What Is Speed to Lead? Why Calling Back in 5 Minutes Wins Sales
Speed to lead is how long it takes you to respond after someone enquires — the clock starts when the form is submitted and stops when a human (or your AI) actually reaches them. The famous finding is that calling within 5 minutes instead of 30 makes you about 21 times more likely to qualify a lead and about 100 times more likely to reach them at all. Two things you should know before you repeat those numbers: they come from a study published in 2007, and they measured contact and qualification, not sales. The sales case is real but rests on different evidence, and I’ll show you both.
Nearly every page on this topic quotes the same two statistics, usually calls them “the MIT study” or “Harvard research,” and presents them as current. I spent a while tracing them back. One is from 2007 and involved about six companies. The other is from 2011. Neither one recorded a single sale.
That is not a reason to ignore speed to lead. The direction of the finding holds up across every dataset I could find, including recent ones, and the practical advice barely changes. But you deserve to know which numbers are solid, which are shaky, and which two of the most-repeated figures in sales have no traceable source at all.
Table of Contents
What is speed to lead?
Speed to lead — also called lead response time — is the time between a prospect raising their hand and you responding. It’s measured per lead and usually reported as a median, because a handful of leads you forgot about for a fortnight will wreck an average.
The calculation is simple:
Speed to lead = time of first meaningful response − time of enquiry
Three details decide whether your number means anything:
| Detail | Get it wrong and… |
|---|---|
| First meaningful response | An instant autoresponder saying “we got your message” isn’t a response. If it can’t answer a question or book a time, the clock is still running |
| Use the median, not the average | One lead you answered after three weeks will drag an otherwise good average into nonsense |
| Measure in the lead’s business hours, or don’t | Pick one and be consistent. A 14-hour “response time” that’s really overnight tells you nothing useful |
Most businesses measuring their lead response time for the first time find the number is far worse than they assumed — not because they’re slow during the day, but because evenings, weekends and lunch breaks quietly add hours to every lead that arrives then.
Picture someone standing in your doorway
A lead is a person standing in the doorway of your shop.
Greet them in five seconds and you’re having a conversation. Look up after a couple of minutes and they’re still there, but cooler. Come back forty-two hours later and the doorway is empty — they walked down the street and bought from whoever said hello.
Nothing about the person changed in those forty-two hours. They still wanted the thing. You just weren’t there when they did.
That’s the whole idea. Everything below is about how much that actually costs and what the evidence really shows.
The 5-minute rule, and what the study actually measured
Here’s the real provenance of the number everyone quotes.
The 2007 Lead Response Management Study, by Dr James Oldroyd with InsideSales.com:
| Sample | About six companies, 15,000+ web-generated leads, 100,000+ call attempts, over roughly three years |
| Finding | Calling at 5 minutes rather than 30 was associated with about 100× higher odds of making contact and about 21× higher odds of qualifying the lead |
| What it did not measure | Sales. Revenue. Conversions. Not one |
| Type | Observational, not a controlled trial |
| Funding | Conducted with a sales software vendor, using its customers’ data |
Two labels get attached to this study and both are wrong. It is not a Harvard study — people confuse it with a later Harvard Business Review article. And it’s only “the MIT study” because Oldroyd was affiliated with MIT Sloan at the time; MIT did not run it as an institution.
The 2011 Harvard Business Review article, “The Short Life of Online Sales Leads” (Oldroyd, McElheran and Elkington, March 2011) is the separate piece of research:
| Part | What they did | What they found |
|---|---|---|
| 1 | Sent test leads to 2,241 US companies | 37% replied within an hour · 23% never replied at all · average reply among those who replied within 30 days was 42 hours |
| 2 | Analysed 1.25 million leads from 42 companies | Contacting within an hour ≈ 7× higher odds of qualifying than waiting one more hour, and 60×+ higher than waiting a day or more |
Again: no sales were measured. Both studies tracked whether you reached the person and whether they turned out to be worth pursuing.
Why “5 minutes wins sales” is still true
So if the famous studies didn’t measure sales, does the headline hold up? Yes — but you have to follow the chain and use the right evidence for each link.
| Link in the chain | Evidence | Strength |
|---|---|---|
| Fast response → you actually reach them | 2007 study, ~100× at 5 vs 30 min | Dated, tiny company sample, right direction |
| Reaching them → more qualified leads | 2007 (21×) and 2011 HBR (7× within the hour) | Two studies, one with 1.25m leads. Reasonably solid |
| More qualified leads → more sales | Not measured by either study | This is the gap |
| Fast response → more closed deals | Velocify 2013 and Optifai 2026 (below) | Vendor data, correlational |
The last row is where the sales claim actually comes from:
Velocify, “The Ultimate Contact Strategy” (2013) — about 3.5 million leads across 400+ companies — found leads called within one minute converted at a rate 391% higher than leads called at two minutes. That’s conversion, not just contact. But it’s vendor data drawn from phone-heavy industries like mortgage, insurance and education, where someone picking up fast genuinely is the product. Don’t carry that multiplier into a slow B2B sale.
A 2026 analysis of 939 B2B companies (Q2 2025 to Q1 2026, by the sales-automation vendor Optifai) reported close rates by response speed:
| Response time | Close rate |
|---|---|
| Under 5 minutes | 32% |
| 5–30 minutes | 24% |
| 30–60 minutes | 18% |
| 1–24 hours | 15% |
| Over 24 hours | 12% |
That’s recent, and it’s about closing. I’d treat the exact percentages carefully, though — the study is self-published by a vendor, doesn’t give sample sizes per tier, and contradicts itself in places (its summary puts 24% in a different bracket than its own table does). Fast-responding companies are also usually better-run companies in every other way, so some of that gap isn’t caused by speed.
Honest summary: the direction is consistent across five datasets spanning nineteen years. The exact multipliers are softer than anyone quoting them admits.
The two statistics you should stop quoting
While tracing sources I found two figures that appear in almost every roundup of speed to lead statistics — probably including pages you’ve read this week — with no traceable primary source.
| Statistic | Usually credited to | Reality |
|---|---|---|
| “78% of customers buy from the first company to respond” | “A Lead Connect survey,” sometimes HBR | No published report, no methodology, no primary source anyone can find |
| “35–50% of sales go to the vendor that responds first” | InsideSales | Equally untraceable |
Both are folklore. They’ve been copied between blogs so many times that they read like established fact, and several pages attribute the 78% figure to Harvard Business Review, which never published it.
Four more things that get misquoted even when the underlying study is real:
| What people say | What the study said |
|---|---|
| “100× more conversions” | 100× odds of making contact |
| “21× more conversions” | 21× odds of qualifying |
| “The average lead waits 42 hours” | 42 hours was the average among companies that replied within 30 days — it excludes the 23% who never replied, so the true picture is worse |
| “23% of leads never get a response” | 23% of companies failed to answer one test lead. Different thing entirely |
I’m pointing this out partly because I got some of it wrong myself. An earlier article on this site credited the 42-hour figure to Drift and described the 7× finding as being about reaching a decision maker. Both were wrong, and I’m fixing them.
How slow is everyone, really?
This is the part that should make you optimistic, because the bar is on the floor.
| Study | Year | What it tested | Headline |
|---|---|---|---|
| HBR | 2011 | 2,241 US companies | 37% replied within an hour · 23% never replied |
| Drift | 2017 | 433 B2B SaaS companies, secret-shopped | Only 7% replied within 5 minutes · 55% never replied within five business days |
| Optifai | 2026 | 939 B2B companies, CRM timestamps | 23% under 5 min · 42% over 24 hours · 47-hour average |
Read the Drift row again. Of 433 companies that had built websites, written content and paid for traffic to generate those leads, seven percent answered within five minutes and more than half never answered at all. Drift also noted that only 14% used live chat — and every one of the ten fastest responders did.
The Optifai and HBR averages are worth comparing: 42 hours in 2011, 47 hours in 2025–26. Fifteen years of CRM software, marketing automation and sales engagement platforms, and the average response time did not improve. Different studies and different samples, so it’s not a clean like-for-like — but nobody can claim this got solved.
Industry averages from that 2026 study, for context:
| Industry | Average response | Within 5 minutes |
|---|---|---|
| B2B SaaS | 38 hours | 28% |
| Professional services | 45 hours | 22% |
| Consulting | 52 hours | 18% |
| Manufacturing | 62 hours | 15% |
What five minutes is actually worth
Let me do the arithmetic nobody does, using the 2026 close rates above. Say you get 100 leads a month and you’re paying $50 a lead — $5,000 a month in acquisition.
| Responding under 5 min | Responding after 24 hours | |
|---|---|---|
| Leads | 100 | 100 |
| Close rate | 32% | 12% |
| Customers | 32 | 12 |
| At a $500 average deal | $16,000 | $6,000 |
| Cost per customer acquired | $156 | $417 |
Same traffic, same ad spend, same website. The difference is who picked up the phone first.
The cost-per-customer line is the one to show your finance person. Slow response doesn’t just lose deals — it nearly triples what each customer costs you to win, because you paid for all 100 leads either way.
Two caveats I’m not going to bury. Those close rates are correlational vendor figures, so this is an illustration of scale, not a forecast — companies answering in five minutes are usually doing six other things right. And if your average deal or lead cost differs, redo the sums with your own numbers. The shape holds; the size depends on you.
Why most businesses can’t do five minutes
If this were easy, that Drift figure wouldn’t be 7%. The reasons are structural, not laziness:
| The real obstacle | Why it bites |
|---|---|
| Leads don’t arrive in business hours | A large share of enquiries to service businesses land outside 9-to-5. Nobody is there to ring back in five minutes |
| Your team is already on calls | A salesperson on a 40-minute call cannot respond to anything that arrives during it |
| Nobody owns the clock | If the lead goes to a shared inbox, the median response time is however long until someone’s curiosity wins |
| The form sends an email, not an alert | Email is checked in batches. Five-minute response needs a push, a ring, or an automation |
| Weekends | A Friday-evening enquiry answered Monday morning is a 62-hour response, and it will look like one bad week in your data |
Staffing your way to genuine five-minute coverage is expensive. Covering 168 hours a week rather than 40 takes roughly four full-time people; at the US median receptionist wage of about $36,920 (Bureau of Labor Statistics, May 2023 data) that’s north of $150,000 a year in salary alone before benefits. For most small businesses that’s not a trade-off, it’s a non-starter.
Which is why this problem got solved with software rather than headcount.
How to actually get to five minutes
Ranked by what they cost and what they’re good for:
| Method | Cost | Covers | Honest limitation |
|---|---|---|---|
| Instant routing + push alerts | Free–cheap | Business hours | Only works if someone’s free to act on the alert |
| Live chat on the page | Low | Whoever’s staffing it | Drift found every one of the 10 fastest responders used it — but it needs a person |
| AI chatbot on your site and channels | Low | 24/7 | Great at qualifying and booking; can’t phone anyone |
| Round-robin with an SLA and escalation | Staff time | Business hours | Needs discipline and someone checking the dashboard |
| Human answering service | ~$1.99–$11.50 per call | 24/7 | The cheapest route to a real voice, but per-call pricing adds up fast |
| AI receptionist / calling agent | Roughly $0.25–$10 per answered call depending on provider | 24/7 | Handles the straightforward 80%; needs a clean handover for the rest |
I’ve broken the phone options down properly elsewhere, with per-call costs for ten providers, in best AI receptionist for small business and AI receptionist vs answering service. If you’re starting from scratch, what is an AI receptionist and what is an AI phone answering service cover how they work.
What I’d actually do, in order:
- Measure first. You cannot fix a number you haven’t got. One week of data is enough to be alarming.
- Fix the alert, not the team. Most slow responses are a notification problem, not a motivation problem.
- Put something on the page that answers at 11pm. A chatbot that qualifies and books beats a form that emails a sleeping inbox.
- Add voice only if your leads expect a call. Home services, legal, medical, real estate, car sales — yes. Most software sales — no, a fast email and a booked slot is better.
- Set one SLA and review it weekly. Five minutes in hours, thirty minutes out of hours, is realistic for most businesses and already puts you in the top tenth.
Speed is only half of it, though. Responding in four minutes to a lead you never should have chased is just being fast at the wrong thing — which is what lead qualification is for, and the two work as a pair.
How to measure your own speed to lead
| Step | What to do |
|---|---|
| 1 | Pick your start event. Form submit, call ring, DM received. Write it down |
| 2 | Pick your stop event. First response that a human could reply to or act on — not the autoresponder |
| 3 | Pull one month from your CRM, or log 50 leads by hand if you have no CRM |
| 4 | Calculate the median, then separately the median in business hours and out of hours |
| 5 | Count what share you answered in under 5 minutes, under 1 hour, over 24 hours |
| 6 | Compare that split against the benchmark table above — not against the 5-minute ideal |
Step 4 is where the insight usually is. Plenty of businesses have a respectable in-hours median and a catastrophic out-of-hours one, and that single split tells you whether you have a staffing problem or an automation problem. They need completely different fixes.
If your leads come through a CRM, wiring that up is mostly a configuration job — I’ve covered the approach in integrating AI chatbots with CRM systems and sales automation.
How I researched this
Every figure on this page is traced to its original study with its original year, because the central problem with this topic is that almost nobody does that.
Verified and used: the 2007 Lead Response Management Study (Oldroyd with InsideSales.com; ~6 companies, 15,000+ leads, 100,000+ call attempts); the March 2011 Harvard Business Review article “The Short Life of Online Sales Leads” (Oldroyd, McElheran, Elkington; 2,241 companies audited plus 1.25 million leads from 42 companies); Drift’s 2017 secret-shopper survey of 433 B2B SaaS companies, published 27 February 2017; Velocify’s 2013 “The Ultimate Contact Strategy” (~3.5 million leads, 400+ companies); and a 2026 study of 939 B2B companies published by Optifai covering Q2 2025 to Q1 2026. The receptionist wage figure is US Bureau of Labor Statistics data for May 2023.
Checked and rejected: “78% of customers buy from the first company to respond” and “35–50% of sales go to the first vendor to respond.” Both are cited constantly and neither has a findable primary source. I have left them out rather than repeat them with a hedge.
Corrections I’m making to my own work: an earlier article on this site attributed the 42-hour figure to Drift rather than to the 2011 HBR study, described the 7× finding as being about reaching a decision maker when it is about qualification, and repeated the unsourced 78% claim while crediting it to HBR. Those are my errors and I’m correcting that page.
What I haven’t done: I haven’t run an experiment, and nor has anyone cited here — every study on this topic is observational, which means some of the measured advantage belongs to the fact that fast companies are better companies, not to speed itself. The two foundational studies are 19 and 15 years old and predate smartphones being universal; the 2007 study involved roughly six companies, which is a very small base for a number quoted as universal law. Three of the five datasets come from vendors selling response-speed software, and I sell software in this space too, which is why the methods table above starts with free fixes and lists the limitations of the paid ones.
Frequently asked questions
How fast should you respond to a lead?
Under 5 minutes is the benchmark the research supports, and under 1 hour is where the measurable advantage still clearly exists — the 2011 HBR data found roughly 7× better qualification odds inside the first hour versus the second. Realistically, 5 minutes in business hours and 30 minutes outside them puts you ahead of about 90% of companies, based on Drift’s finding that only 7% manage 5 minutes at all.
Is the 5-minute rule actually real?
The direction is well supported; the precise multipliers are weaker than usually claimed. The original 2007 study found about 21× better qualification odds at 5 minutes versus 30, but it involved roughly six companies, was observational, was funded by a software vendor, and measured qualification rather than sales. Treat 5 minutes as a well-evidenced target, not a cliff edge at minute six.
Does speed to lead matter for B2B or just consumer sales?
Both, but the mechanism differs. In consumer and phone-heavy industries like mortgage or home services, speed often is the differentiator — the 391% conversion figure comes from exactly those verticals. In considered B2B purchases, speed mostly buys you the chance to shape the requirements before a competitor does. The 2026 B2B data still showed 32% close rates under 5 minutes against 12% over 24 hours.
What counts as a response — does an auto-reply count?
No. A response is something the prospect can act on or reply to: a call, a real message, or a booking link that takes them somewhere. “Thanks, we’ve received your enquiry” doesn’t stop the clock, and counting it is the most common way businesses flatter their own numbers.
How do I respond in 5 minutes outside business hours?
Automation, not staffing. A chatbot that answers questions and books a time covers most of it for very little; an AI phone agent covers people who would rather call. Staffing 24/7 coverage with humans runs past $150,000 a year in wages for a small business, which is why almost nobody does it that way.
Will responding too fast seem desperate?
No evidence suggests it does, and every dataset points the other way. What reads as desperate is the content of the response — a hard pitch thirty seconds after someone downloads a guide. Fast and helpful beats slow and polished; fast and pushy is its own problem.
Should I call or email first?
Call if your audience expects calls — home services, legal, medical, real estate, car sales. Email or chat if they don’t, which includes most software and professional services. The 2007 study measured calls specifically, so its multipliers don’t transfer cleanly to email, and plenty of B2B buyers actively dislike an immediate phone call.
Why do so many articles about this quote the same numbers?
Because there is very little primary research, and almost all of it is old. Nearly every page traces back to the same 2007 and 2011 studies, usually without dating them. I could find no newer large independent study — the only recent figures come from vendors analysing their own customers’ CRM data.
The bottom line
Speed to lead is one of the few things in sales where the advice is simple, the bar is low, and almost nobody clears it. Only 7% of companies in one secret-shopper test answered within five minutes, and more than half never answered at all. You don’t need to be excellent here. You need to be present.
But be careful which numbers you repeat. The famous 21× and 100× figures come from a 2007 study of about six companies and describe contact and qualification, not sales. The 78% first-responder statistic that half the internet quotes has no source anyone can find. The sales case rests on different and more recent evidence — a 32% close rate under five minutes against 12% over twenty-four hours — and that evidence comes from vendors, which you should weigh accordingly.
What survives all the caveats: respond faster, measure it honestly, and fix your out-of-hours gap before anything else. Most of your lost deals are sitting in the hours when nobody was there.
Want something answering at 2am? My plan is $14.50 a month and includes an AI chatbot agent that qualifies leads and books meetings across your website, WhatsApp, Instagram and Messenger, with handover to your team when a conversation needs a person. Start a free trial and point it at the hours you’re currently losing.
Also worth reading: lead qualification, because speed without qualification just means you’re quick at chasing the wrong people, and how to automate your sales funnel for what happens after the first response.
