How to Buy Mortgage Leads: A Buyer Guide for 2026

The mortgage lead market in 2026

US mortgage originations are running at roughly $1.8 trillion annually in 2026, with the pay-per-lead market sized between $400 million and $600 million per year depending on how aggressively you count live transfers, call leads, and aged data. The mix of product types inside that spend has shifted noticeably over the past three years. Refinance leads dominated the market during the low-rate window of 2020 through 2022. In 2026, with rates still elevated relative to that cycle, purchase leads are cheaper per unit than refi was at the peak but carry longer conversion cycles and lower raw volume. Refi spend has compressed into rate-and-term refis for adjustable-rate resets, cash-out refis tied to home equity, and HELOC.

The product breadth is wider than most new buyers realize. Purchase, rate-and-term refi, cash-out refi, HELOC, home equity loan, reverse mortgage, VA, FHA, USDA, jumbo, and non-QM all trade in their own sub-markets with their own supplier ecosystems and their own pricing dynamics. A lender buying leads across three or four of those products will see very different cost per acquisition and very different close rates on each, and treating them as one bucket is the fastest way to lose money.

Major suppliers in 2026 include the consumer-facing portals (LendingTree, Bankrate, NerdWallet, Zillow Mortgage Marketplace, Realtor.com), the direct lenders who sell overflow (Rocket Mortgage and similar national brands periodically release overflow), the credit bureaus selling trigger lead data (Equifax, Experian, TransUnion, Innovis), independent internet publishers and affiliate networks, call centers running their own lead generation, and aged-lead resellers working 60 to 365 day old data. The most disruptive regulatory shift in the last two years was FCC 23-107, which broke the shared-consent model where one form could authorize dozens of unnamed sellers. Post-January 2024 that practice stopped being legal and the market still has not fully absorbed the change.

Lead types and pricing

Mortgage lead pricing in 2026 is heavily segmented by product, exclusivity, and freshness. The ranges below reflect typical retail pricing for a mid-sized lender buying at steady volume. Volume discounts, seasonal pricing, and supplier-specific negotiation all move these numbers.

• Trigger leads (hard inquiry). $15 to $40 retail. Sourced directly from credit bureau hard inquiry data, meaning the consumer just had their credit pulled for a mortgage application. Highest raw intent, shortest conversion window (24 to 72 hours before they commit elsewhere), heaviest compliance burden under FCRA Section 1681b pre-screen rules. Not every operation should buy these.
• Trigger leads (soft inquiry). $5 to $15. Built from soft inquiries (credit monitoring signals, rate-shopping traffic without a full mortgage pull). Lower raw intent but lighter regulatory footprint since the FCRA pre-screen framework is less constraining on soft-inquiry data.
• Refi leads (real-time). $10 to $30 exclusive, $4 to $12 shared. Consumer just filled out a refinance quote form. Exclusive versions sell to one lender. Shared versions sell to three to five lenders on the same offer (multisell). Purchase-intent is higher than aged data but varies a lot by source; a LendingTree refi is priced differently from an affiliate landing page refi.
• Purchase leads. $20 to $60 exclusive, $8 to $20 shared. Priced higher than refi because the buyer conversion cycle runs weeks to months instead of days, volume is lower across the market, and the close-rate variance is wider. Purchase leads often include bonus routing criteria (target price range, timeline, pre-approval status) that further segment pricing.
• HELOC leads. $12 to $30 exclusive. Positioned between refi and cash-out. Filterable by combined LTV, first-lien balance, and available equity.
• Reverse mortgage leads. $30 to $100. Smaller consumer pool (age 62 plus for HECM), higher lifetime value per funded loan, more expensive generation per lead. A niche but profitable segment for lenders with a reverse product line.
• VA and FHA leads. $15 to $40. Filter by military service flag (VA) or FICO floor and DTI ceilings (FHA). Government-loan shops pay more for properly qualified leads because the origination spread is better on those products.
• Aged leads (60 plus days old). $0.50 to $3. Consumer submitted a form two to twelve months ago. Low raw intent but cheap enough to work for email nurture, long-cycle conversion, and re-qualification campaigns. Good for filling gaps, bad as a primary source.
• Live-transfer leads. $75 to $250 per transfer. Consumer is already on the phone, pre-qualified by a call center, and warm-transferred to your loan officer. Highest conversion rate of any lead type, shortest cycle, highest unit cost. Common in purchase and refi, harder to source cleanly in HELOC and reverse.

Where to source mortgage leads

Supplier selection is the single biggest driver of lead economics over a one-year horizon. The supplier categories below differ on cost, quality, volume, and compliance posture. Most operators at scale run a portfolio that includes at least three of these categories at any given time.

Direct publishers. LendingTree, Bankrate, NerdWallet, Zillow Mortgage Marketplace, Realtor.com. These are the consumer-facing sites where borrowers actively shop rates. Quality is generally the best in the market, volume is large and consistent, pricing is at the top of the range. Consent capture on these platforms is well-documented, which matters for TCPA defense. The flip side: every competitor is on these marketplaces, so differentiation on response speed, quote quality, and call cadence matters more than average.

Lead aggregators. Companies that buy leads from multiple upstream sources (publishers, call centers, affiliates, their own owned-and-operated traffic) and resell them as a bundled product. Lower cost than direct publishers, quality varies widely. Vet the aggregator carefully: ask for a sample of their consent capture, TrustedForm coverage, source mix, and return policy before you commit real spend.

Credit bureau trigger programs. Equifax, Experian, TransUnion, and Innovis sell hard-inquiry and soft-inquiry trigger data directly to approved lenders. This is a regulated program under FCRA Section 1681b with firm requirements on pre-approved offer of credit and opt-out list suppression. See the companion post on mortgage trigger leads for the compliance mechanics and operational details.

Affiliate networks. Lead Marketplace, MaxBounty, PeerFly, and similar networks aggregate affiliate publishers and resell the traffic as leads. Variable quality and variable compliance. Works well when you have an affiliate manager who can vet individual offers and reject bad publishers, less well as a set-it-and-forget-it channel.

In-house traffic. Your own landing pages with paid ads (Google Search, Facebook, YouTube, TikTok for younger refi prospects). Highest quality control because you own the consent language, the form, the TrustedForm capture, and the traffic source. Highest operational overhead too. Most lenders running in-house traffic also buy from publishers and aggregators on top, treating in-house as a margin product rather than a primary volume channel.

What to vet before buying

A standard vetting checklist before sending money to a new supplier. Skipping any of these items has burned operators in the past, so build the list into your procurement process and require satisfactory answers before the first test batch.

• TCPA consent documentation. Every lead must come with a TrustedForm certificate URL, a Jornaya LeadiD, or an equivalent verifiable consent record. The certificate proves the consumer saw the consent language at a specific timestamp from a specific IP. Without this you have no TCPA defense.
• FCC 23-107 one-to-one consent. For any lead captured after January 2024, the consent must name the specific calling party. Shared consent language authorizing dozens of unnamed sellers is no longer compliant. Ask the supplier to show you the exact consent text presented to the consumer and confirm your brand is on the list.
• Return policy. Standard windows are 7 to 15 days. Disputed leads (bad phone, already closed with another lender, not-interested on first contact with documentation, fake data) should be refundable. Get the policy in writing with clear dispute reason codes. A supplier that will not refund anything is a supplier you do not want.
• Exclusivity level. Is this lead sold to one buyer (exclusive) or N buyers (shared / multisell)? If shared, how many buyers maximum? The answer should be in the contract, not verbal. Pricing and close-rate math only work when you know the denominator.
• Lead freshness. Real-time (ping-post or direct post inside 60 seconds of form submit) is priced and converts very differently from 24-hour-old or 7-day-old data. Aged data is its own category. Require the supplier to document the age.
• Source attribution. Which publisher ran this traffic? Which campaign? What did the landing page and consent language look like? Suppliers who refuse to disclose upstream sources tend to be hiding quality problems. You do not need publisher names every time, but you should be able to audit a sample.
• Credit bureau opt-out compliance. For trigger leads, the supplier must be suppressing against the bureau opt-out list (the 1-888-5-OPTOUT registry) and any internal opt-out lists. FCRA non-compliance here is a six-figure problem if it goes wrong.
• State-level restrictions. Several states impose additional restrictions on trigger lead outreach, on prerecorded calls, or on text messaging. Florida and Oklahoma notably have stricter state mini-TCPA regimes. Confirm the supplier suppresses against state registries where applicable.

How to structure a lead-buying program

Treat every new supplier as an experiment, not a commitment. The steps below apply whether you are buying $10,000 a month or $10 million a month; the dollar figures scale but the structure does not.

Start with a bounded test budget, typically $2,500 to $10,000 per new source. The point of the test is to generate enough leads to get a statistically meaningful signal on contact rate, intent rate, and close rate, not to profit on the first batch. Under-sized tests lead to decisions on noise, and over-sized tests waste money on suppliers you would have fired after 500 leads anyway.

Measure three numbers per source, broken down by product and by state. Contact rate is the percentage of leads where you reach a live person on an outbound call within your dialer cadence. Intent rate is the percentage of contacted leads that qualify to proceed (want a quote, eligible for your product, not already closed elsewhere). Close rate is the percentage of qualified leads that fund a loan. Multiply them through and you have a revenue-per-lead number you can compare against the cost-per-lead.

Watch quote-to-close ratios by source. A source with high contact rate but low close rate is usually bad-data leads or misqualified consumers. A source with low contact rate but high close rate is usually a timing or cadence issue you can fix internally. Track disposition reason codes (bad number, already closed, not interested, callback scheduled, duplicate) and chart the distribution per source. Patterns emerge fast once you have a month of data.

Negotiate return windows that match your workflow. Standard is 7 to 15 days from purchase. Push for longer on purchase leads where the sales cycle is naturally longer and shorter on real-time refi leads where a 24-hour answered-no-longer-interested return window is reasonable. Build automated return submissions into your CRM so disposition codes flow back to the supplier without manual effort.

Build a multi-source portfolio. No single supplier should exceed 40 percent of your volume. Single-supplier dependencies kill businesses when the supplier changes pricing, gets acquired, gets caught in a TCPA class action, or simply has a bad month. Then use a routing platform to distribute leads to loan officers based on state licensure, product (purchase vs refi vs HELOC vs reverse), credit bucket, and availability. Lead Router does this natively: contracts carry state license matrices, numeric range filters on FICO and LTV, loan-purpose filters, and atomic per-LO caps.

Common buyer mistakes

Scaling a new source before vetting it. The most expensive mistake. A buyer signs up, runs a 200-lead test, sees reasonable contact rate, and immediately writes a $25,000 monthly commitment. Two months in the close rate is half of projections, the supplier mix turned out to be 70 percent aged data dressed as real-time, and there is no contractual exit. Test batches exist for a reason. Do not skip them.

Ignoring consent documentation until a TCPA complaint arrives. TrustedForm certificates cost nothing to archive. Store them per lead from day one. When the first TCPA demand letter shows up (and if you buy enough leads, one will) you will need the cert URL, the consent language, the timestamp, and the IP to mount a defense. Building that archive after the fact is not an option because the certificates expire.

Buying exclusive leads without contractually verifying exclusivity. Exclusive is only worth the premium if it is actually exclusive. Some suppliers sell the same lead to two buyers simultaneously and call both of them exclusive, relying on neither buyer checking. Run your own dedup across suppliers (same email or same phone within 48 hours from two different suppliers is a flag), and put exclusivity in writing with a penalty clause.

Not running dedup across sources. The same consumer fills out three comparison forms in an hour, and three aggregators all sell their version of the lead to your company. Without cross-source dedup you pay three times for one consumer. A routing platform with a 24 to 72 hour dedup window on email plus phone catches this cleanly.

Routing all leads to one loan officer instead of matching to state licensure. An LO licensed in California cannot legally originate a loan in Texas, but mis-routing still happens when routing rules are spreadsheet-driven. State license matrices in the routing platform prevent it structurally.

Regulatory landscape in 2026

TCPA (47 U.S.C. § 227) class action activity continues at high volume. Statutory damages of $500 to $1,500 per unconsented call or text multiply quickly across large call files, and plaintiff firms target mortgage operators specifically because call volumes are high. Documented prior express written consent is the primary defense, and it has to be named, timestamped, and stored.

FCC 23-107 one-to-one consent is fully enforced for consent captured after January 2024. Pre-existing shared-consent leads grandfathered in under the old rule are increasingly scarce in 2026 because they have aged out. Any supplier still selling under the old shared-consent model is selling a liability. Verify the consent language on every lead meets the one-to-one standard for your specific brand.

FCRA § 1681b governs the trigger lead market. Credit bureaus can only sell pre-screen inquiry data to approved lenders who meet the firm offer of credit requirement, suppress against the opt-out registry, and follow specific disclosure rules when calling. Non-compliance is enforced by the CFPB and the FTC and carries both statutory and actual damages.

RESPA Section 8 restricts referral-fee kickbacks in mortgage origination. If you are paying a referral source (real estate agent, financial advisor, anyone positioned as a neutral party) for leads, the arrangement must fit an allowable exception. Consult qualified counsel before structuring referral programs. RESPA enforcement in this area has been active.

NMLS licensure is required for all mortgage-facing staff and state DFIs (Departments of Financial Institutions) add their own layers. Some states impose additional advertising disclosure rules, fee caps, or consumer disclosure requirements. A national lender should have compliance staff tracking state-level changes quarterly.

How Lead Router helps with mortgage lead routing

Lead Router is built for operators buying mortgage leads at scale and routing them to loan officers across multiple lenders, brands, or teams. Contract-based filtering routes by credit tier, LTV, loan purpose, property type, DTI, and state licensure. Each contract can carry its own numeric ranges (for example, 680 to 740 FICO cash-out refis under 80 LTV in 22 states) and the engine only matches leads that fit every filter on the contract.

Real-time ping-post distribution runs sub-100ms contract evaluation in parallel and sells each lead exclusive, multisell, or hybrid by offer. Consent metadata (TrustedForm URL, consent timestamp, IP, user agent, consent text, named sellers) is stored per lead and preserved through the two-call ping-post sequence so the record survives delivery to the buyer. Integration with Encompass, LendingPad, BytePro, Velocify, and any REST or webhook endpoint is first-class; each contract configures its own delivery shape and retry policy.

Atomic cap enforcement per loan officer per day, week, or month prevents oversell during high-velocity periods. Every routing decision is logged for TCPA and FCRA defense, including which contracts were evaluated, what filters matched, what caps were consumed, and which buyer won the lead. See the mortgage solution page for the full feature set.

FAQ

Related reading