your market, with real impressions. This is better than any third-party estimate because it is your actual data. Sort by impressions, filter to queries with zero clicks, and you have a list of things you are nearly winning.
- Google's own autocomplete and "People also ask." Type your seed with the local modifier and read what it suggests. Those suggestions are derived from real query data, and the related-question boxes are a rough public view of the fan-out.
- Ask an assistant the seed query and read what it addresses, per the exercise near the top of this page.
- Your own inbox and call log. The questions clients actually ask you, in their words, are query data that no tool has and no competitor can copy.
If you do want tool-verified figures, they exist and they are real — the point is that they must be pulled for your market. For context on what verified data looks like in this space: our own keyword research on the agent-marketing side, pulled 2026-07-30, found terms like closing gifts for realtors at 720/mo with a difficulty score of 0 and real estate commission calculator at 6,600/mo at difficulty 0. Those are genuine numbers with a date attached — and they are for reaching agents, not buyers, which is exactly the distinction most keyword lists blur.
The mistake this page could cause, so please don't make it
Do not turn each line above into its own page.
A site with fifty near-identical pages differing only by the bracketed city is a doorway-page pattern, it is a documented way to earn a manual penalty, and it also fails on its own terms: fifty thin pages produce fifty weak signals rather than one strong one.
Use it as a coverage audit instead. Take the two or three clusters that match the business you actually want, open the page you already have for each, and ask which fan-out sub-questions it currently answers. Add the missing ones to the existing page. A page that answers seven sub-questions is a fundamentally different asset from seven pages that answer one each — and it is the same asset a human reader prefers.
If you want that sequenced with everything else, the free DIY playbook puts it at step six, after the profile and review work that the agent-selection cluster depends on. And the local SEO guide covers the ranking stack underneath all of this.