What shared mortgage leads do to a borrower

"Who is this? Stop calling me. I never filled out anything."

That is the fourth call. Not the first. By the time it happens the borrower has already spoken to three other companies about a form they filled in once, and has stopped believing any of them.

The wording above is illustrative. Nobody publishes transcripts of those calls, so we are not going to pretend it is a quote. What sits underneath it is documented, and that is the rest of this page.

What the borrower actually did

One thing. They filled in one form, on one page, once. Everything after that is a decision somebody else made about their phone number.

Band 3 is the measured part. Source: CFPB, "Consumers' mortgage shopping experience", January 2015, Figure 6, 1,922 home purchase respondents to the first round of the National Survey of Mortgage Borrowers. Bands 1 and 2 are the mechanism, not a measurement: the number of buyers per lead varies by seller and is not published by any of them.

The regulator has already written this down

This is not a theory about how aggregators behave. It is the thing the Federal Communications Commission set out to stop, in its own words.

"Lead-generated communications are a large percentage of unwanted calls and texts and often rely on flimsy claims of consent to bombard consumers with unwanted robocalls and robotexts."

Source: FCC 23-107, Second Report and Order, CG Docket Nos. 02-278 and 21-402, adopted 13 December 2023, released 18 December 2023, at paragraph 30.

The order would have required consent to be obtained for one seller at a time. It never took effect. The Eleventh Circuit vacated it on 24 January 2025, three days before its compliance date, holding that the Commission had read more into "prior express consent" than the statute allows.

Source: Insurance Marketing Coalition Ltd. v. FCC, No. 24-10277 (11th Cir. 24 January 2025).

So the practice stands, legally and commercially. Two things follow, and they point in opposite directions.

• Nobody buying a shared lead is doing anything wrong. The market is legal and it is priced honestly.
• The regulator that looked hardest at this mechanism concluded it was producing calls consumers did not want. That is the product you are dialling into.

The arithmetic nobody puts on the invoice

Here is the part that decides whether a shared lead pays. Borrowers do not shop the way the multi-buyer model assumes they do.

• About 77 percent of home purchase borrowers applied to only one lender or broker.
• Almost half seriously considered only one before applying at all.
• Of the minority who did apply to more than one, about 80 percent were looking for better loan terms.

Source: CFPB, "Consumers' mortgage shopping experience", January 2015, Figures 5, 6 and 7. Purchase originations from 2013. The survey has run every quarter since and is now published as the National Survey of Mortgage Originations, but this remains the clearest write-up of the shopping question.

Read that against a lead sold four times. Four companies paid. At most one of them ends up with the application, because that is what the borrower does. The other three bought a phone call.

That is not an argument that the shared lead was overpriced. It is an argument that the price is the least interesting number in the transaction.

What it costs the one who dialled first

The first caller has the best moment and the worst job. They arrive before the borrower is annoyed, and they also pay for everything that comes after.

• They do the education. First call explains the process, the documents, the rate environment. Caller three inherits an informed borrower for free.
• They set the number everyone else undercuts. Whoever quotes first is the anchor. Whoever quotes last has the most information.
• They spend the borrower's patience. Every subsequent call is worse for everyone, and the person who caused none of it is the one who arrived on time.

The follow-up assumption sitting underneath all of this deserves a look too. A large study of 3.5 million leads found that 93 percent of the leads that eventually converted were reached by the sixth attempt.

Source: Velocify study of 3.5 million leads, published around 2013. The original paper is no longer hosted and every version in circulation is a secondary citation, so treat it as directional and dated. Taken apart in how many calls it takes to reach a mortgage lead.

Six attempts is fine when you are the only one making them. On a shared lead, your sixth attempt lands somewhere in the region of everyone else's sixth attempt, into a phone whose owner has learned to stop answering unknown numbers.

The number we would most like to print, and cannot

How much does a borrower's willingness to engage fall between the first call and the fourth? That is the whole argument on this page, and there is no primary source for it.

We looked. There is no independent, audited measurement of contact rate, booking rate or preapproval rate for internet mortgage leads at all, let alone one that splits them by how many companies got there first.

Every specific figure in circulation was published by somebody selling leads, or selling software to people who buy them.

The tiering behind that judgement, and what each stage of the funnel can and cannot be sourced to, is set out in what the data supports between a lead and a funded loan.

So we are leaving the gap visible rather than filling it with a number that would look better than it deserves to. Your own dialer answers this question for your own desk in a fortnight, which is more than any published figure will do.

Where the funnel is actually well measured

Worth knowing which end of this is solid. The bottom is, and it is not where deals are lost.

Once an application reaches an origination pipeline, most of them fund: 79.8 percent on purchase and 77.5 percent on refinance, in the last auditable reading of that series.

Source: ICE Mortgage Technology (formerly Ellie Mae) Origination Insight Report, December 2021. The series has no traceable data point after 2021, so any closing rate attributed to it for 2023 or later is not verifiable.

The attrition that decides a month therefore happens above that line, in exactly the stretch a shared lead scrambles. Plan against a 5% average close rate and a funded loan costs you twenty leads. Every one of those twenty has to be a conversation somebody is willing to have.

The 5% average is our own planning figure, not a published industry benchmark. There is no audited one. Use yours if you have it, and you should have it.

What we do instead, stated narrowly

We generate on our own ads and our own landing pages, so there is nothing bought in from a portal to pass along.

• Exclusive means one buyer. One lead goes to exactly one client. What it does not mean is that the person never filled in a form anywhere else, because no vendor can see that.
• The criteria are ours and they are fixed. Qualification runs before the lead leaves our system. Nobody dials the threshold down to make more leads pass, which is the point of having one.
• Speed is measured, not asserted. Median delivery is under one second, and nine of ten land in a client's CRM within five minutes.
• 94.8% of landing-page leads are delivered. The rest miss the criteria and are never sent and never billed.

That is a narrower promise than the market is used to hearing, and it is narrow on purpose. It is also available nationwide.

How to settle this on your own desk

• Ask any seller how many buyers a lead goes to. A straight answer is informative. So is a non-answer.
• Log the borrower's first sentence. Not the outcome, the opening. "Who is this" is a different product from "thanks for calling back".
• Count your own attempts to first conversation. Then compare it against the six from the Velocify work. If yours is materially worse, the lead arrived somewhere else first.
• Compare on cost per funded loan, never on cost per lead. The cheaper lead wins the first metric roughly always, and the second metric is the one that pays salaries.

Two related pages: what the data supports between a lead and a funded loan, and the note on the trigger lead ban. Buying questions are answered on our FAQ.