Speed to Lead in Mortgage: What the Number Really Means

Speed to lead is the elapsed time between a borrower raising their hand and someone actually reaching them.

Not the time to an auto-responder. Not the time to a first dial attempt. The time to a conversation.

That distinction matters more than any benchmark. Most systems report the metric they can measure, not the one that decides anything.

Speed to Lead Is Three Problems, Not One

The category treats this as one number applied to every lead. It is three different mechanisms and they fail in different ways.

Shopped and Shared Leads: the Clock Is Real

A publisher sells the same record to four or five lenders at once. Everyone gets it within seconds of each other.

Here speed is not about the borrower cooling off. It is about who reaches them first. The borrower is about to have four conversations and the first one sets the frame.

This is the only mode where the race is literal. It is also the mode with the least research behind it, which is the subject of the next section.

Purchased leads are the extreme version. You paid for a record four competitors also paid for. The only variable you control is how fast someone dials.

Getting there first is a distribution problem before it is an effort problem. A record that sits unassigned for ninety seconds has already lost, whatever the officer does next. That is what mortgage CRM software is for.

Referrals: They Will Wait, and Silence Still Loses Them

A realtor referral arrives pre-sold. Somebody the borrower trusts already told them to call you.

They are not shopping. They will wait a day. A five-minute rule is the wrong frame and applying it here wastes urgency you need elsewhere.

What loses a referral is silence. Not slowness, silence. Three days of nothing tells the borrower you are disorganized, and it tells the agent the same thing.

The failure mode is different, so the fix is different. Consistency, not speed.

Inbound Calls: There Is No Window

Somebody dials your number. There is no response window at all. Either a human picks up or nobody does.

You cannot be fast at a call you did not answer. This is not a cadence problem and no follow-up sequence repairs it.

Invoca measures 44% of all inbound calls going unanswered, and 29% for calls lasting more than thirty seconds. Those are cross-industry figures with a stated method, which is more than most numbers in this category carry.

The Research Was Run on the Wrong Kind of Lead

Speed matters most on shared leads. That is also the case with the least evidence behind it, and almost nobody says so.

The study underneath most of this category is Oldroyd, McElheran and Elkington, published in Harvard Business Review in March 2011. It is a serious piece of work and it is public.

Here is the method. The researchers submitted test enquiries to 2,241 US companies through those companies’ own websites. One company, one form, one responder.

Nobody else was dialing that number. There was no race.

So what the audit measured is how a lead’s reachability and intent decay over time. That is a real effect and it is worth knowing.

It is not the mechanism operating on a LendingTree or LowerMyBills record sold to five lenders at once. There, the borrower is not cooling off. They are being called by your competitors.

The industry takes intent-decay research from exclusive first-party leads. It then cites that to justify speed on shared leads. There the argument is competitive displacement. The conclusion happens to be right. The evidence does not cover it.

What the research actually maps to

Lead type	What decides the outcome	What the cited research covers
Shopped and shared	Who reaches the borrower first	Nothing cited measures it
First-party inbound web	Intent and reachability decay	HBR 2011
Referrals	Consistency over weeks	Nothing
Inbound calls	Whether anyone answers	Nothing

One square filled out of four.

The part that cannot be checked at all

The 100x and 21x multipliers everyone quotes are not from the 2011 paper. They come from a separate 2007 study by the same lead author with InsideSales.com, presented at a marketing conference.

That study does not disclose where its leads came from. Nobody says whether those call logs were first-party form fills or aggregator feeds.

That is worse than a wrong attribution. If you cannot tell whether the leads were exclusive or shared, the finding cannot be applied to either.

Why the Fix Is Availability, Not Discipline

Most writing on this subject treats slow response as a character problem. Batch processing, the wrong mindset, insufficient hunger.

That reading survives because the people writing it have never run a floor.

A producing loan officer is the least available person in the shop. They are at a closing table. They are on an underwriting call. They are sitting with a realtor who might send three more deals this quarter.

Every one of those is a good reason not to answer. Add them up and a top producer is unreachable for large parts of a working day. That is physics, not discipline.

So coverage is the actual question. Who answers when the assigned officer cannot, and what are they allowed to do when they answer.

There are three honest answers and most shops need more than one.

Route on availability rather than roster. Next in rotation is not the same as next available. An officer set to unavailable should not receive the lead, and the state licensing filter runs before the fairness rule, not after it.

Pool the lead after a defined silence. If nobody has touched it inside the window, it stops belonging to one person. The window is not one number. A shopped lead with one dial and two hours of silence is already late. A referral is not.

Let something always-on answer. Nights, weekends, and the ninety minutes your best officer is at a closing table. This is the only one of the three that covers the inbound call problem, because routing and pooling both assume a record you can call back.

Shape’s AI agent handles the third. It can warm transfer to a licensed officer. It can blast a group until one takes it. It can book on the assigned officer’s calendar. The point is not that the machine sells anything. The point is that nobody hangs up on silence.

None of that works without the routing rules underneath it. That is where automation fits the loan lifecycle rather than sitting beside it as a separate tool.

What the Data Actually Supports

Strip out everything unsourced and one figure survives with its method attached.

Firms responding within an hour were about seven times more likely to hold a qualifying conversation. That is measured against firms responding an hour later. Against firms that waited a full day, more than sixty times.

Seven, not a hundred. Same lead author, different study, an order of magnitude apart. The smaller number is the one with the published method.

The same audit found 37% of companies responded within an hour and 23% never responded at all. Among those who did, the average was 42 hours.

That last figure is the useful one for a lender. Not because 42 hours is a target. Because it tells you the bar in the wider economy is on the floor.

On the borrower side, the CFPB found 77% of mortgage borrowers apply to only one lender. Almost half seriously consider only one before applying. That is the closest thing to a first-responder finding that traces to a real study.

Read it carefully. It says most borrowers do not shop, which is a stronger argument for reaching them early than any conversion multiple.

What to Measure Instead of a Single Number

One average across every lead type tells you nothing, because you are averaging three mechanisms.

Four things worth tracking separately, by lead source.

Contact rate, not response time. Whether you reached a human. Time to first attempt flatters everyone.

Time to first conversation against time to first attempt. The gap between them is the number nobody looks at, and it is where the losses live.

Attempts before contact. Shape’s platform data runs to about 17 attempts to convert an opportunity. That covers calls, texts and email, over fifteen years, across every industry we serve. It is not mortgage-specific and we do not present it as one.

Answer rate in and out of business hours. Split them. An operation that answers well at 2pm and not at all at 7pm has a staffing problem wearing a technology costume.

Which lead goes to which officer, and in what order, decides most of this before any clock starts. That sits upstream in routing, prioritizing, and working leads.

Frequently Asked Questions

What is a good speed to lead time for mortgage leads?+

It depends on the lead type, and one target across all of them is the common mistake. On shopped and purchased leads, minutes, because competitors hold the same record. On first-party web leads, inside the hour is supported by research. On referrals, same day is fine and consistency matters more than speed.

Does the five-minute rule apply to mortgage?+

Partly. The five-minute figure comes from a 2007 vendor study that never disclosed where its leads came from. It cannot be applied confidently to any lead type. The better-sourced finding is seven times more qualifying conversations for responding within the hour. That comes from a 2011 Harvard Business Review audit.

Is speed to lead different for referrals?+

Yes, and treating them the same wastes urgency. A referral arrives pre-sold and will usually wait a day. What loses a referral is silence rather than slowness. Three days of nothing signals disorganization to the borrower and to the agent who sent them.

Why do loan officers respond slowly to leads?+

Usually availability rather than discipline. A producing officer is at a closing table, on an underwriting call, or with a referral partner. Those are good reasons not to answer, and they add up to large parts of a working day. The fix is coverage and routing, not exhortation.

How many attempts does it take to reach a mortgage lead?+

Shape’s platform data averages about 17 attempts to convert an opportunity. That covers calls, texts and email over fifteen years, across every industry we serve. It is not mortgage-specific. The practical point is that most shops stop at three or four. That is well short of where contact happens.