The five-minute lead response rule: what the research actually says

Last updated: 2026-07-30

Almost every article on lead response time cites the same two numbers: respond within five minutes and you are 100 times more likely to reach the lead and 21 timesmore likely to qualify them. Most attribute it to MIT. Some attribute it to Harvard Business Review. A few just say “studies show.”

The numbers are real and the finding has held up. The attribution is usually wrong, and the study is weaker evidence than the confident citation implies. Both things are worth knowing before you build a process around it.

Where the numbers actually come from

The source is research led by Dr James Oldroyd, published around 2007 in partnership with InsideSales.com, a sales software vendor. It analysed roughly 15,000 leads and over 100,000 call attempts across a small number of companies that generate and respond to web leads, over a three-year period.

The headline findings:

• Contacting a lead within five minutes rather than thirty made contact roughly 100 times more likely.
• The same comparison made qualifying the lead roughly 21 times more likely.
• Odds of qualification dropped sharply through the first hour, then flattened out at a much lower level.

Oldroyd was affiliated with MIT’s Sloan School at the time, which is how a vendor-sponsored study acquired the durable nickname “the MIT study.” A separate, later piece of research published in Harvard Business Review in 2011 examined response times across thousands of US firms and found most were slow — that is a different study, with different numbers, and it is not where the 21x comes from.

What is weak about it

We are in the business of selling faster response, so it would be convenient to leave the citation unexamined. It does not survive examination cleanly:

• It is vendor data. The dataset is drawn from the customer base of a company selling software that makes sales teams respond faster. That is not fraud, but it is not disinterested either.
• It is observational, not a controlled trial. Nobody randomly assigned leads to a fast group and a slow group. Which means the causal claim is not established by this data.
• Reverse causation is plausible and rarely addressed. The leads a rep reaches in five minutes may differ systematically from the ones that sit for an hour — higher-intent leads arriving during staffed hours, prioritised by the rep precisely because they looked promising. Some of the 21x is very likely selection, not speed.
• It is nearly two decades old. It predates smartphones being universal, predates WhatsApp entirely, and describes a market where the alternative to a fast call was a slow call, not an instant automated reply.

What survives

The direction is not seriously in dispute, even if the magnitude is inflated. It has been reproduced repeatedly across two decades by parties with different incentives, and the underlying mechanism is not mysterious.

A buyer submitting an enquiry is usually submitting several. The seller who replies first is the one who gets to frame the conversation, and the ones who reply later arrive to a buyer who is already in a conversation with someone else. This does not need a 21x effect to matter; it needs only that buyers contact more than one seller, which in property they reliably do.

The honest version of the claim is therefore: fast response is a large advantage of uncertain size, and the cost of acting on it is low enough that the uncertainty does not change what you should do. That is a weaker statement than the one in the infographics, and it is still sufficient.

The part that is usually left out

The five-minute framing quietly assumes the enquiry arrives during working hours. For property and interiors in India, a large share does not. Enquiries land after site visits, on the commute home, and late at night when someone is browsing listings in bed.

For a lead that arrives at 11pm, the realistic human response time is not five minutes or thirty — it is nine hours, because it is whenever someone opens the CRM the next morning. The gap the research describes is not mainly a gap between fast teams and slow teams. It is a gap between staffed hours and unstaffed hours, and no amount of rep discipline closes it.

That is the actual argument for automating the first reply, and it is a narrower argument than “21x.” An automated first response does not need to close the sale or outperform a good salesperson. It needs to hold the buyer’s attention — acknowledge the enquiry, answer the two or three factual questions they will ask about price, location and availability, and secure a time — so that a human is not starting from cold nine hours later against a competitor who answered at 11:02pm.

What to measure instead

If you are evaluating whether faster response is worth anything in your business, the 21x figure is the wrong thing to plan against. Three numbers from your own data are worth more than any published benchmark:

• Your median first-response time, split by hour of arrival. The average hides the problem. The split shows you whether you have a speed problem or a coverage problem, and those have different fixes.
• The share of enquiries arriving outside staffed hours. This is the ceiling on what automating the first reply can win you.
• Contact rate by response bucket— under 5 minutes, under 1 hour, same day, later — using leads you did not choose to prioritise. The last part is what keeps you from reproducing the original study’s selection problem in your own numbers.

Sources

• Oldroyd, J., with InsideSales.com, Lead Response Management Study(c. 2007) — the origin of the 100x contact and 21x qualification figures; ~15,000 leads and 100,000+ call attempts across a small set of companies. Frequently cited as “the MIT study” on the basis of the author’s MIT Sloan affiliation.
• Harvard Business Review(2011), on US firms’ lead response times — a separate study, commonly conflated with the above.

Both studies are observational and industry-sponsored. Treat the magnitudes as indicative rather than precise, and prefer your own contact-rate data where you have it.

Related: what the WhatsApp Business Platform actually costs in India — where response speed also turns out to drive the bill.