Real Estate Leads That Pay at Closing vs Pay Per Lead: What Each One Costs

Quick answers: Do pay-at-closing leads cost more than buying leads? · What referral fee do pay-at-closing networks charge? · Why do most referral networks not publish a fee? · Are pay-at-closing leads better quality? · How many leads does it take to close one deal? · When is paying per lead the cheaper choice?

What pay at closing actually takes out of your commission

A pay-at-closing network does not sell you a lead. It refers you a consumer and takes a slice of the commission you earn if that consumer transacts with you, which means nothing leaves your account when the lead arrives, when it goes cold, or when the consumer buys with somebody else. The invoice shows up at the closing table or it never shows up at all.

That risk shape is the whole appeal, and it is a real advantage. An agent with time and no marketing budget can work these referrals without ever writing a check. The Consumer Policy Center, a consumer advocacy research group, put the reason plainly in February 2026: “many agents are more willing to pay a percentage of their earned commissions rather than monthly marketing fees that guarantee no sales.”

But the same feature makes the price hard to see. A per-lead invoice states its number in advance. A referral fee states a percentage, and the dollars behind it depend on a sale that has not happened yet, so almost nobody converts it into money before signing.

So convert it. The median existing-home price in June 2026 was $440,600 across all housing types, according to the National Association of Realtors. Assume, according to your own last twelve months rather than any published rate, that you earn 2.5% on your side. That is a gross commission of $11,015, and a 33% referral fee takes $3,635 of it.

The fee is 30 to 40 percent, and it climbs with the price of the house

The best current documentation of what these networks charge is not from the networks. It is a February 2026 report from the Consumer Policy Center by Wendy Gilch and Stephen Brobeck, which found that referral fees “clearly are rising, with most now 30-40 percent of the agent commission (before the agent’s broker takes their share).” The same report tracks the drift: “Since the 1990s, fees from referral companies have tended to increase from 20-30 percent to 30-40 percent.”

The tiering is the part almost nobody knows. Two of the largest charge by price band:

• Zillow, according to the report, “charges 40% on all sales roughly above median area home prices” and “only 25% on sales of lower-priced homes whose ceilings range from $150,000 to $450,000.”
• Redfin’s fee, according to the same report, is “33% for all sale prices between $100,000 and $500,000 but 40% for all sale prices at or above $900,000.”

That inverts the usual intuition. You are not paying a fixed percentage of a bigger commission on a bigger house; on the expensive house the percentage itself steps up, so the fee rises twice over.

The Consumer Policy Center is an advocacy organization arguing that these fees keep commissions inflated, and it should be read as such. Zillow rejects the findings, telling Real Estate News that referral fees are “a business and marketing expense agents pay in the same way they might otherwise have spent thousands on billboards, print ads, and other marketing tools.” That dispute is about whether the fees inflate what consumers pay, not about the percentages themselves, and the percentages are what concern you.

Where companies do publish a rate, they corroborate the range:

• HomeLight, according to its own help site current as of January 23, 2025, describes its terms as covering “commission (either 30 or 33%).”
• Agent Pronto states on its own site that its “referral fee typically ranges from 25 to 35% of the gross referred-side commission,” and defines that base as what the receiving broker gets before any split.
• Sold.com charges “a 30-35% referral fee based on your gross commission,” according to HousingWire, updated June 29, 2026.

Most networks still will not publish their own number

Two companies stating a rate is not transparency. When The Close reviewed eight pay-at-closing platforms in July 2026, five disclosed nothing: Clever appeared as “Varies by agreement,” Zillow Preferred as “Varies by market and transaction,” and SOLD.com, ReadyConnect Concierge and ReferralExchange each simply as “Varies.”

Realtor.com’s referral arm shows how complete that silence can be. Move, Inc.’s twenty-page agent deck for ReadyConnect Concierge, hosted on NAR’s own site, pitches throughout that you pay nothing until a referral closes: “ReadyConnect only gets paid when your referral closes. We win when you win.” That slide carries one asterisk, and the asterisk reads “See your broker for specific questions about your referral fee.” Twenty pages selling a payment model, with the payment handed off to somebody not in the room.

Clever is the same, promising “there’s never any upfront fees for agents, we get paid a referral fee at closing” and stating no percentage anywhere.

The practical consequence is that you cannot comparison-shop this market from public information. Get your actual percentage in writing, ask whether it steps up at a price threshold, and ask where that threshold sits in your market.

The break-even sits between 61 and 209 leads per closing

Now price the other side, because somebody does publish it. CINC, which says its team “manages almost $30 million in search and social advertising spend annually for real estate teams and agents,” reported that “the overall Google Seller CPL in CINC’s portfolio came in at $17.40 in Q4 2025”. According to the same report, that was “a 22% increase versus the same period in 2024.”

Read the fence around that number. It is seller leads, not buyers, on Google search rather than social, in one company’s book of business rather than the market. Nor does the 22% rise mean lead prices are climbing that fast everywhere, according to CINC itself, which says “a meaningful portion of that increase reflects a deliberate shift in CINC’s campaign strategy toward higher-quality seller leads, not simply market-driven inflation.” It is also a raw lead, where a referral network hands you a consumer somebody has already called and screened. Those are different products at different stages of the funnel, and what the comparison establishes is the scale of the gap between them.

The arithmetic is three lines.

Run the same lines at the ends of the Consumer Policy Center range and, according to that report, the 30% floor gives a fee of $3,305 or 190 leads, while the 40% ceiling gives $4,406 or 253 leads.

Which seller lead you are buying changes that denominator, so check it before you use the number. CINC’s $17.40 is media cost, what its clients’ campaigns paid Google for a form fill, across a book its own team describes as “almost $30 million in search and social advertising spend annually.” Buy the same lead as a finished product and you pay for the campaign management too. Real Geeks lists its Seller package at $599 a month against a contractual guarantee of 10 to 15 seller leads a month, which is $39.93 to $59.90 each. The same $3,635 buys 61 to 91 of those, not 209. Both figures are right. They price different products.

So the question stops being “which is cheaper” and becomes one you can answer: does your desk work more or fewer leads per closing than the 61 to 209 that fee would have bought?

Where the 25-leads-per-sale figure comes from

According to Move’s own deck, the market-wide ratio sits far below that break-even, and the decade-long shift is the real story:

Label that figure honestly. Move calls it an estimate and discloses the method on the slide, saying the data “includes assumptions on competitor lead conversion rates.” It is a vendor’s estimate of a market, published in 2021, and your desk is not the market. Use it to sanity-check your own ratio, never to replace it.

The distance is still hard to argue away. According to those figures, a market-wide ratio near 25 sits roughly eight times below the 209 that would make a 33% fee the cheaper option. For the referral model to win on price alone, your leads-per-closing would have to run several times worse than the market average. None of the networks publish that comparison, though nothing about it is difficult to work out.

What the referral fee is really buying is the first minute

The honest case for paying at closing was never that it is cheap. It is that the network does work you are probably not doing. According to Move’s 2021 agent deck, its team reaches consumers “within 10 seconds of online inquiry” and calls them from 9 a.m. to 7 p.m. local time, 365 days a year. Treat all of it as five-year-old vendor marketing, because that is what it is.

Set that against what the same 2021 deck says agents actually manage:

• “Over half of the agents surveyed were unable to connect with the consumer that submitted the inquiry” (Move internal lead-quality study, July 2018).
• According to InsideSales.com research from January 2016, most real estate leads are contacted 1.5 times before agents give up. Reaching 90% of them takes six to nine calls, according to the same research.
• According to Vendasta in May 2018, quoted in that deck, conversion rises 391% when contact happens inside the first minute.

Put those together and the fee looks less like a lead price and more like an outsourcing charge for speed and persistence. If leads bought at that price sit untouched for three hours and then get a call and a half before you give up on them, the cheaper model on the spreadsheet is quietly the more expensive one in practice. The saving you booked never existed.

One mechanic to check before you sign. The same deck shows a referral going to a first agent with exclusive access, then opening to a second and third agent with shared access. Ask how long your exclusive window lasts, because a referral you share is a different product from one you do not, and exclusive and shared leads do not convert alike.

Which model fits your desk

Neither model wins outright, and the conditions that decide it are specific enough to check against your own book.

Pay at closing fits an agent or team with genuine capacity and no working capital, or one testing a new geography without committing spend. It fits best at lower price points, where the percentage is smaller in absolute terms and, at Zillow and Redfin, a lower rate outright. It fits anyone whose leads-per-closing ratio is poor, because a fee owed only on success caps your downside at zero in a quarter where nothing closes, and that protection is worth real money the arithmetic above does not capture.

Paying per lead fits anyone whose ratio is near the market-wide 25, anyone working higher price points where the percentage steps up against them, and anyone who wants the lead to belong to them. It demands cash before anything closes, and the speed-to-contact discipline you would otherwise be renting. It also lets you verify a lead before you bid on it, which the pre-packaged referral does not.

The same tension runs through every vertical that buys volume, which is why the retainer against pay per lead argument in home services reads so familiarly here.

Frequently Asked Questions

Do pay-at-closing leads cost more than buying leads?

On the published arithmetic, usually. According to NAR’s June 2026 median of $440,600, a 33% referral fee on a median-priced home comes to about $3,635 on a commission assumption of 2.5%. That buys 209 seller leads at CINC’s Q4 2025 Google cost per lead of $17.40, or 61 to 91 at the $39.93 to $59.90 Real Geeks publishes for a contractually guaranteed seller lead. Move’s own deck estimates the market-wide ratio at roughly 25 leads per sale, several times below that break-even. The exception is a desk that converts far worse than the market, or one with no cash to buy leads at all, where a fee owed only on a closing is worth paying for the certainty.

What referral fee do pay-at-closing networks charge?

According to a February 2026 Consumer Policy Center report, most charge 30% to 40% of the agent’s commission, up from 20% to 30% in the 1990s. According to their own sites, HomeLight describes its terms as “either 30 or 33%” as of January 2025, and Agent Pronto states 25% to 35% of the gross referred-side commission. According to that same report, Zillow charges 40% above roughly median area prices and 25% on lower-priced homes, while Redfin charges 33% from $100,000 to $500,000 rising to 40% at $900,000 and above. Get your own rate in writing.

Why do most referral networks not publish a fee?

They do not say, so this is better described than explained. Of eight platforms The Close reviewed in July 2026, five listed no number, using wording such as “Varies by agreement” or “Varies by market and transaction.” Move, Inc.’s twenty-page ReadyConnect Concierge deck sells the pay-at-closing model throughout and handles the fee with a single asterisk reading “See your broker for specific questions about your referral fee.” A fee that genuinely varies by market and price tier is harder to publish than a flat one. It is also harder to compare, which is the practical effect on you.

Are pay-at-closing leads better quality?

They are further along the funnel, which is a different claim. A referral network screens the consumer and hands over someone who has agreed to speak to an agent, while a Google form fill is a name and a number. Move, Inc.’s 2021 agent deck says its team contacts consumers within 10 seconds of the inquiry, seven days a week, and that head start is most of what the fee buys. What no network publishes is the number that would actually prove a quality claim, which is how many of its referrals close. Ask for it, and ask how long your exclusive window runs before the referral is shared with other agents.

How many leads does it take to close one deal?

Nobody can tell you your number, and your number is the only one that decides this. As a market-wide reference, according to Move’s own deck, the industry generated roughly 25 leads for every home sale in 2021, up from about one lead per sale a decade earlier. Move discloses that the estimate rests on assumptions about competitor conversion rates, so treat it as a sanity check rather than a benchmark. Contact discipline moves the ratio hard: according to research that deck cites, most leads get 1.5 calls before agents give up. Reaching 90% of them takes six to nine calls, according to the same research.

When is paying per lead the cheaper choice?

Whenever your desk works fewer than 61 to 209 leads per closing. The low end of that band applies if you buy vendor-guaranteed seller leads at $40 to $60, the high end if you run your own Google campaigns nearer $17. At the market-wide ratio near 25, buying leads is cheaper by a wide margin. The comparison assumes you actually work them, and it holds only if you can fund the spend before anything closes. Higher price points push the answer further toward paying per lead, because Zillow and Redfin both raise the percentage on more expensive homes. Recalculate whenever your own cost per lead moves, using the price you actually pay rather than any published figure.

The number that settles this is already in your CRM

Neither side of this argument can tell you what you need to know. Most networks will not publish a fee, and none publishes the close rate that would justify one. The lead vendors publish a cost per lead that says nothing about what happens after the lead lands on your desk. The one figure that decides the question is how many leads your team worked for each closing over the last twelve months, and you already own it.

Run it before your next sales call, because it changes what the call is about. You stop asking what the fee is and start asking what a lead costs in writing, and which product that price buys: a raw form fill, or a screened one somebody has already spoken to. If we are the ones on the call, ask us for cost per lead in writing on our real estate lead generation programs.