The short answer: inside the hour, and in minutes when you can
The best-documented finding is about the first hour. In a study of 1.25 million sales leads reported in Harvard Business Review in 2011, firms that tried to reach a lead within an hour were nearly seven times as likely to have a meaningful conversation with a decision maker as firms that tried an hour later. An older study of six companies found a five-minute call beat a thirty-minute one by a wide margin. Neither counted sales, and neither involved a homeowner phoning a trade business. So reply within minutes when you can, never let an inquiry pass the hour without a person on it, and treat the famous multipliers as direction, not a forecast.
Below: each quoted figure next to its study, the popular numbers that don't hold up, what the evidence means for a 5-to-30-person service business, and how to measure your own response time. The system that makes fast replies happen is its own post, how to automate lead follow-up. This page is the evidence behind it.
Seven sources, and what each one actually measured
Nearly every speed-to-lead number online traces back to one of these. The last column is my read for a small service business, not a rating from the study.
| Statistic as usually quoted | What the study actually measured | Source and year | How much weight to give it |
|---|---|---|---|
| “Call within 5 minutes and you’re 21x more likely to qualify the lead, and 100x more likely to reach them.” | Three years of data from six companies on InsideSales.com’s calling system: 15,000+ web-form leads, 100,000+ call attempts. It compared the odds of reaching and qualifying a lead called at 5 minutes vs. 30. Each company defined “qualified” its own way, and close rates weren’t studied. | InsideSales.com/MIT Lead Response Management Study, Oldroyd and Elkington, October 2007 (archived copy) | Low for the multipliers. Six companies, one vendor’s customers, 19 years ago. Good for direction only. |
| “The average company takes 42 hours to respond to a lead.” | An audit: one web-generated test lead sent to each of 2,241 U.S. companies. 37% replied within an hour, 16% in one to 24 hours, 24% after more than a day, and 23% never. The 42 hours is the average among companies that replied within 30 days. | Oldroyd, McElheran and Elkington, “The Short Life of Online Sales Leads”, Harvard Business Review, March 2011 | Moderate, as a 2011 snapshot. Shows how common slow replies were. Says nothing about what speed earns you. |
| “Respond within an hour and you’re 7x more likely to qualify the lead.” | 1.25 million leads at 29 consumer and 13 B2B U.S. companies. Firms that tried to make contact within an hour were nearly seven times as likely to qualify the lead (a meaningful conversation with a key decision maker) as firms that tried an hour later, and over 60 times as likely as firms that waited a day or more. | Same HBR article, March 2011 | The strongest in the set. 1.25 million leads across 42 companies, mostly consumer-facing. Still counts conversations, not sales, and the published method is one sentence. |
| “Calling within one minute raises conversions 391%.” | Almost 3.5 million leads at 400+ Velocify client companies, generated in the first half of 2012. A chart shows the conversion lift by time to first call: 391% at one minute, down to 17% at 24 hours. The report doesn’t say what the lift is measured against. | Velocify, “The Ultimate Contact Strategy” (2012 lead data) | Low. Big data set, but a lift with no stated baseline can’t be checked or applied. |
| “Only 7% of companies respond within five minutes.” | A secret shopper filled in lead, demo, and sales-inquiry forms at 433 B2B software companies. 7% replied within five minutes, and 55% hadn’t replied after five business days. | Drift lead response survey, February 2017 | Low for trades. Software demo forms, not service calls. It does show the 2011 pattern held six years on. |
| “Reply within 30 minutes and you’re 25% more likely to be hired.” | Thumbtack’s data on local service pros: those who replied to a lead within 30 minutes were 25% more likely to be hired than those who replied within three hours. No sample size or method is published. | Thumbtack, “3 takeaways from the 2025 Pro Summit”, updated November 2025 | Moderate. The closest match to a local service business, and the only row that clearly measures hiring. Unpublished method, and the post promotes Thumbtack’s own fast-reply tools. |
| “A third of contractors never answer a web inquiry.” | An analysis of 100 plumbing and HVAC contractor websites in Dallas and Houston: 34% never responded to a web inquiry. The top 10% averaged a 23-minute first response, the bottom group over six hours. How the inquiries were sent and timed isn’t described. | Michael Carpenter (founder of Forge, a contractor marketing platform), Plumbing & Mechanical, September 2026 | Low to moderate. The right trades and recent, but small, two cities, and no published method. |
Three things repeat down the middle column. Most studies measured whether a salesperson reached or qualified a lead, not whether anyone bought. The two most quoted are from 2007 and 2011. And every one was run or co-written by a company selling sales, lead, or marketing tools: InsideSales.com's CEO at the time co-presented the 2007 study and co-wrote the 2011 one. I sell a fast-reply tool too. That doesn't make the findings wrong, but it's a reason to lean on the ones that show their method.
Popular numbers that fall apart when you trace them
These show up across speed-to-lead roundups. For the first two I couldn't find any study at all. The last two trace back to a real study that says something different. Don't base a decision on any of them.
“78% of customers buy from the company that responds first.”
Usually credited to a “Lead Connect” survey. I found no report, sample, or year for it, only blogs repeating it. Some credit the 391% figure to the same survey; it’s Velocify’s. And the 2026 contractor piece in the table says the first contractor to respond wins the job more than 70% of the time, and cites nothing.
“35–50% of sales go to the vendor that responds first.”
Usually credited to InsideSales.com. It isn’t in the InsideSales.com/MIT study summary, and the oldest repetition I found, a 2015 blog post, cites nothing for it.
“Every 10-minute delay cuts your odds by 400%.”
A garbled version of a real line. The 2007 study found the odds of qualifying a lead fell fourfold between minute 5 and minute 10. That was one early window, about qualifying, not buying, and a fourfold drop isn’t a 400% drop: nothing falls by more than 100%.
“Leads reached in 5 minutes are 21x more likely to become customers.”
The 21x is real, but it measured qualifying, at six companies, in 2007. That study says outright it didn’t look at close rates.
What carries over to a 5-to-30-person service business
Almost all of this evidence comes from sales teams following up web forms, in fields like insurance, lending, cars, and software. A homeowner with water coming through the ceiling is a different buyer. They're local, often in a hurry, and often calling rather than filling in a form.
What carries over
- Direction. Every study here that compared fast and slow replies found fast did better, across nearly two decades and very different businesses.
- The first hour. It’s the best-documented threshold, and Thumbtack’s local-pro figure points the same way.
- Slow is normal. From the 2011 audit to the 2026 contractor sample, a big share of businesses never replied, so a fast reply stands out.
What doesn't
- The multipliers. 21x, 100x, and 391% describe specific sales teams studied in 2007 and 2012. They aren’t a forecast for your phone.
- Sales. Thumbtack’s is the only figure that clearly measures getting hired, and it’s 25%, not 21 times. The windows differ, so they can’t be compared directly, but it hints the effect on hires is smaller than on reaching someone.
- Texts to missed callers. None of these studies timed an automatic text to someone whose call went unanswered.
- Job type. No study I found splits the results by how urgent the job is.
On that last point, reasoning rather than data: a no-heat call in January probably has a shorter window than a fence quote for next spring. The first caller is working down the list now; the second is collecting bids. If your work is mostly planned, a same-day reply may hold up. If it's mostly emergencies, assume the window is minutes.
And who can skip all this: if a person answers your phone live all day and you're booked weeks out, response time probably isn't what's holding you back.
Response-time targets worth setting, and why
These are judgment calls built on the studies above, not findings from them.
- 01
A missed call during working hours
A text within a couple of minutes, then a call back within the hour. The text holds the caller while you finish the job; the hour is where the best evidence sits.
- 02
A web form or email
A reply within five minutes during working hours, then a person within the hour. If the first reply can’t answer the question yet, it should at least say when someone will call.
- 03
Anything after hours
A clear promise of when you’ll call, kept first thing the next morning.
- 04
Inquiries that never get a reply
Zero. The no-reply share was 23% in the 2011 audit and 34% in the 2026 contractor sample. Getting it to zero takes no extra speed, only one place every inquiry lands.
A text back covers the first target and only that. Missed-Call Text Back texts the caller within seconds when you can't pick up (a minute or two if they leave a voicemail), and holds the text until 8 am for calls between 8 pm and 8 am. Whether that wins the caller back depends on who calls you; see when a missed-call text actually works and when it doesn't.
Measure your own response time in one afternoon
Your own numbers beat every row in that table. You need your call history, your inbox, and a spreadsheet.
- 01Pull 30 days of call history from your business phone app or carrier portal. List every missed call that wasn’t a current customer or a robocall.
- 02Next to each, note when you first called or texted back and the gap in minutes. Calls nobody returned get their own column.
- 03Do the same for web forms and email: when the inquiry arrived against your first reply in your sent mail or CRM.
- 04Split the rows into working hours, evenings, and weekends to see where the slow ones cluster.
- 05Count three numbers: the median gap, the share answered within an hour, and the share never answered. The 2011 audit’s 37% and 23% make a rough, old benchmark.
- 06Then test yourself the way the 2011 audit did: have a friend fill in your form and call while you’re on a job. Time what happens.
If the never-answered column isn't empty, that's a capture problem before it's a speed problem, so fix it first. To put a dollar figure on the misses, run them through the ROI calculator. If it turns up one queue, the same afternoon method finds the other bottlenecks in your business.