Speed-to-lead: why the first five minutes decide the sale.

The two most-cited studies on this are both old, and both trace back to one research lineage: a 2007 analysis of roughly 15,000 leads across six companies, run by Dr. James Oldroyd with data from InsideSales.com, and a 2011 Harvard Business Review follow-up analyzing 1.25 million leads across 2,241 companies. Both found the same shape: contact and qualification odds fall sharply as response time grows, and fall fastest in the first hour.

The 2007 study, later summarized as a collaboration with MIT’s Sloan School, where Oldroyd held a faculty fellowship, tracked three years of activity across six companies: more than 15,000 leads and over 100,000 recorded call attempts. Its two headline figures: a company that called a lead within five minutes was roughly 100 times more likely to make contact than one that waited thirty minutes, and about 21 times more likely to actually qualify that lead. The original study write-up is still hosted here.

Harvard Business Review’s 2011 follow-up, “The Short Life of Online Sales Leads” (Oldroyd, McElheran and Elkington), analyzed a far larger set: 1.25 million leads across 2,241 U.S. companies. It found the bigger problem wasn’t really the five-minute window. It was the hour around it. Only 37% of companies responded within an hour at all, 23% never responded, and average response time among the companies that did respond was 42 hours. Firms that called within an hour were about seven times more likely to qualify a lead than those that waited one more hour, and roughly 60 times more likely than firms that waited a full day.

“The lead that sits for six hours isn’t losing to a better pitch. It’s losing to whoever picked up the phone first.”

Both studies are worth reading with their limits in view. Neither is a randomized trial. They’re analyses of real call and CRM data from inbound B2B sales leads specifically, run by or with a company that sells lead-management software, not audited independent research. The underlying data is now roughly two decades old, predating smartphones, SMS-first communication, and chat as a default channel. And a lot of what gets presented online as a “current” statistic on this topic just cites the same 21x and 100x multiples back to this one 2007 dataset rather than re-deriving them, which is worth noticing when a 2026 blog post quotes a twenty-year-old number as this quarter’s finding.

One partial exception: a 2025-2026 vendor benchmark analysis of CRM timestamp data across 939 B2B software companies reports the same order of magnitude from a newer dataset: average response time of 47 hours, only 23% of companies responding inside five minutes, and roughly 2.6× higher close rates for leads contacted inside five minutes versus after 24 hours. See that analysis here. Worth flagging honestly: the exact “21 times more likely to qualify” figure this newer benchmark reports is identical to the 2007 number. That either means the underlying pattern has held remarkably stable for two decades, or means the newer figure was borrowed rather than independently derived. Nothing published settles which, and neither should you assume. What has actually held up across every dataset available, old and new, is the shape of the curve rather than any single exact multiplier: waiting costs you the lead, and the cost compounds hardest in the first hour.

Inside five minutes for anything that arrives while the business is open, and inside a minute or two for a missed call, where an automated text can go out within seconds. Past roughly thirty minutes, both your odds of reaching the person and their odds of still wanting to talk drop hard, in every study on this topic since 2007. The precise number matters less than the shape: minutes, not hours, and never “whenever someone gets to it.”

The five-minute figure gets treated as gospel because it’s memorable, but the operational target underneath it is simpler: be first, or be automated fast enough that it doesn’t matter who technically called first. A homeowner with storm damage typically calls more than one roofer and books whichever one answers first; in a 2025 homeowner survey cited by CallRail, 97% said response speed influences who they hire, which is why Kova’s roofing page treats the first answer as the thing that decides who books the estimate. Kova’s own review of 21 Sacramento roofing websites found the other half of that gap sitting in plain sight: only 2 of the 21 put a number on how fast they respond, and 8 of them gave a visitor arriving at 9pm no after-hours signal of any kind. The exact minute count shifts by industry. The requirement that a human, or a convincing automated stand-in, respond before the prospect moves to the next name on their list does not.

Almost never because anyone is being lazy. In practice, four infrastructure gaps account for most of the delay: the lead notification lands in a shared inbox nobody’s actively watching, the form only sends an email instead of triggering an action, no single person owns the response queue after five o’clock, and the CRM isn’t connected to the ad platform that generated the lead in the first place. Each one is a wiring problem, not an effort problem, and each one is fixable without hiring anybody.

Picture the actual failure sequence at a typical small or midsize business. A lead comes in through a website form. The form’s only integration is an email notification, so it lands in a shared inbox (sales@, info@, or a single rep’s personal inbox) that nobody’s covering while they’re on a job site or in a client meeting. If it arrives after 5pm, or on a Saturday, there’s often no one assigned to check that inbox until Monday. And even when someone finally does see it, there’s usually no system connecting that email back to the ad campaign, keyword, or creative that produced it, so nobody downstream ever learns which channels are producing leads that actually get answered.

None of that is a motivation problem. It’s a gap between the moment a lead is captured and the moment a human or an automated system acts on it. A form that emails instead of triggering an action, a queue with no named owner, and a CRM that doesn’t talk to the ad platform will produce the same slow response no matter how hard the team behind it is working.

Five specific mechanisms, roughly in order of leverage: an instant automated response that sets a real expectation, missed-call text-back so an unanswered call still gets a reply, routing rules that assign a named owner to every lead source, an escalation path for when the first attempt goes unanswered, and time-to-first-touch tracked as an actual reported number instead of a vibe. None of these require new headcount. They require the notification, the routing and the queue to be built as a system instead of assumed.

1. 1

   Instant auto-response. The moment a form is submitted or a lead ad converts, an automated SMS or email goes out immediately confirming receipt and setting a real expectation (“we’ll call within the hour”) before a human is even involved.

2. 2

   Missed-call text-back. A call that rings out or goes to voicemail triggers an automatic text within seconds, so a missed call stops being a dead end and becomes the start of a conversation instead.

3. 3

   Source-based routing, with an owner. Every lead source (form, missed call, chat, an ad platform’s native lead form) routes to one accountable person or queue, never a shared inbox nobody’s specifically assigned to check.

4. 4

   An escalation path when the first attempt fails. If the first outreach goes unanswered, a second touch, then a third, fire automatically on a defined schedule instead of the lead going quiet after one unreturned call.

5. 5

   Time-to-first-touch as a tracked metric. Measured and reported the same way a marketing team already tracks cost per lead, using the same tracking and attribution layer that ties a lead back to the ad that produced it, not left as an assumption nobody checks.

It varies by channel, but the target is almost always minutes, and the mechanism that hits it is almost always automated rather than manual. A website form and a missed call both need a response inside about five minutes. Live chat needs one inside the conversation itself. A lead ad on Meta or Google needs a text the moment the platform syncs it to the CRM. And anything that arrives after hours needs an immediate automated acknowledgment with a specific next-morning follow-up time, not silence until someone happens to open the inbox.

These are operational targets, not results measured for a specific client. Kova has no published client data. The discipline is the same regardless of vertical: automate the acknowledgment, route by source, and track the number.

None of this requires reinventing a marketing plan. It requires auditing one number most businesses aren’t tracking yet: the actual elapsed time between when a lead arrives and when a human first responds, broken out by source and by hour of day. Most teams that measure it for the first time are surprised by what they find. The fix, once the number is visible, is rarely more ad spend. It’s closing the gap between the ad and the first real response, and that gap sits with whoever answers the leads.