How fast should you call back? What the speed-to-lead studies measured

Where the five-minute rule comes from

The figures we traced lead back to two studies with the same two names on them, James Oldroyd and David Elkington.

• 15,000+web-form leads from six unnamed companies, behind the 100-times figure

  Source: InsideSales.com, Lead Response Management Study, dated October 2007. Vendor study, six companies, web-form leads.

• 1.25 milliononline leads at 42 US companies, behind HBR’s “nearly seven times”

  Source: Oldroyd, McElheran and Elkington, “The Short Life of Online Sales Leads”, Harvard Business Review, March 2011. Business press; one author was the chairman and CEO of InsideSales.com.

A 2007 study by a software company. InsideSales.com sold dialing and lead software. Its CEO, Elkington, presented the study at a MarketingSherpa summit in 2007 with Oldroyd, then a visiting research fellow at MIT, and the report is dated October 2007.

The data came from InsideSales.com’s own system: six unnamed companies that answer web leads, three years, more than 15,000 leads and more than 100,000 call attempts. This is where the line comes from: “The odds of contacting a lead if called in 5 minutes versus 30 minutes drop 100 times.” The next sentence gives 21 times for the odds of qualifying a lead.

A 2011 article in Harvard Business Review. Oldroyd, Kristina McElheran and Elkington wrote “The Short Life of Online Sales Leads”, and HBR lists Elkington as chairman and CEO of InsideSales.com. The article reports two things:

• An audit of 2,241 US companies that each got a test lead from a web form. Of those, 37 percent replied within an hour and 23 percent never replied.
• A study of 1.25 million leads at 42 US companies. Firms that tried to reach a lead within an hour were “nearly seven times as likely to qualify the lead” as firms that tried even an hour later. To qualify meant “having a meaningful conversation with a key decision maker.” The article does not say whose data this was.

Source: Oldroyd, McElheran and Elkington, “The Short Life of Online Sales Leads”, Harvard Business Review, March 2011. Business press; one author was the chairman and CEO of InsideSales.com.

What they counted

Both studies timed replies to leads that came in online. HBR names financial services, automobiles, education, software, health care and professional services among the industries that had turned to the internet for sales leads. The 2007 report says InsideSales.com’s customers were concentrated in mortgage and insurance. Neither study mentions a trade shop.

Success meant a conversation. In the 2007 study, a “contact” was a call that reached a live person and lasted a set time, 2 to 6 minutes depending on the company.

A lead counted as qualified once it was willing to enter the sales process, and each company marked that its own way. The report is plain about the rest: “This study did not address close ratios.”

The 100-times figure comes from pooled data. The report says Oldroyd found its clear patterns “only when data from several companies is combined together,” and that “Patterns vary significantly from company to company.”

Its own case study shows how far. When FranklinCovey called back people who had dropped out of a purchase, “the worst time to call back was within the first 24 hours.”

The report prints no contact rate for any five-minute slot either. Its charts are bars on a scale with no unit, so a reader cannot tell whether 100 times means going from most leads to a few, or from a few to almost none.

The report’s first guess at why speed matters is that someone who just sent a web form is “at their computer desk, probably right near their phone.”

A customer whose call rang out had the phone in hand a minute ago, so the guess fits them at least as well. No study we found measured how long that customer waits before calling the next shop.

What the trade data measures

We found no study that timed trade shops’ call-backs against the jobs they booked. The trade figures we did find come from software vendors, and they measure other things:

Source: ServiceTitan, call booking rates by trade, shop size and time of day, October 25, 2022. Vendor data, more than 3,000 businesses in the US and Canada.

• A typical trade shop booked 42 percent of its calls in June 2022. In the same data, shops with fewer than 5 technicians booked 24 percent. Mornings booked best. At small shops the rate ran from 26 percent at the peak to 9 percent after 6 p.m., and the article does not say why (ServiceTitan, 2022, vendor data, more than 3,000 businesses in the US and Canada).
• At residential AC shops, 14.1 percent of calls in June 2025 came in outside Monday to Friday, 8 a.m. to 5 p.m., against 9.8 percent in October. The summer rise was largest between 5 and 9 p.m. (ServiceTitan, 2026, vendor data).
• Of calls to home service businesses, 52 percent were answered by a person (Invoca, 2026, vendor data).

• 52%of calls to home service businesses were answered by a person

  Source: Invoca, Home Services Lead Conversion Benchmarks Report, July 2026. Vendor data, averages across Invoca’s customers.

None of it says what a slow call-back costs. That number is in your own call history.

Time your own call-backs

You need one month of call history and your job list.

1. 1Step 1 of 5

   Download the past 30 days of call history soon, because phone companies keep it only so long. AT&T’s Business Phone history shows “either your last 100 calls or calls for the last 60 days.” Cox Business, which sells phone service in Gainesville, says its online call records for IP Centrex customers “go back as far as 120 days.” Our call-history article covers where to find it, company by company.

2. 2Step 2 of 5

   Mark every call from a real customer that nobody answered. Skip spam and wrong numbers.

3. 3Step 3 of 5

   For each one, find your call back to that number and put the gap in one of five groups: within the hour, the same day, the next morning, later, never.

4. 4Step 4 of 5

   Check each number against your job list and mark the ones that booked.

5. 5Step 5 of 5

   Compare the share that booked in each group.

Then put a dollar figure on the slow group. Here it is with example numbers, using the 4-in-10 booking rate and $151 profit a job from our sample walk sheet.

Say 10 callers last month got their call back later than the next morning, and 1 of them booked. Callers you got back to within the hour booked 4 in 10.

• Had the late 10 booked like the fast ones: 10 × 40 percent = 4 jobs.
• They booked 1, so they came up 3 jobs short.
• 3 jobs × $151 profit a job = $453 a month.

Late quotes, counted the same way

Both studies timed the first call. Neither looked at quotes, and we found no trade study that ties days-to-quote to jobs won. So split your quote export by days from your visit to the quote, and compare how many were accepted in each group.

On our sample sheet, 9 of 22 quotes went out 48 hours or more after the visit and 5 of those went cold: 5 jobs × $380 profit a job = $1,900 a month. Those are illustrative numbers too.

Where to start

If you return most calls later than the next morning, look first at a rule on the line you already pay for: calls that ring out go to a second phone or a callback list, and someone works the list first thing.

If you also send a missed-call text, give a call-back time you can keep. Our texting article covers what a missed-call text may say in Florida.

Pull last month’s call history this week, mark the gaps and run the calls nobody returned through our loss calculator. It asks for weekly numbers, so divide each 30-day count by 4.3. It runs on your phone or computer and sends us nothing.

Sources, in plain words

• James B. Oldroyd, Kristina McElheran and David Elkington, “The Short Life of Online Sales Leads”, Harvard Business Review, March 2011. Business press; one author was the chairman and CEO of InsideSales.com. https://hbr.org/2011/03/the-short-life-of-online-sales-leads
• InsideSales.com, Lead Response Management Study, dated October 2007 (report pages marked 2009). Vendor study, six companies, web-form leads. Archived copy of the original report: https://web.archive.org/web/20160322104646/http://www.leadresponsemanagement.org/images/lrm_study.pdf
• ServiceTitan, call booking rates by trade, shop size and time of day, October 25, 2022. Vendor data, more than 3,000 businesses in the US and Canada. https://www.servicetitan.com/blog/data-call-booking-rates
• ServiceTitan, the summer after-hours call spike in residential AC, June 3, 2026 (2025 data). Vendor data from its own customers; ServiceTitan says the figures “are not nationally representative.” https://www.servicetitan.com/toolbox/state-of-the-trades/trends/hvac-summer-after-hours-call-spike
• Invoca, Home Services Lead Conversion Benchmarks Report, July 2026. Vendor data, averages across Invoca’s customers. https://www.invoca.com/reports/the-invoca-home-services-lead-conversion-benchmarks-report-2026
• AT&T, “Access your Business Phone call history”, updated December 12, 2025. Vendor help page. https://www.att.com/support/smallbusiness/article/smb-digital-phone/KM1194880/
• Cox Business, “Viewing Call Detail Records in MyAccount” and “Business Phone Service in Gainesville, Florida”. Vendor help and service pages, no date shown. https://www.cox.com/business/support/viewing-call-detail-records-in-myaccount.html and https://www.cox.com/business/local/services/fl/gainesville/phone.html
• Lumora research, 2026: our trace of the speed-to-lead figures to their sources.

Questions and answers

Is it true that calling back within five minutes makes you 100 times likelier to reach a customer?

It is what one 2007 study reported after pooling web-form leads from six companies it does not name. The study was by InsideSales.com, a dialing software company. It compared the odds of reaching a person at 5 minutes and at 30, not the odds of booking a job, and it says it “did not address close ratios.” The report does not say what the six companies sold.

Where does the Harvard Business Review speed-to-lead figure come from?

From “The Short Life of Online Sales Leads”, HBR, March 2011, by James Oldroyd, Kristina McElheran and David Elkington, the CEO of InsideSales.com. It audited how 2,241 US companies answered a web-form test lead and studied 1.25 million online leads at 42 companies. It measured qualifying a lead, meaning a meaningful conversation with a decision maker, not a sale.

How fast should a trade shop return a missed call?

As fast as it can; the studies point that way. None of them timed call-backs against booked jobs, and we found no study that did it for trade shops, so time your own from 30 days of call history and your job list.

How do I measure my own call-back times?

Download 30 days of call history, mark each real customer nobody answered, write the gap to your call back, check which numbers booked and compare the share that booked in each group.

Do late quotes cost jobs too?

We found no trade study that ties days from visit to quote to jobs won. Split your own quote export by days from visit to quote, and compare how many were accepted in each group.

Read next

• How to get 30 days of call history from your phone companyCarrier by carrier, where a Gainesville shop downloads the past 30 days of calls, and one pass through the file to find the calls nobody returned.
• Texting your customers in Florida: five rules in plain wordsWhat a missed-call text may say, when you need written consent and how to handle STOP.