Speed to Lead Benchmarks, the Research Behind the 5-Minute Rule, and Routing Rules for B2B Teams

Speed to lead is the time between a prospect submitting a form and your team's first real contact with them. The research behind the 5-minute rule found the odds of contacting a lead drop 100 times when the call comes at 30 minutes instead of 5. A 2026 summary put the median business-to-business inbound first response at roughly 42 hours. In our reading, that gap is where recoverable demand sits.

What is speed to lead, and what did the research find?

Speed to lead is the time between a prospect submitting a form and your team's first real contact with them. The clock starts at the submission timestamp and stops at the first human (or capable agent) touch. Practitioners also use the phrase more loosely: one 2026 benchmark article separates lead response time from speed to lead, describing the latter as "the whole capture-to-contact motion including routing and assignment." The same article says auto-responders do not count in any serious measurement.

The 2007 study behind the 5-minute rule

The figure most teams quote comes from the InsideSales.com and MIT Lead Response Management study, presented at a MarketingSherpa B2B demand generation 4th Annual Summit 2007. The researchers examined 3 years of data across six companies, more than fifteen thousand web leads and more than one hundred thousand call attempts.

The headline finding: the odds of contacting a lead called within 5 minutes versus 30 minutes drop 100 times, and the odds of qualifying that lead drop 21 times. The study measured contact and qualification only. It did not address close ratios, so treat any closed-won claim attributed to it with care.

What 2026 benchmarks report

A 2026 summary put the median first response at roughly 42 hours for B2B inbound, and found that between 7% and 23% of companies reply inside five minutes, depending on the study. In our reading, the gap between a 5-minute target and a 42-hour median is where most of the recoverable demand capture sits.

Measuring it on your own data

Report a median and a 90th percentile, not an average, because a few untouched leads distort the mean.

speed_to_lead_minutes = first_human_touch_timestamp - form_submit_timestamp

Example:
form submitted  10:04
first dial      10:31
speed to lead = 27 minutes  (outside a 5-minute SLA)

Pair this with stage conversion rates from our business-to-business funnel conversion benchmarks to see what slow response costs downstream.

A different product with the same name

Search results for the phrase also surface real estate businesses. iSpeedToLead, for example, describes itself as a real estate lead marketplace selling motivated seller leads. That is a lead-buying product, not a response-time metric, so filter it out when researching benchmarks.

Where the 5-minute rule and the 21x figure come from

The 5-minute rule and the 21x figure trace back to the 2007 summit presentation covered above. That presentation bundled three pieces of work, and the report says so directly: it set out to "summarize key results from the survey, study, and case study". For each tier of delayed response in that survey, "leads qualified dropped 4.3% and close rates fell nearly 2%."

The same 2007 study measured how much its best calling day and time slot beat the worst one for contact and qualification.

Best calling slot versus worst slot, Lead Response Management Study (2007)

Item	Value
Best day, contact	49.7 % better than worst slot
Best day, qualify	24.9 % better than worst slot
Best day, contact to qualify	19.1 % better than worst slot
Best time, contact (4 to 6pm)	114 % better than worst slot
Best time, qualify	164 % better than worst slot
Best time, contact to qualify (4-5pm)	109 % better than worst slot

Source: InsideSales.com/MIT, Lead Response Management Study, 2007

How the study defined its terms

Before you compare your numbers to the 21x figure, check that your terms match the study's.

• Contact was a call that connects with a live person and lasts a set length. That length was different for each company, ranging from 2 minutes to 6 minutes.
• Qualify was the stage where the lead is willing to enter the sales process. Each company had their own way to indicate a qualified lead.
• The sample covered only leads captured through a web form, and attempted or called at least one time. Leads nobody called fell outside the data.

The hourly view matters as much as the 5-minute view. The odds of contact decrease by over 10 times in the first hour, and the odds of qualifying decrease by over 6 times. The study also found that after 20 hours, every additional dial your salespeople make actually hurts the chance of contact that leads to a qualification.

Map the study's timestamps to your CRM

The right-hand column shows how we would map each one to fields in a customer relationship management (CRM) system.

Set your contact threshold on call duration. Write down your qualification rule before you measure anything. Without these two settings, we think a team's speed to lead number cannot be compared to the study's, because the study let each company define both.

What the 2011 Harvard Business Review article studied

The second piece of research people cite is a magazine article, and it is a separate publication from the study covered in the previous section. The Short Life of Online Sales Leads ran in the Harvard Business Review (HBR) issue for March 2011. Its authors were James B. Oldroyd, Kristina McElheran and David Elkington.

The same record files it as a peer-reviewed article. Published in March 2011, the article is now more than fifteen years old.

The authors' interests matter when you weigh the figures. HBR's author note describes Elkington as the chairman and CEO of InsideSales.com, the company that also worked on the earlier study. The article's opening claim is simple: most companies are not responding nearly fast enough.

Which numbers belong to which work

The company-level figures circulating today usually trace to this article, not to the earlier study. LeanData describes the HBR work as having analyzed response times across 2,241 U.S. companies. On that page, firms that reached leads within an hour were seven times more likely to have a meaningful conversation with a key decision maker. The same page's benchmark table restates the result as 7x less likely to qualify vs. responding within the hour (HBR). In our reading, these are two different outcomes, conversation and qualification, attached to one multiplier. Check which one a deck means before you quote it.

The lead and call counts belong to the earlier study, not to the article. Mixing the two creates a citation that no single source supports.

The 24-hour figure is the third row: companies that waited a day or longer were, per LeanData, 60 times less likely to qualify the lead at all.

Before a benchmark goes into a board slide, run it through a short check:

Speed-to-lead citation check
1. Name the publication: magazine article or research study.
2. Name the unit counted: companies, leads, or dials.
3. Name the outcome: contact, conversation, or qualification.
4. Name the window: minutes, first hour, or first day.
5. Link the original, not the blog that repeated it.
If any line is blank, drop the figure from the slide.

How fast B2B companies respond in practice

The averages covered earlier came from mystery-shopper style audits. The table below lines up the published audits by what each one counted, because the samples and definitions differ enough that the headline figures do not compare directly.

Three audits reported the share of companies that never replied: 23% in 2011, 27% in 2018 and 63.5% in 2024.

Share of companies that never responded to an inbound lead

Item	Value
HBR 2011	23%
Drift 2018	27%
RevenueHero 2024	63.5%

Source: Digital Applied, Speed-to-Lead Benchmarks 2026

Two caveats apply. The Drift figures survive only through secondary aggregators. And InsideSales, Velocify, and Blazeo all sell lead-response software, so their numbers come from interested parties.

The channel split in the Workato test is useful for routing design. Personalized email took 11 hours and 54 minutes, on average, and just 31% of companies responded to leads by phone. Companies using a lead routing tool averaged 3 hours and 32 minutes, against nearly 13 hours for those without one.

RevenueHero heard back from only 365 companies, and those responders averaged one day, five hours, and seventeen minutes.

Running the same audit on your own funnel

Before you quote any row, measure yourself the same way. Submit test leads through every form and channel, and log the first human touch. One common definition excludes auto-responders from the count.

For each test lead:
  response_minutes = first_human_touch_timestamp - form_submit_timestamp

No-response share = leads with no human touch after 7 days / leads submitted
Five-minute share = leads with response_minutes <= 5 / leads submitted
Report: median and 90th percentile of response_minutes (responders only)

Worked example: you submit 20 test leads. 4 get no human touch, and 3 get one within five minutes. In this example, your no-response share is 4 / 20 = 20%. In this example, use the full 20 as the denominator for both figures, so they line up with audits that counted silent companies.

How fast should sales respond to a demo request?

Start from a published ladder, then add owners and timers. LeanData's tiered framework puts a demo request or pricing inquiry in its top tier and a trade show contact at within 2 to 3 business days. Its organizing rule is to match response speed to the temperature of the signal. A ladder on its own leaves two questions open: who picks the lead up when the owner does not, and when a manager hears about it.

Writing the rule down has value even before you tune it. Blazeo's 2026 report, as summarized by LeanData, found that companies with a defined SLA respond within 15 minutes at 54.9% versus 29.5% for those without one.

The cutoffs, backup owners and escalation times below are our view. Adjust them to your coverage hours.

Here is how the demo row runs. A form arrives at 2:00 p.m. and routes to the assigned rep. With no logged touch by 2:05, routing moves the lead to the on-call rep and posts a reassignment notice. In this example, with still no touch by 2:10, the manager is alerted and the lead is flagged in the weekly SLA report. The same timers in rule form:

RULE demo_request, pricing_inquiry
  clock_start  = form_submitted_at
  target       = 5 min
  at 5 min     -> reassign to backup_owner, notify original owner
  at 10 min    -> alert manager, flag record SLA_BREACH

RULE content_download WHERE account_fit = high
  target       = 60 min
  at 120 min   -> alert manager, flag SLA_BREACH

RULE event_lead
  target       = 1 business day
  at 2 business days -> alert manager, flag SLA_BREACH

Start the clock at submission, not at assignment. LeanData defines lead response time as lead processing time plus representative response time, so slow routing counts against the SLA. Report each tier as a median and a 90th percentile, the convention Perspective AI describes. As covered earlier, auto-replies do not count as a touch.

Some teams push the top tier lower. LeanData notes top teams targeting sub-60-second response times on demo and pricing requests. Perspective AI's modeled scenario put a 45-second SLA at about 9% more opportunities than a 5-minute one, at the cost of 24/7 coverage. In our reading, a 5-minute target with a working backup is the first rung to hold before chasing seconds. For the full handoff document, use the marketing and sales SLA template and the service level agreement definition.

How should inbound leads be routed?

A response target only holds if the lead lands with a rep who can act on it. Without automatic assignment, LeanData notes that 30 minutes or more may have passed before the lead reaches the right person. That alone breaks a five-minute target before any rep is involved.

HubSpot's Rotate record to owner workflow action is a common example of how a CRM handles assignment. For lead and ticket records, it offers three distribution options:

Who is eligible

Eligibility rules can quietly remove reps from the pool. In HubSpot, users must be active with an assigned Sales or Service Hub Professional or Enterprise seat. For team-based assignment, extra team members will not be included. Deactivated users, or users who lose their paid seat, show with a red border and are skipped. After any rep leaves or changes seats, audit every routing rule.

Out-of-hours coverage and fallback

The setting that assigns to available users only skips reps marked Away. If everyone in the pool is away, the owner field is set to unassigned or (No owner). An unowned demo request at 7 p.m. waits until someone notices it. HubSpot advises adding a user for the fallback option so assignment failures still reach a person.

Two further limits matter under load. Only one record passes through the rotate action at a time while others retry, which can result in objects taking longer. If the owner property syncs to Salesforce, Salesforce may try to update HubSpot with the previous owner.

A routing rule template

Write each rule down before you build it. This template follows our own method:

RULE: Demo request, business hours
Trigger: form = demo request AND fit score >= threshold
Pool: named reps with paid seats, status = available
Method: load balanced
Target: first touch within the tier window
Fallback 1: on-call rep (rotates weekly)
Fallback 2: team lead
Out of hours: on-call rep; auto-confirm to buyer with booking link
Alert: owner = unassigned for 2 minutes -> notify sales manager
Audit: monthly check for deactivated or seatless users in pool

Worked example: a pool of four reps has one on leave and one marked Away. Only two reps are eligible. Load balancing then splits leads between them, so their targets need checking that week. Pair the rule with your lead scoring model so fit scores decide which tier a lead enters.

How to measure lead response time

Two timestamps define the metric. Most arguments about a team's speed to lead turn out to be arguments about which two timestamps someone used.

Start the clock at submission. Take the time from the form, chat or scheduling tool at the moment the buyer submits, not from the CRM record's creation time. Routing, matching and enrichment happen between those two moments, and that delay belongs in the number.

Stop the clock at the first genuine reply. Perspective AI defines this as the first substantive reply from a human or an AI agent capable of advancing the conversation. A connected call, a personal email that answers the request, or an agent that qualifies and books the meeting all count. A confirmation email does not, as covered earlier. Log unanswered dials as attempts in a separate field so they do not stop the clock.

Write both timestamps to the lead record as fields. The calculation then runs inside the CRM and nobody rebuilds it in a spreadsheet each week.

Report the median and the 5-minute share, not the mean

Response times have a long tail, so a few leads answered the next day drag the mean far from what a typical buyer experienced. The 2026 figures Perspective AI collected show this: one benchmark put the median first response at about 42 hours, while a second sample's average of 47 hours was skewed by long tails.

Here is a worked example with our own numbers.

The median and the 5-minute share show a team that is mostly fast, with one lead that went to the next day. Keep never-answered leads in a separate count. They should not quietly fall out of the median.

Weekly report layout

LeanData advises teams to calculate current average lead response time across lead types and rep groups. We break the report out further, and we use medians:

WEEKLY LEAD RESPONSE REPORT   Week of: ____

Cut               | Value          | Leads | Median (min) | % within 5 min | Never answered
------------------|----------------|-------|--------------|----------------|---------------
Source            | Demo form      |       |              |                |
Source            | Pricing page   |       |              |                |
Source            | Chat           |       |              |                |
Lead type         | Demo request   |       |              |                |
Lead type         | Content        |       |              |                |
Owner             | Rep A          |       |              |                |
Owner             | Rep B          |       |              |                |
Hour submitted    | 00:00-07:59    |       |              |                |
Hour submitted    | 08:00-12:59    |       |              |                |
Hour submitted    | 13:00-17:59    |       |              |                |
Hour submitted    | 18:00-23:59    |       |              |                |

Clock: submission timestamp to first genuine reply (auto-responses excluded)

Read the hour-of-submission rows first. They show coverage gaps that a single team-wide median hides. To track these rows next to pipeline metrics, add them to the demand gen dashboard template.

A worked example and the steps to set your own target

The steps below are our method. Every number in the example is illustrative.

The baseline arithmetic

Next, split the requests by arrival time. The fix for each group is different.

Then check capacity. The problem is the queue, not headcount.

Picking a tier

Routing and review

Route business-hours demo requests to a single on-point rep per two-hour block, with a named fallback. Queue off-hours requests for the first block of the morning, ahead of new arrivals. Write both rules into your marketing and sales SLA.

Review weekly with one short sheet:

Week of: ________
Demo requests received:            ___
  Business hours:                  ___
  Off hours:                       ___
Median first response, in hours:   ___
90th percentile, in hours:         ___
Requests outside target:           ___
  Top reason (no owner / rep away / off hours / other): ___
Change agreed for next week:       ________
Owner of the change:               ________

Five steps to set your own target

1. Pull a recent window of demo requests, for example 90 days, and compute volume per working day and per hour, split by business hours and off hours.
2. Compute the median and 90th percentile for each split, so you see which group drives the delay.
3. Check capacity: requests per hour times minutes per reply, against the minutes one rep has.
4. Choose a target per tier, set on-point blocks and a fallback owner, and write them into the SLA.
5. Run the weekly sheet for a quarter, then tighten the target if your 90th percentile already meets it. Feed the new response rates into the reverse funnel calculator to see the pipeline effect.

What this page does not cover

This page does not rank or recommend routing tools or lead-response software, and it leaves out lead-buying products such as real estate lead marketplaces that share the name. It also does not establish how response time affects close rates, since the 2007 study measured contact and qualification only.