Research
50K+

leads analyzed in this study

Speed to Lead: The Harvard Study, Explained for 2026
H
Hayk Muradyan
RevOps, SayVo.ai · March 4, 2026 · 8 min read
speed to lead harvard studyspeed to leadlead response time studyMIT lead response management study21x lead response statistic5 minute lead response

The “Harvard speed-to-lead study” is really two pieces of research by the same lead author. The famous numbers come from the first one: a 2007 study by Dr. James Oldroyd and InsideSales.com found the odds of reaching a web lead drop 100 times, and the odds of qualifying it drop 21 times, when the first call comes at 30 minutes instead of 5.[1]The second is the 2011 Harvard Business Review article “The Short Life of Online Sales Leads,” which made the idea famous.[2] Both still matter: 35% of the mortgage brokerage owners we talk to name slow lead response as a top problem.[3] This guide covers what each study measured, what people get wrong, and what to do about it in 2026.

TL;DR
  • The 100x and 21x numbers trace to a 2007 InsideSales.com and MIT study, not to the 2011 Harvard Business Review article.
  • They are odds of contacting and qualifying a lead. The study did not measure closed deals.
  • Most of the loss happens fast: the odds of contact fall 5 times between minute 5 and minute 10.
  • No person can call every lead in under five minutes all day. A system can.
100x
drop in the odds of contacting a lead called at 30 minutes instead of 5
Source: InsideSales.com / MIT, 2007
21x
drop in the odds of qualifying a lead called at 30 minutes instead of 5
Source: InsideSales.com / MIT, 2007
35%
of mortgage brokerage owners name speed to lead as a top problem
Source: SayVo, 2026

What is the Harvard speed-to-lead study?

When people say “the Harvard study,” they usually mean an article in the March 2011 issue of Harvard Business Review called “The Short Life of Online Sales Leads.” It was written by James B. Oldroyd, Kristina McElheran of Harvard Business School and David Elkington of InsideSales.com.[2]Its public summary says it plainly: “most companies are not responding nearly fast enough.”

The full article sits behind Harvard Business Review's paywall. We only repeat numbers we can open and check, so this page does not quote figures from inside it. If you see a statistic credited to Harvard, ask where it was actually published. Often it comes from the earlier study below.

Where do the 100x and 21x numbers really come from?

They come from the Lead Response Management Study, presented in October 2007 by Dr. Oldroyd, then a faculty fellow at MIT, and Dave Elkington, the CEO of InsideSales.com.[1]That is why it is often called “the MIT study.” Four years later the same two researchers co-wrote the Harvard Business Review article, and the two got blended together online.

The 2007 study looked at three years of data from six companies that call web leads. It covered more than 15,000 leads and more than 100,000 call attempts. It asked one question: when should a company call a web lead to get the best contact and qualification rates?[1]

Lead Response Management StudyThe Short Life of Online Sales Leads
Year20072011
Where it appearedPresented at a MarketingSherpa summit, published by InsideSales.comHarvard Business Review, March 2011 issue
AuthorsDr. James Oldroyd (then at MIT) and Dave ElkingtonJames Oldroyd, Kristina McElheran and David Elkington
Known forThe 100x and 21x numbers, best days and best hours to callMaking slow lead response a boardroom topic
Can you read it free?Yes, copies of the summary are publicNo, the full text is paywalled

What did the study find about calling within five minutes?

The study sliced the first three hours after a lead came in into five-minute blocks. The drop was steep and it started right away.[1]

What the 2007 study found
The odds of contacting a lead drop 100 times when the first call comes at 30 minutes instead of 5.
Between minute 5 and minute 10 alone, the odds of contact fall 5 times.
The odds of qualifying a lead drop 21 times at 30 minutes instead of 5, and 4 times between minute 5 and minute 10.
Measured by the hour, the odds of contact fall more than 10 times in the first hour. The odds of qualifying fall more than 6 times.
After 20 hours, each extra dial made it less likely the lead would be reached and qualified.

The study also found that timing in the week matters, though far less than speed. Wednesday and Thursday were the best days to call. Late afternoon, 4 to 6 pm, was the best time to reach a lead, and 8 to 9 am was the best time to qualify one.[1] Response time mattered far more than either one. For more numbers like these, see our speed-to-lead statistics page.

What do people get wrong about it?

The 21x number is one of the most repeated statistics in sales, and it is usually repeated wrong. Three mistakes show up again and again.

1
Crediting it to Harvard

The 100x and 21x numbers appear in the 2007 InsideSales.com and MIT study. The Harvard Business Review article came four years later.

2
Saying "21 times more likely to close"

The study measured contact and qualification. In its own words, it did not address close ratios.

3
Treating odds as a promise

The numbers describe how the odds fell across six companies. They are a direction and a size, not a forecast for your shop.

None of this weakens the point. It makes it sharper. The honest version of the claim is that a lead you call in five minutes is far easier to reach and qualify than one you call in thirty, and most of that gap opens in the first ten minutes.

Does a 2007 study still hold in 2026?

The data is old, and buyers have changed. People screen unknown numbers more than they did in 2007. But the reason speed works has not changed. A person who just filled out a form is still holding the phone, still thinking about the problem, and has not yet talked to anyone else.

The study includes a small mortgage test that still rings true. The president of InsideSales.com asked a large lead provider for mortgage quotes. He got seven calls. The first came 30 minutes later. The last came three days later.[1] Shared leads still work this way: several lenders get the same borrower, and the first real conversation sets the bar for everyone after.

The speed to lead is really the name of the game. So I need to have the system built out before I even pay for one lead.

Owner of an independent mortgage brokerage, on a SayVo discovery call

The same logic applies outside mortgage. A homeowner asking for an HVAC estimate, a driver asking for an insurance quote and an injured person looking for a lawyer all send the same form to more than one company. The first one to call gets the conversation.

Why do most teams still miss the five-minute window?

Owners already know speed matters. In our sales conversations, 35% of mortgage brokerage owners and branch leaders name speed to lead as a top problem, and 43% say their loan officers don't follow up on leads the way they should.[3] The gap is not knowledge. It is who is free to dial.

Your best loan officers are on the phone with borrowers, at closings or chasing documents. Leads also arrive when nobody is at a desk, which is why so many are lost at night. And dialing strangers is the job loan officers like least.

Any skilled loan officer, they don't want to spend their time dialing out, getting hung up on… sooner or later they die out.

Mortgage branch leader, on a SayVo sales call

A lead that waits is not free, either. You paid for it, and our breakdown of what unworked leads cost shows how fast that adds up.

How do you call every lead in under five minutes?

Three things have to be true at once, and each one fails on its own in most shops.

What a five-minute response takes
The lead reaches the caller in seconds. That means a webhook from your CRM or lead source, not a sync that runs every 15 minutes.
Someone, or something, is free to dial right now. Not after the current call, not after lunch.
The first call is a real conversation that can qualify and book, not a voicemail drop.
Leads that don't pick up get called again on a schedule, inside the calling hours you set.
Every attempt and outcome is written to the lead record, so nobody calls the same borrower twice by accident.

This is the job an AI voice agent was built for. SayVo calls every new lead within 60 seconds, which is inside the first block the study measured. It asks your qualifying questions and books the appointment on the right loan officer's calendar. The trigger and the write-back run inside your CRM, so your team sees every call where they already work.

Speed is only worth having if the call itself is good. Before you trust any system with your leads, read how to evaluate an AI voice agent, and decide whether you want a platform to configure or a done-for-you build.

Where this breaks

The research has limits, and so does speed. The 2007 study covered six companies and phone calls to web leads. It measured contact and qualification, not funded loans. Its authors could not explain why dials after 20 hours hurt, and said so.

Speed also does not fix a bad lead, a weak script or a missing follow-up plan. A fast call to someone who never asked to be called is a problem, not a win. Consent and calling-hour rules are your brokerage's responsibility, and our TCPA guide for AI calling explains what to check. And for referral or past-client business, a thoughtful call from the loan officer still beats a fast one.

Hear a real call, then decide

Get a live call from a SayVo agent on your own phone. Then book a demo and we'll show you how it would call your leads.

Book a demo

SayVo is the premium, done-for-you AI voice agent for independent mortgage brokerages. Built on 200,000+ real mortgage calls, it calls every new lead within 60 seconds, works inside your CRM and goes live in 2 to 4 weeks.

Frequently Asked Questions

Is the 21x speed-to-lead statistic from Harvard?

It traces to the 2007 Lead Response Management Study by Dr. James Oldroyd and InsideSales.com, often called the MIT study. The Harvard Business Review article by the same lead author came out in 2011. The two are often blended together online.

What does the 21x number actually mean?

In the 2007 study, the odds of qualifying a web lead dropped 21 times when the first call came at 30 minutes instead of 5. Qualifying meant the lead agreed to enter the sales process. The study did not measure closed deals.

How fast should you call a new lead?

As close to instantly as you can. In the 2007 study, the odds of reaching a lead fell 5 times between minute 5 and minute 10, and 100 times by minute 30. SayVo calls every new lead within 60 seconds.

Is there a best day or time to call leads?

The 2007 study found Wednesday and Thursday were the best days, 4 to 6 pm was the best time to reach a lead, and 8 to 9 am was the best time to qualify one. Response time mattered far more than either.

Can a loan officer team hit a five-minute response without AI?

During a quiet hour, yes. All day, every day, it is very hard, because the people who would make the call are busy with borrowers or off the clock. That is why owners use an AI voice agent for the first call and keep loan officers for the conversations that follow.

Sources & References

Related Articles

How to Evaluate an AI Voice Agent: The SayVo Standard
Guide

How to Evaluate an AI Voice Agent: The SayVo Standard

AI Voice Agent for Mortgage Brokers: Enterprise or Done-for-You?
Guide

AI Voice Agent for Mortgage Brokers: Enterprise or Done-for-You?

AI Voice Agents That Work Inside Your Mortgage CRM
Guide

AI Voice Agents That Work Inside Your Mortgage CRM

Ready to convert more leads?

Join mortgage teams already using SayVo to respond to every lead in seconds.

Book a Demo