Lead response time: do 5 minutes give 21x better odds?

The line “responding to a lead within 5 minutes makes you 21 times more likely to qualify it” has a traceable origin: a study presented in October 2007 by the then CEO of InsideSales.com and researcher James Oldroyd at a MarketingSherpa B2B marketing event, with three years of data from six companies, more than 15,000 leads and more than 100,000 call attempts (Elkington and Oldroyd, 2007). The 21x compares the odds of qualifying a web-form lead when the call happens at 5 minutes versus 30 minutes, uses data from the vendor’s own system and, in the document’s words, left close rates out. I found no peer-reviewed version. The article published in Harvard Business Review in 2011 confirms the direction of the finding on a scale of hours and shows something more useful for leadership: companies lose the lead in the process, when the CRM releases the queue once a day and distribution follows territory and rotation. Use the 21x as a warning about the direction of the effect and solve speed with a response SLA, routing and automation.

In diagnostic meetings, the 21x shows up on the slide of whoever sells dialers or AI agents, credited sometimes to MIT, sometimes to Harvard. When I open the client’s CRM, the median time between form conversion and first attempt usually sits in hours, sometimes in days. I went to the original document to separate what it supports from what repetition added.

Where does “21 times more likely” in 5 minutes come from?

The number was born in a 35-page report dated November 2007, still hosted on MarketingSherpa’s server. It summarizes a questionnaire survey Oldroyd ran while finishing his PhD at Kellogg, at Northwestern, and a study of call data carried out when he was a Faculty Fellow at the MIT Sloan School of Management. The “MIT” label comes from that affiliation. MIT did not publish the study, and the page that announced the results, now offline, survives in an archived 2009 copy.

What the study measured

The question was when to call a web-generated lead for the best contact and qualification rates, using records from the InsideSales.com dialing system. The definitions weigh on the reading. A contact was an answered call that lasted a minimum time, from 2 to 6 minutes depending on the company. Qualification was the moment the lead agreed to enter the sales process, in some cases by booking a meeting, and each company used its own definition.

In the analysis by hour, the odds of contact fell more than 10 times and the odds of qualification more than 6 times over the first hour. In the analysis by 5-minute intervals, the odds of contact at 5 minutes were 100 times higher than at 30, and the odds of qualification 21 times higher. Between 5 and 10 minutes, the odds of qualifying fell 4 times. The study also reports that, after 20 hours, each additional dial reduced the odds of a contact that leads to qualification, a useful data point for the exit rule of a prospecting cadence.

What the document does not show

The report does not present the statistical model, confidence intervals, number of leads per time band or the industries of the six companies. It states that the InsideSales.com customer base was concentrated in mortgage and insurance, without saying whether the six companies belonged to those markets. According to the text, Oldroyd himself said the patterns only became clear when the companies’ data were combined and that they varied a great deal between them. The company section also says the study changed the positioning of InsideSales.com, which sold a patented web-form callback technology. The study should travel with the label of vendor research with academic participation.

The report talks about odds, the ratio between the probability that something happens and the probability that it does not, which stays close to probability when qualification rates are low. The distortions that matter are in the table below, and there is a causality problem: nobody randomized which leads would get a call at 5 minutes, and reps may have called the most promising ones first, such as price quote requests, which the report itself links to higher qualification. The authors explain the effect through presence: someone who has just filled out the form is still near the phone. The fast call buys the chance of a conversation, and qualification depends on what the rep does with it.

Version in circulation	What the original document says	Difference
“MIT study”	InsideSales.com data analyzed by an MIT Faculty Fellow, presented at a MarketingSherpa event	Vendor research with academic participation
“21 times more likely to close”	Odds of qualifying 21 times higher at 5 than at 30 minutes	The study left closing out
“Harvard research”	The 2011 HBR article uses another dataset, windows of hours and does not mention 5 minutes	Studies mixed up
Applies to any company	Six companies, undisclosed industries, patterns that vary between them	Generalization without basis

What the 2011 HBR study added

In 2011, Oldroyd published “The Short Life of Online Sales Leads” with Kristina McElheran and David Elkington (Oldroyd, McElheran and Elkington, 2011), drawing on two datasets. In the first, the authors sent a web test lead to 2,241 US companies and measured response time: 37% responded within an hour, 16% within one to 24 hours, 24% after more than 24 hours and 23% never responded. Among those that responded within 30 days, the average was 42 hours.

The second dataset covers 1.25 million leads from 42 US companies, 29 selling to consumers and 13 to businesses, and supports the comparison between attempting contact in the first hour and later, which I use in the article on lead scoring to justify the deadline for direct requests. For B2B sellers the composition matters, because most of the dataset sells to individuals. The article works with windows of hours and does not mention 5 minutes or the 21x.

The most useful passage for leadership deals with the causes of delay. The authors point to retrieving leads from the CRM database once a day instead of as they arrive, sales forces focused on the leads they generate themselves and distributing leads among reps and partners by geography and by “fairness.” These are process and configuration decisions that leadership controls.

What other evidence exists on B2B response time?

Peer-reviewed research on response speed to B2B leads is scarce, and I found no experiment that randomized response time to measure the effect on sales. What exists sheds light on what happens to the lead after it reaches the rep.

The 2007 survey of 495 companies

The same 2007 report includes Oldroyd’s questionnaire survey, with 495 responses from companies in more than 40 industries, most of them in North America. Each additional band of delay to first contact was associated with a 2.31% drop in the qualification rate, and companies that could not report that time had a larger drop, 7.29%. Sending the lead to the first available rep and distributing by skill were associated with more qualification, and round robin showed no relationship. These are correlations on self-reported data.

The sales lead black hole

The study by Sabnis, Chatterjee, Grewal and Lilien in the Journal of Marketing analyzed responses from 461 reps at four B2B companies, in scientific instruments, chemicals, copiers and computers, on how they split their time between marketing leads, their own leads and other activities (Sabnis et al., 2013). The article opens by stating that reps never pursue about 70% of marketing leads, a number the authors cite from earlier publications without measuring it.

The measured results matter to whoever designs the SLA. The better a rep rated the company’s prequalification, the more time they devoted to marketing leads. Managerial tracking of follow-up, which the authors expected to have a positive effect, was associated with less time on those leads, and reps with better performance in the previous year spent less time on them and more on their own. A stopwatch and pressure without improving the queue tend to push away the best reps.

More recent vendor research

In 2021, InsideSales released an analysis of 5.7 million inbound leads at more than 400 customer companies (InsideSales, 2021). According to the company, 57.1% of first call attempts happened after more than a week, only 0.1% of leads were engaged in under 5 minutes and conversion was more than 8 times higher in the first 5 minutes than between 5 minutes and 24 hours. The page does not define conversion or describe the method. The multiplier dropped from 21 to 8 within the same family of studies, and the exact number depends on the dataset and the definition.

Summary of the research cited

Source	Year	Sample and method	Main finding
Elkington and Oldroyd, report for MarketingSherpa (call study)	2007	Dialing records from six companies over three years, more than 15,000 leads and 100,000 attempts; vendor data	Odds of qualifying 21 times higher at 5 than at 30 minutes; closing outside the study
Same report (Kellogg questionnaire survey)	2007	495 companies in more than 40 industries, self-reported, regression	Each band of delay associated with less qualification; those who do not measure drop more
Oldroyd, McElheran and Elkington, HBR	2011	Test lead sent to 2,241 US companies; 1.25 million leads from 42 companies, 13 of them B2B	23% never responded; 42-hour average; causes in distribution
Sabnis, Chatterjee, Grewal and Lilien, Journal of Marketing	2013	Survey of 461 reps at four B2B companies	Better prequalification increases follow-up; managerial tracking reduces it
InsideSales, vendor research	2021	5.7 million inbound leads at more than 400 customers; method not disclosed	57.1% of first calls after a week; 0.1% in under 5 minutes

What these data do not tell you

None of the studies lets you say how much your revenue grows if the team goes from two hours to five minutes. The 21x measures qualification, with a different definition at each company, and HBR measures a meaningful conversation with a decision maker. None followed leads through to the sale or randomized response time, and part of the effect may come from reps’ choices, since they call the most promising leads first.

The datasets are old, American and phone-based. In 2007 and 2011, the lead filled out a form and waited for a call. In a Brazilian operation in 2026, a large share of leads arrives through WhatsApp, chat or direct booking, and I found no study with a public method on the decay curve in those channels. In my experience, a WhatsApp lead expects a faster answer than a form lead, but that is field observation.

For B2B, the samples matter: the 2007 study does not disclose industries and the HBR lead dataset is mostly B2C. A demo request with a buying committee and a months-long cycle behaves differently from a car insurance quote, and the size of the multiplier does not transfer to your base.

How do you turn speed into a response SLA?

The SLA between marketing, sales and customer success already sets the deadline for the first attempt and the number of attempts. The studies add two precautions. The first is to measure separately the time to first attempt, which depends on the process, and the time to first actual contact, which also depends on the lead. The second is to distinguish the type of inbound request: the lead who asked for a demo declared a project, and the curve of hours weighs on that lead. The table is the starting point I use with leadership teams, adjusted for team size.

Type of inbound	First attempt target	Ceiling	How the CRM measures it
Demo or quote request, during business hours	Minutes	One hour	Conversion versus first logged activity
Conversation started on chat or WhatsApp	Immediate reply, human or automated	One hour to talk to a rep	First message versus rep’s reply
Direct request after hours	Automated reply with booking link	First hour of the next business day	Attempts by the end of the first business hour
Content, webinar or event, with ICP fit	Same business day, through the SDR queue	Two business days	Time between MQL and first activity
Content without fit or recent interest	Automated nurturing	No call deadline	Entry into the nurturing flow

Measure by the median and the 90th percentile, because the average hides the lead that waited three days. These numbers go into the RevOps metrics matrix, next to conversion by source, so leadership can see whether meeting the deadline changed qualification.

CRM routing: where the lead loses the hours

The causes HBR described are still in the CRMs I audit. Daily retrieval became the spreadsheet the coordinator distributes in the morning, and territory-based distribution became the rule that sends a lead from Minas Gerais to the only rep in the region, who is on vacation. Routing that sustains a short deadline assigns the lead on conversion, without passing through anyone’s hands, considers availability and skill, consistent with what the 2007 survey associated with more qualification, and includes reassignment: once the ceiling passes without an attempt, the CRM returns the lead to a shared queue or to the next available rep and alerts the manager.

The study by Sabnis and colleagues sets a limit on pressure. A response-time dashboard with weekly pressure, without improving what arrives, tends to reduce the time the best reps devote to marketing leads. The deadline holds when the queue deserves it, which is why the routing rule goes hand in hand with the lead qualification criteria.

What automation solves and what it does not

Automation removes the dead time between conversion and assignment, where most of the hours are lost. A notification on the rep’s phone, a task with a deadline in the CRM, an automated reply with a booking link and an alert to the manager when the ceiling expires solve most of the delay, including for the lead that arrives at 7 p.m. on a Friday. An AI agent that replies on WhatsApp in seconds catches the lead at the moment of presence, asks the first qualification questions and books the conversation. The 2007 and 2011 studies, however, measured a call from a person, and I found no research with a public method comparing the two responses. Treat the agent as a first step, with a handoff rule to the rep, and measure the time to that conversation separately, so you do not confuse an instant reply with qualification. At the other end, the 20-hour finding suggests switching channels after the first attempts and closing the lead with a logged reason.

What leadership does with these numbers

1. Set the baseline. Median and 90th percentile of time to first attempt, by source and by time of arrival, and the share of leads with no attempt over the last three months.
2. Map the lead’s path. Draw the route from the form to the rep’s phone: who distributes, with what rule, at what time and with how many manual steps.
3. Separate the types of inbound. Define with marketing and sales which requests get a deadline in minutes and which go to the SDR queue or to nurturing.
4. Configure automatic assignment and reassignment. Assignment on conversion, by availability and skill, with a ceiling, a shared queue and a manager alert.
5. Write the deadline into the SLA and take it to the monthly review. Target, ceiling, after-hours rule, median and 90th percentile on the leadership dashboard.
6. Compare qualification after a quarter. Check whether qualification changed by time band and adjust the deadlines with your own data instead of the 21x.

Response time and L&D: what HR needs to assess

Much of speed is configuration, and the rest is behavior, such as the SDR who leaves the form for after the prospecting block or the rep who does not log the attempt made from a personal phone. Because the study by Sabnis and colleagues linked follow-up to the quality of prequalification, the program needs to include marketing leadership, which answers for the queue. To measure, record time to first attempt and the share of leads with no attempt before the program, and compare after one or two cycles of the monthly review.

If your company invests in demand generation and the CRM shows leads waiting hours for the first call, that is the work of the in-company Revenue Operations training. I run the program with the marketing, sales and CS leaders and the operations person, on your operation’s data, and the group leaves with the SLA between marketing, sales and CS in writing, with deadlines by type of inbound and routing rules, and with the RevOps metrics matrix that tracks response time in the CRM you already use.

Frequently asked questions

Does responding to a lead within 5 minutes make a sale 21 times more likely?

The number comes from a 2007 study with data from six companies, which compared the odds of qualifying a lead with a call at 5 and at 30 minutes. The study left closing out, used data from the tool’s vendor and has no peer-reviewed version I could locate. It indicates that speed matters, without allowing you to predict the effect on your company’s sales.

Is the 5-minute study from MIT or Harvard?

The study was presented by the CEO of InsideSales.com and James Oldroyd, then a Faculty Fellow at MIT, at a MarketingSherpa event in 2007. The 2011 Harvard Business Review article lists Oldroyd among its authors, uses other datasets and does not mention 5 minutes or the 21x.

What is the ideal response time for a B2B lead?

For demo or quote requests, I work with a target of minutes and a ceiling of one hour during business hours, with an automated reply and booking link after hours. Content leads go to the SDR queue on the same business day or to nurturing. The right deadline for your operation comes from comparing response time and qualification in your CRM.

How do you measure lead response time in the CRM?

Record the conversion date, the assignment date and the first activity for each lead, and calculate the median and 90th percentile by source and by time of arrival. Also track the share of leads with no attempt at all. If reps work outside the CRM, the report measures the delay in logging.

Does an AI agent solve response time?

The agent solves the first reply and after-hours coverage, and it can ask the first qualification questions. The studies cited measured human calls, and there is no research with a public method comparing the two. Measure the time to the conversation with the rep separately and define when the agent hands the lead to a person.

How do you prepare the team to meet the response deadline?

Train marketing, presales and sales together, with the company’s own leads and funnel, because the deadline depends on queue quality and routing as much as on SDR effort. The team learns to log every attempt in the CRM and leaders learn to read the dashboard in the monthly review. Measure time to first attempt before and after the program.