Lead Response Time Beats Clever Follow-Up — And How to Get Fast

Nearly every team trying to fix its conversion rate rewrites the follow-up email first. It is the wrong end of the problem. The expensive failure is almost never what you said in the third message; it is that the first message took nine hours, by which time the enquiry had been answered by whoever replied in four minutes.

The useful part is that this is rarely a sales-discipline problem. In most Indian SMBs the delay is plumbing — a form that emails an inbox that nobody owns — and plumbing is cheap to fix compared with hiring better salespeople or writing better sequences.

The research behind all this is genuinely strong on direction and genuinely old on specifics, and the gap between those two things is where most of the nonsense on this topic lives. So let us start there.

What does the research actually say?

Three studies do most of the work, and they are not interchangeable.

The InsideSales.com/MIT Lead Response Management study (2007). This is the source of the famous numbers, and it is worth reading rather than quoting. Dave Elkington of InsideSales.com and Professor James Oldroyd, then at MIT Sloan, presented it at MarketingSherpa's B2B Demand Generation Summit in October 2007. The method: three years of call data from InsideSales.com's own system, across six companies that generate and respond to web leads, covering more than fifteen thousand leads and over one hundred thousand call attempts. The findings, in its own terms:

• The odds of contacting a lead called at 5 minutes versus 30 minutes drop by 100 times; from 5 minutes to 10 minutes alone they fall by 5 times.
• The odds of qualifying a lead drop 21 times between 5 and 30 minutes, and 4 times between 5 and 10 minutes.
• Across the first hour, the odds of contact fall by over 10 times and of qualifying by over 6 times.
• After 20 hours, each additional dial actively hurt the chance of making contact to qualify.

The original site is gone; the full text survives as an archived copy of the study page.

"The Short Life of Online Sales Leads" (Harvard Business Review, March 2011). By Oldroyd, Kristina McElheran and Elkington — the peer-reviewed citation record is public and the article is on HBR, behind a paywall. Its own abstract says plainly that most companies are not responding nearly fast enough. The widely circulated figures attributed to it — an audit of a few thousand firms, an average response time measured in days, a multiple for qualifying within the hour — come from secondary summaries that disagree with each other, and I could not verify them against the article. So they are not in this post. If you want to cite HBR on this, buy the article.

Workato's response-time test (published March 2026). The most recent thing with a stated method: demo requests submitted to 114 B2B companies, with the responses timed. Their findings: exactly one of 114 sent a personalised email within five minutes and none called within five minutes; personalised email averaged 11 hours 54 minutes; only 31% ever responded by phone at all, averaging 14 hours 29 minutes. The number that matters most here: companies using a lead routing tool averaged 3 hours 32 minutes, against nearly 13 hours for those without one.

Study	Year	What it measured	Main limitation
InsideSales/MIT LRM	2007	Phone contact and qualification odds by minute of first dial	Six companies, US B2B, vendor's own data, phone only
HBR, Oldroyd et al.	2011	How fast companies actually respond to web leads	Paywalled; circulated numbers unverifiable
Workato	2026	Observed response times at 114 B2B firms	Small sample, B2B software, no close-rate data

How old is too old?

The 2007 study is the one everybody cites and nobody reads, so here is what you are actually citing.

It is three years of data ending around 2007, from six companies, in US B2B, on the telephone. The data came from the software vendor that sold lead-response tooling, which is a real conflict of interest even when the analysis is honest. "Contact" meant a connected call lasting between two and six minutes depending on the company. And the study itself says it "did not address close ratios" — it measured contact and qualification, not revenue.

The clearest evidence of how badly its specifics travel is in its own secondary findings: it concluded that Wednesday and Thursday were the best days to call, 4-6pm the best window for contact, and Monday consistently poor. Those are artefacts of US office hours in 2006. Nobody should run an Indian sales team off them in 2026, and the fact that people still repeat them should make you equally sceptical of the 100x.

What survives, and what I would be willing to defend: the relationship between delay and contact rate is very steep and most of the damage happens in the first hour, not the first day. That direction is consistent across all three studies and matches what we see in client data. The precise multiplier does not survive, and anyone presenting "100x" as a current benchmark is quoting a nineteen-year-old vendor study they have not read.

The honest version of the claim is this: you do not need to know whether it is 100x or 6x to know that replying in five minutes rather than five hours is worth more than any rewrite of the email. That is enough to act on.

Where do the minutes actually go?

This is the part worth mapping, because almost nobody does. A realistic chain for an Indian SMB with a website form and a small sales team:

Step	Typical delay	Why
Form submitted → email sent	0–2 min	Usually fine
Email lands in a shared inbox	0–15 min	Spam filtering, shared-inbox rules
Somebody notices it	20 min – 4 hrs	Nobody owns the inbox; it competes with everything else
Decides whose lead it is	10 min – 1 day	No routing rule, so it is settled on WhatsApp
Entered into the CRM	1 hr – never	Manual re-typing, often skipped
First contact attempt	2 hrs – 2 days
Out of hours, or weekend	+ 12–60 hrs	No rule for what happens at 9pm on Saturday

Three things in that table are worth noticing.

No single step is anyone's fault. Every delay is small and defensible. The sum is two days.

The longest delays are decisions, not work. "Whose lead is this" and "has anyone picked this up" are the two biggest contributors, and neither requires any skill to resolve — only a rule. This is why routing beats everything else you could spend the money on.

The CRM entry step is where the data dies. When leads are typed in later, or not at all, you lose the ability to measure response time, which means you cannot tell whether anything you do next helps.

Before changing anything, measure the one number: the median minutes from form submission to first genuine human contact attempt, split by weekday and weekend. Most teams guess 30 minutes and discover four hours. If you cannot produce that number today, that is the first thing to build.

What to automate, and in what order

Roughly this order, because each step is worth more than the one after it.

1. Instant acknowledgement, with substance. An auto-reply within seconds that confirms receipt, names the person who will call, and says when. This is not the response — it is a holding action that buys you the next hour. It does nothing if it is a generic "we have received your enquiry"; it works when it is specific.

2. Routing with a named owner. Every lead gets assigned to one person by a rule — territory, service, round-robin, whatever fits — at the moment of submission. Shared responsibility is the single largest source of delay, and this removes it. On Workato's numbers this is also the only intervention with a directly observed effect: roughly 3.5 hours against 13.

3. Notification that actually interrupts. Email does not. A WhatsApp or push notification to the assigned person, with the lead's name, number and question in the notification body, so they can act without opening anything. Most "slow response" problems are notification problems wearing a motivation costume.

4. Escalation on a timer. If the owner has not logged an attempt in N minutes, reassign and notify someone else. Without this, routing just moves the bottleneck onto one person's bad day.

5. An explicit out-of-hours rule. Decide what happens to a 10pm Sunday enquiry: acknowledged immediately, queued to the top of Monday's list, or routed to whoever is on. Any of those beats the default, which is that it surfaces on Tuesday.

6. Capture into the CRM automatically. Partly so nobody re-types anything, mostly so response time becomes a number you can see weekly.

Notice what is not on the list: better email copy, a longer nurture sequence, a chatbot that tries to answer product questions. Those are improvements to the part of the funnel that is not broken.

None of this is a large build. It is webhooks, a routing rule, a notification channel and a write into whatever system of record you have — days of work, not months, and it generally sits inside the range we publish for marketing automation work. It gets more involved when the routing logic is genuinely complex or has to reflect a process nobody has written down, which is the point at which it becomes a CRM question rather than an automation one. For Students Traffic, the CRM was one of five core systems built across a three-year partnership, and routing logic of this kind belongs in that layer rather than bolted on beside it — mostly because it changes as the business does.

What does not work

SLAs without notification. "Respond within 15 minutes" written in a document changes nothing. The timer has to be enforced by something that reassigns the lead.

Chatbots that hold the conversation hostage. A bot that qualifies in three questions and hands over fast is useful. A bot that tries to answer everything adds a stall between the enquiry and a human, which is precisely the cost you were trying to remove.

More follow-up, later. The 2007 study's most uncomfortable finding is that past about 20 hours, additional dials actively reduced the chance of making contact to qualify. Treat the number sceptically; treat the direction seriously. Attempt cadence is not a substitute for being first, and a team making its fifth call on day four is usually doing it because nobody made the first on minute four.

Buying speed you cannot staff. Cutting response time raises contact volume, and a team already at capacity will simply be fast and then unavailable. Check the second-order effect before launching.

Optimising speed when the real problem is lead quality. If most enquiries are unqualifiable, responding in two minutes gets you to the same place faster. Fix the source first.

When this does not apply

If your sale is a long considered purchase with a procurement process — enterprise software, capital equipment, anything with a tender — speed of first response matters much less than being credible by week three. The research above is about web-generated leads to inside sales teams, and it does not transfer to deals that take nine months.

If your enquiry volume is a handful a week and one person owns all of them on their phone, you are already fast. Build nothing; spend the money on getting more enquiries.

And if your buyers have explicitly chosen a time — they booked a slot, they asked to be called Thursday — then responding in four minutes is not a win, it is an intrusion. Speed is a proxy for availability, not a virtue in itself.

What remains true for most Indian SMBs with a website form and two to ten people selling: the first hour is where the leads go, nobody currently knows their own median response time, and the fix is a routing rule and a notification rather than anything clever.

Send us your current numbers — even guessed ones — and we will tell you whether this is a plumbing problem worth a week of work or whether you have a lead-quality problem that speed will not touch. Those two get confused constantly, and only one of them is worth paying to fix.