The Complete Guide
How Real Estate Agents Actually Get Clients
What the national survey data says about where clients really come from, and why most of the money spent chasing them is aimed at the wrong step.
Most real estate clients do not come from advertising. In the National Association of REALTORS 2025 Generational Trends Report, 40% of buyers found their agent through a referral from a friend, neighbour or relative and another 17% used an agent they had worked with before, so 57% arrived already knowing the name. Only 7% found their agent by inquiring about a property they saw online. The same report found 75% of buyers interviewed just one agent before choosing, and 81% of sellers contacted only one. That combination changes what the work is. The contest is usually decided before any comparison happens, so the job is to be the name that gets mentioned, and to hold up when someone checks it.
Key takeaways
- 57% of buyers used an agent they already knew or were referred to: 40% by a friend, neighbour or relative, 17% an agent they had used before (NAR 2025 Generational Trends Report, Exhibit 4-4).
- For sellers the pattern is stronger at 66%: 38% referred by someone they know, 28% an agent they had used before (Exhibit 7-1).
- Only 7% of buyers found their agent by inquiring about a specific property seen online, and 6% through a website with no specific reference. Those two numbers are the honest size of the portal-and-form channel.
- 75% of buyers interviewed only one agent, and 81% of sellers contacted only one. Most decisions are not comparisons, they are confirmations.
- 51% of buyers found the home they bought on the internet and 29% found it through their agent, so buyers search independently and then still need an agent (Exhibit 3-4).
- New agents cannot rely on the referral channel. NAR's 2024 Member Profile shows agents with two years of experience or less get less than 1% of business from repeat clients, against 42% for agents with 16 years or more.
- The widely quoted claim that responding in 5 minutes makes you 100 times more likely to reach a lead does not come from Harvard Business Review, and neither study behind these numbers covered real estate.
Start with who is actually choosing you
There is a large industry selling real estate leads, and it has shaped how agents talk about getting clients. The vocabulary is all about capture: funnels, lead magnets, cost per lead. That language assumes the buyer starts as a stranger who has to be caught.
The survey data describes something different. NAR mails its buyer and seller survey to people who actually completed a transaction, and asks how they found their agent. In the 2025 Generational Trends Report, covering purchases from July 2023 to June 2024, 40% of buyers were referred by a friend, neighbour or relative, and 17% used an agent they had already worked with. Add those and 57% of buyers were not strangers at any point.
Sellers lean even further that way. 38% came through a referral from someone they know and 28% returned to an agent they had used before, so 66% arrived with the relationship already in place.
Now put the two channels side by side. Referral and repeat business account for 57% of buyers. Inquiring about a specific property seen online accounts for 7%, and finding an agent through a website with no specific reference accounts for 6%. The channel the industry sells hardest is roughly a fifth the size of the one it barely discusses.
The number that changes the strategy: most people interview one agent
This is the finding that reorganises everything else, and it is the one most rarely quoted. 75% of buyers interviewed only one real estate agent before choosing. Among sellers, 81% contacted only one agent.
Read that carefully, because the usual mental model is a bake-off. Agents prepare for a competitive listing presentation, sharpen their differentiators, and practise answers to why they should be hired over someone else. For roughly four out of five sellers, that competition never happens.
What happens instead is a confirmation. Someone gets a name, usually from a person they trust. They look that name up. If nothing they find contradicts the recommendation, they call, and the process ends there. The other agents were never in the room.
So the two moments that decide most transactions are the mention and the check. Being mentioned is relationship work: staying in contact with people who already know you, and being specific enough about what you do that someone can recommend you in a sentence. Surviving the check is mostly your website, your Google Business Profile, and your reviews, because that is what a person finds when they look you up before calling.
Why buyers still need you after finding the house themselves
A second pair of numbers explains why the referral pattern holds even though buyers do their own searching. 51% of buyers found the home they purchased on the internet. 29% found it through their real estate agent.
The first number is often used to argue that agents are being replaced. The second, and the fact that 88% of buyers still purchased through an agent or broker in NAR's 2025 Profile, says something more precise. Finding a listing and completing a purchase are different jobs. Buyers took over the first one years ago. They did not take over the second.
This matters for how you spend effort on your website. Competing with the portals on listing search is competing on inventory size, which you will lose. The questions buyers cannot answer from a portal are the ones about judgment: whether a building has assessment history worth worrying about, what a specific neighbourhood is like for a family, whether an asking price is defensible. Those are answerable on your own pages, and they are also the questions an AI assistant has no portal data to resolve.
The age split is worth knowing if you are deciding where to put your time. Among buyers aged 26 to 34, the internet found the home 63% of the time. Among buyers aged 79 and over, it was 28%, and the agent found the home 45% of the time.
The honest problem with referral advice for a new agent
Advice built on referral data has a gap, and it is worth naming rather than working around, because the people most desperate for clients are the ones it fails.
Referrals and repeat business come from past clients. An agent in year one has none. NAR's 2024 Member Profile is the only free NAR document that breaks this down by experience, and the table is blunt. For agents with two years of experience or less, the median share of business from repeat clients is reported as less than 1%, and the same for referrals from past clients. For agents with 16 years or more, those figures are 42% and 29%.
The same table shows the consequence. 62% of members with two years of experience or less earned under $10,000 in the survey year, against 9% of members with 16 or more years.
So the strategy genuinely differs by stage. An established agent should protect and systematise the channel already producing most of the business. A new agent has to manufacture the first relationships through methods that do not require a past client list, then convert those into the referral base that eventually carries the business. The mistake is a new agent copying a veteran's advice about referrals, or a veteran buying leads they do not need.
What this means for the website
It is tempting to read all of this as evidence that a website barely matters. That reading gets the mechanism backwards.
If most clients arrive by referral, and the step between the referral and the call is someone looking you up, then the website's main job is not to capture strangers. It is to confirm a recommendation that has already been made. That is a different design brief and it changes what belongs on the page: proof you have done this work, clarity about what you specialise in, an answer to the questions a cautious person has before calling, and a way to make contact that does not feel like submitting to a sales machine.
It also raises the cost of an unconvincing site in a way that is invisible in your analytics. When a referred person looks you up and quietly decides not to call, nothing appears in a report. There was no form abandonment and no bounce worth investigating. The referral simply did not convert, and the usual conclusion is that the lead was weak.
The same logic now extends to AI assistants. When someone asks ChatGPT or Google's AI mode about an agent by name, or asks for a recommendation in a specific area, the answer is assembled from what those systems can read about you across your site and your profiles. That is the modern version of the check, and it fails in the same silent way.
How to use this hub
The five pages below follow the order the decisions actually happen in. The first two settle where clients come from and whether buying leads makes sense for your situation. The middle two cover being the name that gets mentioned and surviving the look-up. The last one covers what to do when someone does reach out, including the response-time research that is misquoted almost everywhere.
Every figure on these pages is attributed to the document it came from, and no company that sells leads is used as a source for a claim about how well lead generation works.
| How they found the agent | Buyers | Sellers |
|---|---|---|
| Referred by a friend, neighbour or relative | 40% | 38% |
| Used an agent they had worked with before | 17% | 28% |
| Referred by another agent or broker | 7% | 4% |
| Inquired about a specific property seen online | 7% | Not reported separately |
| Website with no specific reference | 6% | 4% |
| Saw contact information on a For Sale or Open House sign | 5% | 2% |
NAR 2025 Home Buyers and Sellers Generational Trends Report, Exhibit 4-4 (buyers, p. 64) and Exhibit 7-1 (sellers, p. 119), published April 2025 and covering the July 2023 to June 2024 survey year. Rows shown are the largest categories, so columns do not total 100%.
| Number of agents | Buyers who interviewed | Sellers who contacted |
|---|---|---|
| One | 75% | 81% |
| Two | 16% | 15% |
| Three | 7% | 6% |
| Four or more | 3% | 2% |
NAR 2025 Generational Trends Report, Exhibit 4-6 (p. 66) and Exhibit 7-2 (p. 120). Buyer percentages are rounded and sum to 101%.
| Experience | From repeat clients | From past-client referrals | Median transactions |
|---|---|---|---|
| 2 years or less | Less than 1% | Less than 1% | 2 |
| 3 to 5 years | 8% | 17% | 8 |
| 6 to 15 years | 22% | 24% | 12 |
| 16 years or more | 42% | 29% | 12 |
| All members | 20% | 21% | 10 |
NAR 2024 Member Profile highlights, p. 5. Figures are medians. This is the most recent free NAR document that splits repeat and referral business by experience level; later editions report the all-member figure only.
Frequently asked questions
How do most real estate agents get their clients?
Most real estate agents get clients through referrals and repeat business rather than advertising. In NAR's 2025 Generational Trends Report, 40% of buyers found their agent through a friend, neighbour or relative and 17% used an agent they had worked with before, totalling 57%. For sellers the same two categories reach 66%. By comparison, 7% of buyers found their agent by inquiring about a property seen online.
Is buying real estate leads worth it?
Buying real estate leads is worth it only when you can answer two questions with your own numbers: what a closed transaction from that source costs you after every unconverted lead, and whether you have the follow-up capacity to work them. Purchased leads arrive as strangers, while NAR data shows 57% of buyers pick an agent they already know or were referred to. Paid leads compete on speed and volume, not on trust.
How many agents do most sellers interview before choosing one?
Most sellers interview only one agent. NAR's 2025 Generational Trends Report found 81% of sellers contacted just one agent before selecting the one they worked with, with 15% contacting two and 6% contacting three. Among buyers, 75% interviewed only one agent. This means most listing decisions are confirmations of an existing recommendation rather than competitive comparisons between several agents.
Do real estate agents still matter if buyers find homes online?
Yes, because finding a listing and completing a purchase are separate jobs. NAR's 2025 Generational Trends Report found 51% of buyers located the home they purchased on the internet, yet NAR's 2025 Profile of Home Buyers and Sellers found 88% of buyers still purchased through an agent or broker. Buyers took over property search years ago. They did not take over negotiation, pricing judgment, or transaction management.
How does a brand new real estate agent get clients without referrals?
A new real estate agent has no past clients to refer them, and NAR's 2024 Member Profile confirms the gap: agents with two years of experience or less report less than 1% of business from repeat clients, against 42% for agents with 16 years or more. The realistic path is to manufacture first relationships through methods that need no client history, such as open houses, a specific geographic or property-type focus, and consistent contact with people who already know you personally.
Why does my real estate website get traffic but no calls?
A real estate website that gets traffic but no calls is usually failing the confirmation step rather than the discovery step. Because 57% of buyers arrive already holding your name, much of your traffic is people checking you out after a referral. If the site does not show proof of relevant work, state what you specialise in, or answer the questions a cautious person has, they leave without calling and nothing appears in your analytics to explain it.
Is it true that responding to a lead in five minutes makes you 100 times more likely to reach them?
That figure is real but it is almost always misattributed. The 100 times claim comes from a 2007 Lead Response Management study by James Oldroyd conducted with InsideSales.com, a company selling lead-response software, and the comparison is 5 minutes versus 30 minutes. It is not from Harvard Business Review, whose 2011 article reported that contacting within an hour made firms about seven times more likely to qualify a lead. Neither study included real estate.
What percentage of real estate business comes from repeat clients?
Across all NAR members, the median share of business from repeat clients was 20%, with another 21% from referrals by past clients, according to the 2024 Member Profile. The figure depends heavily on experience: agents with 16 years or more report 42% from repeat clients, while agents with two years or less report less than 1%. NAR's 2026 Member Profile reports 28% of business coming from past clients and customers.
Should I focus on getting more leads or converting the ones I have?
Work out which step is actually failing before spending anything. If people contact you and do not become clients, more leads multiply the same loss. If almost nobody contacts you despite people knowing your name, the problem sits in the look-up step, usually your website, reviews, or Google Business Profile. NAR data showing 75% of buyers interview only one agent means a lost enquiry is rarely a lost comparison, it is a call that never happened.
How much do real estate agents spend on lead generation?
There is no reliable public figure isolating lead-generation spend, and any specific number quoted without a source should be treated as marketing. NAR reports total median business expenses of $8,010 for 2024 and $9,530 for 2025 across all categories, with vehicle costs the single largest at $1,580, but the detailed expense breakdown sits behind a paywall. Calculate your own cost per closed transaction rather than benchmarking against an unsourced average.
Does a website still matter if most clients come from referrals?
It matters more, because the referral does not end at the mention. Someone hearing your name looks you up before calling, and that check is where a recommendation is either confirmed or quietly dropped. With 75% of buyers interviewing only one agent, there is no second round to recover in. The website's job in a referral-driven business is confirmation rather than capture, which changes what belongs on it.
Related reading from Plot
Sources
Every claim on this page that could be checked against a primary source is linked below. Where something is not publicly documented by a vendor, the page says so rather than filling the gap with an estimate.
- 2025 Home Buyers and Sellers Generational Trends Report (National Association of REALTORS)
- 2025 Profile of Home Buyers and Sellers (highlights) (National Association of REALTORS)
- 2024 Member Profile highlights (National Association of REALTORS)
- Experienced REALTORS Anchor the Industry as Housing Affordability Remains Top Hurdle (National Association of REALTORS)
- The Short Life of Online Sales Leads (March 2011) (Harvard Business Review)
Want to know how your own site scores?
Every check on this page applies to a real website. Paste in your current site and Saige rebuilds it with readable text, stated facts, and structured data already in place, so you can see the difference rather than reading about it.
See how the rebuild works