Real Estate Tech and AI

How Buyers Use AI in the Home Search

What buyers ask AI before they call an agent: payments, values, neighborhoods and process. What 2026 surveys show and how to run a better first call.

AssistCall research teamPublished Reviewed by Faruque Amin, VP of Sales and Marketing9 min read

On this page
  1. Homework done before the phone rings
  2. What buyers ask the chatbot
  3. Why the surveys disagree on how many
  4. Four things buyers bring to the first call
  5. A first-call script for researched buyers
  6. When the call comes after hours
  7. What buyers still want from a person
  8. Be open about your own tools
  9. Frequently asked questions
  10. References
Real Estate Tech and AI: How Buyers Use AI in the Home Search45%: of prospective buyers have used AI tools in their home search, up from 37% a year earlierREAL ESTATE TECH AND AIHow Buyers Use AI in theHome Search45%of prospective buyers have used AI tools in their homesearch, up from 37% a year earlierAva for Real Estate TeamsInsightsReal Estate Tech and AI: How Buyers Use AI in the Home Search45%: of prospective buyers have used AI tools in their home search, up from 37% a year earlierREAL ESTATE TECH AND AIHow Buyers Use AI inthe Home Search45%of prospective buyers have used AItools in their home search, up from37% a year earlierAva for Real Estate Teams
Figure source: Veterans United Home Loans, 2026[1]

Key takeaways

  • Surveys disagree on how many buyers use AI, from 20% to 45% depending on who was asked and what counted[1][2], but they agree on what it is used for: payments, values, neighborhoods and the process.
  • Most of that research happens before the first call: 74% of buyers installed a real estate app, most of them before contacting an agent or a lender[3].
  • Treat the buyer's numbers as a draft to check, not a mistake to correct. Ask what a payment or value figure includes before you answer it.
  • Buyers still want people for tours and contract guidance[2]. The first call is where you show the difference.

Homework done before the phone rings

A buyer who calls about a listing at nine on a weeknight has usually done an evening of work first. They have a saved search, a rough monthly payment, a view on whether the asking price is fair and a short list of questions about the area. More of that homework now runs through AI chat tools as well as the portals. In Zillow's 2026 buyer survey, 82% of buyers said they use AI in general, and 61% of those had used it while buying a home, most often to learn how the process works[3].

The research also starts before any agent hears about it. Seventy-four percent of buyers installed a real estate app during the purchase, most of them before contacting an agent or a lender[3]. Contacting an agent was still the most common first step, at 46%, but that share fell from 52%, while the share who went to a lender first rose from 18% to 29%[3]. Buyers reach the agent later in the process and with more numbers already in hand.

Where buyers are when they first call
  • 74%

    installed a real estate app, most before contacting an agent or lender[3]

  • 29%

    contacted a lender before anyone else, up from 18% a year earlier[3]

  • 45%

    of prospective buyers have used AI tools in their home search[1]

Zillow Consumer Housing Trends Report 2026 and Veterans United Q2 2026 survey.

That changes the job of the first call. The buyer is rarely asking you to explain what a pre-approval is. They want you to check what they already think they know, and they will judge you on how you do it.

What buyers ask the chatbot

Several 2026 surveys asked buyers what they use AI for, and the answers line up even where the headline numbers do not. In a Veterans United survey of prospective buyers, the most common uses were finding homes (52%), estimating monthly payments (43%) and checking property values (39%)[1]. In Bank of America's 2026 report, as summarized by NAR, buyers who used AI most often used it to estimate affordability, payments or closing costs (57%), to learn about the buying process (55%) and to research neighborhoods, market trends or property values (52%)[2].

What prospective buyers use AI for in the home searchShare of prospective buyers in a Q2 2026 online survey of 859 people run for a mortgage lender.
What prospective buyers use AI for in the home searchBar chart. Finding homes: 52%; Estimating monthly payments: 43%; Checking property values: 39%; Learning about housing trends: 35%; Mortgage calculator: 33%; Virtual home tours: 32%; Visualizing renovations: 29%; Comparing lender reviews: 28%.Finding homesFinding homes: 52%52%Estimating monthly paymentsEstimating monthly payments: 43%43%Checking property valuesChecking property values: 39%39%Learning about housing trendsLearning about housing trends: 35%35%Mortgage calculatorMortgage calculator: 33%33%Virtual home toursVirtual home tours: 32%32%Visualizing renovationsVisualizing renovations: 29%29%Comparing lender reviewsComparing lender reviews: 28%28%What prospective buyers use AI for in the home searchBar chart. Finding homes: 52%; Estimating monthly payments: 43%; Checking property values: 39%; Learning about housing trends: 35%; Mortgage calculator: 33%; Virtual home tours: 32%; Visualizing renovations: 29%; Comparing lender reviews: 28%.Finding homesFinding homes: 52%52%Estimating monthly paymentsEstimating monthly payments: 43%43%Checking property valuesChecking property values: 39%39%Learning about housing trendsLearning about housing trends: 35%35%Mortgage calculatorMortgage calculator: 33%33%Virtual home toursVirtual home tours: 32%32%Visualizing renovationsVisualizing renovations: 29%29%Comparing lender reviewsComparing lender reviews: 28%28%

Source: Veterans United Home Loans, 2026[1]

Show the data
UseShare of buyers
Finding homes52%
Estimating monthly payments43%
Checking property values39%
Learning about housing trends35%
Mortgage calculator33%
Virtual home tours32%
Visualizing renovations29%
Comparing lender reviews28%

Read across the lists and four questions keep coming back: what can I afford, what is this house worth, what is this area like, and how does buying actually work. Each one turns into a predictable moment on your first call, and each one can go well or badly depending on how you handle it.

Why the surveys disagree on how many

The headline shares range widely because each survey asked a different group a different question. Quote the one that matches the point you are making, and say who was asked.

Four surveys, four definitions. Veterans United is a mortgage lender and ran its survey through a research firm.
SurveyWho was askedWhat the number measuresFigure
Zillow Consumer Housing Trends, 2026Recent buyersUsed AI while buying, among buyers who use AI at all61%[3]
Veterans United, Q2 2026859 prospective buyers, online surveyUsed AI tools in their home search45%[1]
Cotality, 2026Prospective and recent buyersUse generative AI tools at least monthly, for anything55%[2]
Bank of America, 2026Prospective buyers and current homeownersUsed AI tools or chatbots for homebuying research20%[2]

For a brokerage, the practical reading is simple. Somewhere between one in five and one in two of the buyers who call have run part of their search through AI, and younger buyers more often: Bank of America's figure rises to 28% for millennials and 32% for Gen Z[2]. Assume the next caller has done it, and check their numbers politely.

Four things buyers bring to the first call

Each kind of AI question leaves a trace you can hear on the call. The table lists what tends to come up, what to check and what to do next.

Example first-call checklist for buyers who have researched with AI tools.
What the buyer bringsWhat to checkYour move
A monthly payment estimateWhether it includes property tax, homeowners insurance, mortgage insurance and HOA dues for this homeWalk through the inputs, then introduce a lender for real figures
A value opinion on the houseWhat it compared: condition, recent sales, homes now under contractAsk where the figure came from; bring your comparables to the showing
Questions about schools, safety or who lives in an areaFair housing limits on what an agent can say about people and placesDescribe the property, point to public sources for the rest
A view of the processYour state's rules and your brokerage's buyer agreementExplain the written agreement and the next three steps, in order

Payments come up most, so start there. A chatbot's estimate is only as good as what the buyer typed in. The CFPB notes that a total monthly mortgage payment can include homeowners insurance, taxes and possibly mortgage insurance through escrow, while condo, co-op or HOA fees are usually paid separately[4]. A buyer who priced only principal and interest will be surprised by the first Loan Estimate. Ask which parts their figure includes, then hand off; the pre-approval handoff to a lender covers how to do it without losing the buyer.

Example. A buyer says an AI tool put the payment on a $420,000 townhouse at $2,350 a month. You ask whether that includes taxes, insurance and the $310 HOA fee. It does not. You say: “That looks like principal and interest only. With taxes, insurance and the HOA the real figure will be higher, and a lender can give you the exact number. Shall I introduce you to one today?”

Value questions need the same care. If a buyer says the house is overpriced, ask where that view came from and what it compared. Then bring your own comparables to the showing. You do not need to win the argument on the phone; you need the buyer to come and look, because condition and layout never show up in an estimate.

Neighborhood questions carry the most risk. Buyers ask AI tools about schools, crime and who lives where, then ask the agent the same thing in the same words. Answer with facts about the property, point to the school district's and police department's own published information, and never describe an area by the people in it. Avoiding steering when buyers ask about schools and safety has wording that works on the phone.

Process questions are where you earn the next meeting. Buyers who learned the steps from a chatbot often know the words, such as contingency, earnest money and buyer agreement, without knowing how they apply in your state or at your brokerage. Explain the written agreement plainly before the first showing; explaining a buyer agreement to a first-time buyer gives a structure for that conversation.

A first-call script for researched buyers

  1. Ask what they have already looked at

    “Before I repeat anything you know, what have you found so far, on the apps or anywhere else?” It tells you where they are and shows you respect the work they did.

  2. Confirm the numbers they quote

    Ask what a payment or value figure includes. Do not call it wrong; call it a starting point that needs local inputs, then supply them.

  3. Answer what is confirmed, flag what is not

    Give the facts on the listing sheet. For anything else, say you will check with the listing agent and say when you will call back.

  4. Book the next real step

    A showing, a lender introduction or a buyer consultation, with a date and time. The first-call qualifying questions fit here.

  5. Send a short written recap

    Two or three lines by email: what you confirmed, what you are still checking and the next appointment. Researched buyers keep notes, so give them accurate ones.

Speed still matters more than polish. Among buyers who contacted an agent, 54% contacted only one before hiring them[3]. A buyer who has done an evening of research and then reaches voicemail can open the app again and call the next name on the screen.

When the call comes after hours

Research happens in the evening, so a share of these calls land when nobody is at a desk. If the listing line forwards to an SI Employee outside office hours, the call can still produce something useful by morning instead of a voicemail that says “call me back”.

Example: an evening listing call from a buyer who has done the research
  • Caller
  • SI Employee
  • System
  • Your team
  1. 01Caller

    Buyer calls the listing line

    8:50 pm, after comparing payments and values for three homes.

  2. 02SI Employee

    Asks which home

    A forwarded call does not say which listing the buyer saw, so it asks.

  3. 03SI Employee

    Answers from listing sheet

    Confirmed facts only; anything else is marked not confirmed for the agent.

  4. 04SI Employee

    Takes the showing request

    Preferred times and whether the buyer already works with an agent.

  5. 05System

    Summary reaches the agent

    Recording, transcript and summary by email, and to the CRM by webhook.

  6. 06Your team

    Agent calls back

    Next morning: answers the open questions and confirms the showing.

Example sequence. The agent confirms every showing; nothing is promised on the agent's behalf.

The value is in the summary. When the agent rings back, they already know the buyer's payment figure, the question about the HOA and the two times that suit, so the call starts at the useful part. The solo listing agent example walks through the same pattern for sign calls taken during showings, and you can see what the SI Employee does on a call in more detail. Whatever answers the line, check its notes before they become CRM records; AI call summaries and CRM notes lists what to look for.

What buyers still want from a person

None of this research is pushing buyers away from agents. In NAR's 2026 profile, 90% of buyers used an agent or broker, the highest share in four decades[5]. Bank of America's report found that more than half of respondents still prefer real estate professionals for touring homes and for legal or contractual guidance, and Cotality found 44% of buyers would pay more for a human professional to verify AI-generated information[2].

That last figure is the opening for an agent. Buyers trust the research enough to use it and not enough to sign on it. An agent who checks their numbers without condescension, and says clearly which parts are confirmed and which are not, is doing exactly what those buyers say they would pay for. The longer argument is in will AI replace real estate agents.

Be open about your own tools

Researched buyers also ask what you use. Pew found that 76% of Americans say it is extremely or very important to be able to tell whether content was made by AI or by people[6]. If an automated assistant answers your line, say so in the greeting. If you draft listing copy or emails with AI, read them before they go out, because a buyer who has spent a month with these tools will recognize the style.

Before anything automated talks to buyers on your behalf, run it through a test plan built from real listing questions, including the ones it should decline: price negotiation, school rankings and anything about the seller's circumstances.

Frequently asked questions

How many home buyers use AI in their search?

It depends on the survey: 20% of prospective buyers and homeowners in Bank of America's 2026 report and 45% of prospective buyers in Veterans United's Q2 2026 survey[2][1]. Among buyers who already use AI, Zillow found 61% used it while buying[3].

Should I tell a buyer their AI payment estimate is wrong?

Treat it as a starting point. Ask what it includes: a total monthly payment can include taxes, insurance and mortgage insurance, and HOA dues are usually paid separately[4]. Then introduce a lender for the real figure.

Do buyers who research with AI still hire agents?

Yes. 90% of buyers used an agent or broker in NAR's 2026 profile[5], and more than half of Bank of America's respondents prefer professionals for tours and contract guidance[2].

What should I say when a buyer asks about schools or crime?

Give facts about the property and point the buyer to the school district's and police department's own published information. Do not describe neighborhoods by the people who live there; that is steering.

Can an answering assistant handle these calls after hours?

It can answer from an approved listing sheet, mark anything else as not confirmed, take a showing request and send the agent a summary with the recording. The agent confirms the showing and answers the open questions.

References

  1. Veterans United Home Loans (2026). New Survey: More Homebuyers Turning to AI Tools in 2026 (opens in a new tab) ↑

  2. National Association of REALTORS® (REALTOR® News) (2026). AI Becomes Early Step in Homebuying Journey (opens in a new tab) ↑

  3. Zillow Research (2026). Buyers: Results from the Consumer Housing Trends Report 2026 (opens in a new tab) ↑

  4. Consumer Financial Protection Bureau (2023). On a mortgage, what's the difference between my principal and interest payment and my total monthly payment? (opens in a new tab) ↑

  5. National Association of REALTORS® (2026). Top 10 Takeaways from NAR's 2026 Profile of Home Buyers and Sellers (opens in a new tab) ↑

  6. Pew Research Center (2025). How Americans View AI and Its Impact on People and Society (opens in a new tab) ↑

Book a demo

See Ava handle your calls.

Book a short demo. We’ll walk through your calls and workflow, and agree what Ava should take care of.

Book a demo