Speed to Lead: The 5-Minute Rule Is 19 Years Old

Someone fills in your form at 21:40. You see it at 08:15. Every guide you have ever read says you have already lost the deal — and every one of them is citing the same study from 2007.

Speed to lead is the most repeated number in sales. Reply within five minutes and you are 21 times more likely to qualify the lead. It is on vendor landing pages, in SDR onboarding decks, in LinkedIn posts dated last week. Almost nobody who repeats it has read where it came from, and the few who check find something uncomfortable: the figure is real, it is old, it is about phone calls, and it was never a measurement of whether you win.

This is not an argument for being slow. Responding fast genuinely matters, and the 2024 data on this is more damning than the 2007 data. But if you are a founder running outbound alone, you cannot build an SLA around a number you do not understand. So here is the provenance, the parts the benchmark pages leave out, and the version of the rule that survives contact with a one-person sales team.

Where the speed to lead number actually comes from

The 21x figure traces to a single study: The Short Life of Online Leads, presented by Dave Elkington of InsideSales.com and Dr. James Oldroyd, then a faculty fellow at MIT Sloan, at MarketingSherpa's B2B Demand Generation Summit on 16 October 2007.

Read the deck and the shape of it becomes clear. Six companies. Over 15,000 leads and over 100,000 call attempts across three years. The outcome measured was whether a rep reached a live person for two to six minutes and whether that call qualified. Calling at five minutes versus thirty minutes was associated with roughly 100x the odds of contact and 21x the odds of qualification.

Four things about that are worth holding onto:

• These are odds ratios from a statistical model whose specification, p-values and confidence intervals the deck does not publish.
• The mechanism is the telephone. Every data point is a dial. Nothing in it describes email, and nothing in it could describe messaging, because WhatsApp did not exist.
• The study explicitly did not address close ratios. Qualification is not revenue.
• It was published by a company selling software that solved the problem it identified. That does not make it wrong. It does mean nobody independent has re-run it.

The companion citation, the 2011 Harvard Business Review article by Oldroyd, Kristina McElheran and Elkington, is the source of the "42 hours average" and "23% never respond" figures. I could open the article's opening section but not its data behind HBR's paywall this session, so I am not going to quote numbers I have not read. Treat anything you see attributed to it as second-hand until you buy the reprint.

The 2026 speed to lead "benchmarks" are mostly 2007 in a new jacket

Search the keyword today and you get a wall of pages titled with this year. I opened two of the better-ranking ones.

One is published as a 2026 Speed to Lead Benchmark Study. Every figure in it is from 2007, 2011 or 2017. It says so, to its credit, and links what it can. It contains no 2026 data at all.

The other reports a 939-company dataset, Q2 2025 to Q1 2026, with close rates by response tier — the most current thing I could find. It is also vendor CRM data mixed with "industry reports" without saying which number came from where, it attributes the 21x claim to "research shows" with no study named, and its own table does not divide: it lists a 24% close rate against a 12% close rate as 1.9x. That is 2.0x. A rounding error is not fraud, but it tells you how much checking went into the page you are about to base an SLA on.

What the most recent checkable test found, and it is not about speed

In March 2024, RevenueHero submitted demo requests to 1,000 B2B SaaS company websites and timed the replies. The headline is not the five-minute rule.

635 of the 1,000 never replied at all.

Of the 365 that did, the average was one day, five hours and seventeen minutes. 172 answered in about two minutes. 109 took half a day. Twelve took almost a month.

The automation split is the part worth printing out. Companies that handled the first response manually averaged two days, three hours. Companies that automated at least the first touch averaged seventeen hours, twenty minutes — about three times faster. And response time tracked with how much the company was hiding: firms that published pricing on all plans replied in two hours, fourteen minutes; firms that made you request a quote took over a day.

So the real distribution is not fast versus slow. It is answered versus ignored, and almost two thirds of a thousand funded software companies were in the second group. If you reply the same day, you are already in the top third of your market. That is a far more useful target for one person than five minutes.

The thing every speed to lead guide skips: which channel you reply on

Here is the gap in the whole literature. All of the foundational data measures a phone call. None of the benchmark pages break response time down by channel at all — one of the two I read mentions channel exactly once, to say email alerts are too slow for reps.

But in 2026 the channel is most of the effect. Five minutes is not a single thing:

• A five-minute email lands in an inbox that is structurally saturated, behind spam filtering and AI triage. Speed does not buy you attention there; it buys you a better position in a queue nobody is reading. We have written at length about why cold email reply rates collapsed.
• A five-minute phone call is what the 2007 study measured, and it still works — The Bridge Group's 2025 survey of 351 B2B companies found phone-centric teams averaging 4.6 quality conversations a day against 3.4 for email-centric ones. It also requires you to be free to dial.
• A five-minute WhatsApp reply arrives as a notification on a device in someone's hand, in a thread they can answer with one thumb. That is a different mechanism entirely, and it is why WhatsApp B2B outreach behaves unlike email in every market where the phone is the primary computer.

There is a 2026 wrinkle on that last one, and it is a cost, not a feature. When someone messages you first, WhatsApp opens a 24-hour customer service window in which you can send free-form replies. Meta's own non-template pricing documentation now states that any non-template message is charged as of 1 October 2026 — service messages, which previously were free inside that window, are billed per message, and AI replies routed through Meta's own Business Agent platform are billed per token from 1 August 2026. Replying instantly on WhatsApp is still cheap. It is no longer free, and an always-on auto-responder that fires on every inbound now has a line item.

The solo operator version of the rule

If you are the founder and the whole sales team, the five-minute SLA is not a target, it is a fantasy. The Bridge Group's 2025 numbers put a full-time SDR at a median 112 activities a day — 44 phone, 41 email, 19 LinkedIn, 8 other — to produce 4.1 quality conversations. You are doing that between a support ticket and a standup.

So commit to what you can hold:

1. Pick one channel you can answer in minutes. For most of our readers that is WhatsApp, because the notification reaches you. Put the form's "best way to reach you" field there.
2. Automate the first touch, not the conversation. RevenueHero's data says a templated acknowledgement inside two minutes beats a thoughtful manual reply a day and a half later. Say what happens next and when.
3. Set a same-day floor, not a five-minute one. Two thirds of your competitors never reply. Beat that first.
4. Protect two windows a day for live replies. The 2007 study's own timing data put late afternoon and 08:00–09:00 as the best hours for qualification. Use them.
5. Route by intent, not arrival. A pricing-page visitor who filled a three-field form is worth a dial. A gated-whitepaper download is worth a message. Both get the same SLA in most tools, which is why most tools waste your afternoon.

Point 2 is the one with real leverage, and it is the same persistence argument behind sales follow-up automation: machines are better than humans at the touches that are boring and timed.

Where speed stops helping

Three honest limits, because the vendors in this category will not give you them.

Qualification is not closing. The 2007 study says in its own text that it did not look at close ratios, and the 2026 pages that convert its odds ratios into revenue claims are inventing that bridge. Faster replies get you to a meeting. They do not make the meeting good — which is a problem we have written about separately in the context of AI SDRs and meeting quality.

Speed amplifies whatever your qualification is. If your form attracts the wrong people, a five-minute SLA means you now meet the wrong people faster and with more of your calendar gone.

And an instant reply that is obviously a robot can cost you the lead the slow reply would have kept. The fast message has to read like a person who actually looked.

Where BOSRAI fits, and where it does not

BOSRAI handles the mechanical half of this. It watches for the inbound reply across email, WhatsApp and LinkedIn, drafts the first response in the prospect's language, and queues the follow-up sequence so touch three and touch seven actually happen. Because outreach runs on a human-in-the-loop approval model, a draft waits for you rather than firing blind — which is slower than a pure auto-responder by design, and the tradeoff we argue for in human-in-the-loop AI sales.

What it does not do: hit a five-minute SLA while you sleep if you want to read every message before it sends. You pick one. BOSRAI's own pricing runs Free, then $79.99, $199, $499 and $999 a month, discounted annually — see pricing. And we have no published case studies or response-time benchmarks of our own, so nothing above is a claim about our results. It is a reading of other people's data, with the sources listed so you can disagree with us.

If you take one thing: stop quoting the 21x. Go and answer the 635.

Sources

• The Short Life of Online Leads (Oldroyd / InsideSales, 2007) — the original deck behind the 21x and 100x odds ratios; 6 companies, 15,000+ leads, 100,000+ call attempts, no close-rate analysis.
• The Short Life of Online Sales Leads, Harvard Business Review, March 2011 — author attribution for the companion study; data behind the paywall and not quoted here.
• RevenueHero B2B lead response test, March 2024 — 1,000 B2B SaaS demo requests; 635 non-responses, automation and pricing-transparency splits.
• The Bridge Group 2025 SDR Models and Metrics report — 351 B2B companies; 112 daily activities, 4.1 quality conversations, 4.6 vs 3.4 phone versus email.
• Meta WhatsApp non-template message pricing documentation — the 24-hour customer service window, per-message charging for service messages from 1 October 2026, per-token Business Agent charging from 1 August 2026.
• Artemis GTM "2026 Speed to Lead Benchmark Study" — a 2026-titled page whose figures are all from 2007, 2011 and 2017.
• Optifai lead response time benchmark — the 939-company Q2 2025–Q1 2026 vendor dataset, including the 24% vs 12% close rates it labels 1.9x.
• Lead response time statistics roundup (Ainora) — where the surviving links to the 2007 deck and the archived Drift and Velocify studies were traced.