Lead Response Time Benchmark 2026: 7 Million Leads, 70 Million Calls, One Pattern

What the open research says, study by study

Most lead response statistics online are copies of copies, and a few are inventions. So this benchmark starts with the studies you can open yourself, each with its sample size, and it names the numbers that fail that test. Together they cover more than 7 million web leads, more than 70 million inbound phone calls, 2,241 audited companies and 433 secret-shopped companies. Studies from 2020 onwards carry the argument and come first; the older foundational studies follow, each with a line on how much weight it still deserves. The table is the short version; the detail follows.

• Invoca, Lead Conversion Benchmarks, 2026, drawn from more than 70 million tracked calls. In home services, 48 percent of callers never speak with a person. Invoca's Home Services Buyer Experience Report adds that 60 percent of consumers waited more than an hour for a response after submitting a lead form, and 79 percent say they will switch to the business that responds fastest.
• CallRail, Small Business Marketing Benchmark Report, January 2025. 1.1 million leads across automotive, financial services, healthcare, home services, legal and real estate. Missed call rates: healthcare 32 percent, legal 28 percent, home services 14 percent, real estate 9 percent. The same company's September 2025 analysis puts the all-business average at 28 percent of calls unanswered, and states that up to 85 percent of callers whose calls go unanswered will not call back.
• InsideSales, Lead Response Study, 2021. More than 55 million sales activities on 5.7 million inbound leads at more than 400 companies. Conversion rates are more than eight times higher when the first attempt is made inside five minutes than between five minutes and 24 hours. 57.1 percent of first call attempts happen more than a week after the lead arrives. Only 0.1 percent of inbound leads are engaged in under five minutes.
• Pew Research Center, 2020. Eight in ten Americans do not generally answer their cellphone when an unknown number calls; 67 percent let it ring and check for a voicemail.
• Drift and Salesloft, Lead Response Survey, 2017. A secret shopper submitted forms to 433 business-to-business software companies. 7 percent responded within five minutes. 55 percent had not responded after five business days. 14 percent offered live chat. Weight today: secondary. A pre-2020 audit of software companies; the 2021 and 2026 datasets show the same pattern at far larger scale.
• Harvard Business Review, 2011, The Short Life of Online Sales Leads (Oldroyd, McElheran, Elkington). Two studies. First, an audit of 2,241 US companies that received a test web lead: 37 percent responded within an hour, 16 percent within one to 24 hours, 24 percent took more than 24 hours, and 23 percent never responded. The average response time among companies that replied within 30 days was 42 hours. Second, 1.25 million leads at 29 consumer and 13 business-to-business companies: firms that tried to contact the prospect within an hour were nearly seven times as likely to qualify the lead as those that tried an hour later, and more than 60 times as likely as firms that waited 24 hours or longer. Weight today: the 2011 audit of response rates is a historical baseline, not a current rate; the qualification odds describe how buyers behave after asking a question and were reproduced on 5.7 million leads in 2021.
• MIT Sloan and InsideSales, Lead Response Management Study, 2008 (Oldroyd and Elkington). Three years of data, six companies, more than 15,000 web leads, more than 100,000 call attempts. The odds of contacting a lead called at five minutes versus thirty minutes are 100 times higher; the odds of qualifying it, 21 times higher. The odds of contact fall more than tenfold and qualification more than sixfold in the first hour. After 20 hours, every additional dial reduces the chance of contact. Best hours: 4 to 6 PM for contact, 8 to 9 AM and 4 to 5 PM for qualifying. Best days: Wednesday and Thursday. Weight today: the origin of the five-minute rule and still the clearest statement of the mechanism. Treat the exact multiples as directional; every dataset since confirms the direction.
• Not verifiable: the claim that replying within one minute lifts conversion by 391 percent. It is attributed to a Velocify analysis of 3.5 million leads that no longer exists at any open address. Quote it as a vendor claim or not at all.

Finding one: the odds are brutal and they have not changed in fifteen years

The 2021 InsideSales study, run on 5.7 million leads, lands exactly where the 2008 MIT study and the 2011 HBR study did, on different data and thirteen years later. Five minutes is the window. One hour is the cliff. A day later the lead is effectively gone, with contact odds a sixtieth of what they were. What changes when you answer in five minutes walks through the mechanism; the figure below shows the shape.

That consistency is the point. These are not findings about software or about one industry. They describe how a person behaves after asking a question: they are available for a few minutes, they move on, and whoever reaches them first while they are still available gets the conversation. Everything else in this benchmark is about how far the typical business sits from that window and how a run account closes it.

Finding two: most businesses are hours or days away from the window

In 2011, 23 percent of 2,241 audited companies never answered a web lead and the average was 42 hours. In 2017, 55 percent of 433 software companies had not answered after five business days. In 2021, across 5.7 million leads, 57 percent of first call attempts came more than a week later and one lead in a thousand was engaged inside five minutes. In 2026, 60 percent of home services consumers say they waited more than an hour for a form response.

Three different methods, fifteen years apart, one answer: the typical business responds in days, not minutes, and a large minority never responds. If you are reading this to find out whether you are behind, the honest prior is yes, and the useful question is by how much, which the measurement section answers. The usual reason is not effort. It is that nobody owns the queue, as the CRM follow-up system for a service business sets out.

What a run account achieves against that research

To show what the standard looks like in practice, we read 90 days of CRM message timestamps, July 6 to October 4, 2026, across 376 inbound leads at five Megawebvision-run service businesses in the United States and Canada: window and door companies in New Jersey, Ontario and Quebec, a property manager in Texas, and a janitorial company in California. Each runs the response system described in AI for lead generation: instant acknowledgement on every channel, missed-call text-back, and a named person on the queue. Only aggregates are published, no business is identified, and the definitions are in the data file under Sources.

• Channel mix across the 376 leads: phone calls 52 percent, text 13 percent, live chat and web chat 12 percent, phone-menu calls 9 percent, Facebook and Instagram 9 percent, web forms 5 percent, email 1 percent. Half the demand still arrives by phone.
• Busiest arrival hour: 9 AM local, then 10 AM, noon and 2 PM. Almost nothing before 7 AM or after 9 PM, but one business saw 53 percent of its leads outside staffed hours, so the after-hours share is a property of each business's channel mix and ad schedule, not of its trade.

Finding three: automation wins the first ten seconds, a person still has to win the five minutes

The gap between 0.1 percent and 92 percent inside five minutes is automation: an acknowledgement on every channel, a text-back on every missed call, a chat that answers at 9 PM. That is the floor every business can install this week. It is not the whole job, and the research says why.

Nielsen Norman Group's interface limits are 0.1 seconds to feel instant, 1 second to keep the flow of thought, and 10 seconds to keep attention. A chat widget or text thread is an interface, so acknowledgement inside ten seconds is a design requirement. In the same group's usability study of chatbots, participants found customer-service bots less helpful than human representatives, valued them mainly for speed, and were pleased when a business said plainly that it was a bot. Salesforce's research found 89 percent of customers want to know whether they are talking to AI, and 80 percent say it is important that humans validate AI output.

So the bot buys ten seconds and should say what it is. A person still has to arrive inside five minutes during staffed hours, and that arrival needs its own timestamp. Standard CRM message logs stamp automation and people the same way, which is why the human first-touch field is the first thing a Megawebvision engagement adds: without it, a five-minute target is measured on the bot and the person is invisible. The AI receptionist guide covers the handoff from the automated reply to the named person.

Finding four: the phone is the biggest channel and the leakiest

Half of the leads in the sample came by phone, CallRail's benchmark shows phone as the primary conversion path for local services, and Invoca's 70-million-call data puts the share of home services callers who never reach a person at 48 percent. In any call log, a share of connected calls ends inside ten seconds: robocalls, wrong numbers, and prospects who met a menu and left. A call log that reads answered is not a call log that reads handled.

The callback does not rescue the miss. Pew found eight in ten Americans do not answer unknown numbers, and the office line calling back is an unknown number. CallRail's figure that up to 85 percent of unanswered callers will not call back is a vendor claim, but it is consistent with Pew. A missed-call text inside two minutes, from the number the prospect dialled, is the floor that turns the miss into a thread, and it is the single practice that separated the run accounts from the industry figures above.

1. Ring a named person's mobile before voicemail, not after. A ring group with a 20-second handoff removes most misses.
2. Send the missed-call text from the number the prospect dialled, signed with a first name, with a time to call back, so the return call is expected rather than unknown.
3. Count connected calls over 30 seconds as your answer rate. Report that number, not the raw one.
4. Read the ten shortest connected calls each week. If they are prospects hitting a menu, the menu is costing you leads.

What customers expect now

The expectation surveys agree on direction, with the caveat that each publisher sells a cure. Salesforce, 2023: 77 percent of customers expect to interact with someone immediately when they contact a company. Zendesk CX Trends, 2025: 74 percent of consumers now expect customer service to be available 24/7 because AI exists, and 85 percent of CX leaders say customers will drop a brand over an unresolved issue even on first contact. HubSpot: 58 percent say their expectations are higher than a year ago, 67 percent of those expect faster service, and 92 percent will stop buying after three poor experiences. Invoca, home services: 79 percent will switch to the business that responds fastest.

One more shift matters for how leads arrive at all. BrightLocal's 2026 consumer survey found use of ChatGPT and similar tools for local business recommendations rose from 6 to 45 percent in a year. A growing share of your leads will have been told which business to call by an AI assistant, and will expect the same speed from you that they just got from it. How to get cited by ChatGPT and Google AI covers the other half of that shift.

What a slow response costs: run your own numbers

This is an illustrative example with made-up inputs, built so you can swap in your own. A roofing company gets 120 inbound leads a month. 40 percent arrive outside staffed hours, which is 48 leads. Today those get a voicemail greeting and a next-morning callback, so the first human attempt lands 12 to 15 hours later. The MIT and HBR odds put contact after the first hour at less than a tenth of the five-minute rate.

If the business reaches 60 percent of its in-hours leads and closes 25 percent of those at an average job of $9,000, the in-hours math is 72 leads, 43 contacted, 11 jobs, roughly $97,000 a month. The after-hours math at a tenth of the contact rate is 48 leads, 3 contacted, less than one job. Move the after-hours contact rate to even half the in-hours rate, with a receptionist, human or AI, that books the next-morning slot while the prospect is still on the page, and the same 48 leads produce about 14 contacts and 3 to 4 jobs: $27,000 to $36,000 a month from leads already paid for. Your inputs will differ. The structure of the loss will not. Which stage is your constraint is the question that decides whether this is the number to fix first.

Benchmarks to hold yourself to

These targets follow from the studies above. The middle column is what the evidence says the typical business does; the last is what a run account does.

The questions owners ask, answered in one line each

These are the questions people type into Google and ask AI assistants about lead response, with the short answer the evidence supports.

• What is the five-minute rule for leads? Attempt human contact within five minutes of the lead arriving. In the MIT study, contact odds were 100 times higher at five minutes than at thirty, and qualification odds 21 times higher.
• How quickly should you respond to leads? Acknowledge inside ten seconds by automation, reach a person inside five minutes during staffed hours, and book or text a named callback inside two minutes outside them.
• What is considered a good response time? Five minutes to a human attempt. One hour is the outer limit; after it, contact odds drop more than tenfold and qualification odds sevenfold.
• What is the average response time for inbound leads? The only open audit, 2,241 companies in 2011, found 42 hours among companies that replied at all, with 23 percent never replying. No newer open study measures it across industries. Measure your own from CRM timestamps.
• What percentage of businesses never respond to leads? 23 percent in the 2011 audit of 2,241 companies; 55 percent of 433 software companies had not responded after five business days in 2017.
• What is the biggest mistake in lead generation? Buying more leads before measuring what happens to the ones you have. The research says the typical business loses most of its leads to delay, and the fix costs less than the next campaign.
• Does an automated text count as responding? It counts as acknowledging. The contact and qualification odds in every study are about a person reaching the prospect.

The seven-step plan for next week

Each step is one hour or less and uses data you already have.

1. Export 90 days of inbound conversations from the CRM with timestamps, direction, channel, call status and call duration.
2. Tag each lead with its local arrival hour and weekday. Count the share outside your staffed hours. That share is your after-hours decision.
3. Count missed calls and connected calls under 10 seconds. Together they are your real missed rate. Compare it with 14 percent for home services and 28 percent all-business.
4. Add one field or tag that a person sets the moment they first engage a lead. From next month, lead response time means time to that field, not time to the bot.
5. Set two targets: automation acknowledges inside 10 seconds and says it is automated; a named person attempts contact inside 5 minutes during staffed hours.
6. Cover the after-hours share. Under 15 percent, a ring group to a mobile is enough. Over a third, test an AI receptionist or answering service for 30 days against your voicemail baseline.
7. Review the week every Friday: after-hours share, missed rate, median time to the human field, and the ten shortest calls. Fix the largest gap first.

What this means for your next decision

Fifteen years of research on millions of leads says five minutes. Three audits say most businesses take days or never answer. Seventy million calls say a third to a half of callers never reach a person. Accounts running a response system reply in seconds and text back every missed call in minutes. None of that is a software problem. It is one field in the CRM, one named person per hour of the day, and one weekly review.

Megawebvision runs this measurement at the start of every engagement, because response is the stage most owners believe is fine and the one the timestamps most often contradict. If you want yours read against the research and this benchmark, the diagnostic starts there.