How fast should you respond to an inbound lead?

Answer high-intent inbound within five minutes, and treat one hour as the floor for everything else. Those two thresholds come from the two studies most often cited on this question, and they describe different parts of the same curve: the drop between five and thirty minutes is its steepest section, and the one-hour mark is where responders separate clearly from stragglers. Past 24 hours, replying is close to not replying.

That is the short answer. The rest of this is the detail behind it — what the evidence does and does not support, which leads justify which target, and what to do when covering every hour is not realistic.

What does the research actually say?

Two studies do most of the work in this genre. Both are worth reading directly rather than through the posts that cite them, because the numbers get mangled in transit.

Harvard Business Review, March 2011. James Oldroyd, Kristina McElheran and David Elkington audited how long 2,241 US companies took to respond to a web-generated test lead, published as The Short Life of Online Sales Leads. Firms attempting contact within an hour were nearly seven times as likely to qualify the lead — a meaningful conversation with a key decision maker — as those attempting it an hour later, and more than 60 times as likely as those who waited a day or more.

The Lead Response Management study. Oldroyd again, with InsideSales.com, covering more than 15,000 leads and over 100,000 call attempts across six companies (study summary). It measured the first half-hour specifically: the odds of contacting a lead fell roughly 100-fold between a five-minute and a thirty-minute response, and the odds of qualifying one fell about 21-fold across the same window.

The useful conclusion is not any single number. It is that the penalty for waiting is front-loaded: minutes early on the curve are worth more than hours later on it. Which is what makes a flat, one-size target the wrong instrument.

Is the target five minutes, or an hour?

Both, applied to different signals.

Five minutes is a genuine competitive edge and an expensive promise. Guaranteeing it across every inbound channel, every hour of the week, means either continuous staffing or automation you trust unsupervised. Most small teams can afford neither, and trying to buy it everywhere is how the target quietly gets abandoned.

One hour is achievable during covered hours by an ordinary team with a clear owner, and it captures most of the documented benefit.

So the workable policy is a tier, not a threshold: five minutes for the signals that justify it, one hour for everything else, same day as the absolute back-stop. Choosing which signals go in which tier is the actual decision.

Which leads deserve a five-minute response?

The ones where intent is unambiguous and the buyer is in-market right now:

• A demo or trial request from your target segment. The highest-intent action anyone takes on your site.
• A pricing enquiry. Someone building a business case, usually with a shortlist open.
• A reply on a live thread. They are at their desk, in the conversation, waiting. This is the cheapest fast reply you will ever send and the one most often missed.
• An inbound call you did not answer. They tried the highest-effort channel available. Ring back before they try a competitor.

And the ones that do not, where a fast reply mostly signals eagerness:

• Content downloads and newsletter signups. Interest in a topic, not in a purchase.
• Event and webinar registrations. The intent is the event.
• Enquiries clearly outside what you sell. Answer honestly, unhurriedly, and point them somewhere useful.

Most of the value of a tiering exercise is the argument it forces about what actually counts as intent in your business. Have it once, write it down, then instrument it.

Does the clock run outside business hours?

Yes. The prospect does not experience your operating calendar.

A lead arriving Friday at 5:10pm and answered Monday at 9:15am is a 64-hour response. Many CRMs will report it as fifteen minutes, because they measure elapsed business hours. That setting is the single most common reason a team's reported response time bears no relation to what buyers experience.

This matters more than it first appears, because inbound does not arrive evenly. Evenings and weekends are when people research vendors without interruption. If those hours are excluded from your measurement, your worst response times are also your least visible ones.

Set the clock to wall time. If the resulting number is uncomfortable, that discomfort is the finding.

What counts as a response?

Something the prospect could act on. A reply that answers what they asked, or asks a question that moves the conversation forward.

An autoresponder is not a response. "Thanks for getting in touch, someone will be with you shortly" is a receipt — it confirms delivery and provides no information. Plenty of tooling stamps the first outbound message on a thread as the first response, which means a well-configured autoresponder can make a four-day reply look instantaneous.

Check what your system counts. If it counts the autoresponder, your response-time metric is measuring your mail server.

What if you cannot hit these numbers?

Most teams cannot hit them everywhere, and pretending otherwise is how targets get quietly dropped. Three things that work better than a stretch goal nobody owns:

Cover the windows that produce the damage. Plot response time by arrival hour and weekday. Nearly every team finds one or two windows — Friday afternoons, Monday backlog, weekday evenings — that generate most of the worst cases. Fixing four hours of coverage beats a general instruction to be faster.

Narrow the promise instead of weakening it. Five minutes for demo requests during covered hours, held reliably, is worth more than a five-minute target across all channels that holds about a third of the time.

Decide the escalation rule in advance. What happens to a lead nobody has answered in four hours? If the answer is "it depends who notices", you have found the source of your multi-day tail.

How do you know whether you are hitting it?

Report the distribution, never the average. The percentage answered within an hour, plus the size of your worst decile, tells you something a mean cannot: one four-minute reply and one four-day reply average out to two days and describe neither.

Count the leads that were never answered at all. They are excluded from most response-time reports precisely because they have no response timestamp, which means the metric systematically omits its own worst cases.

Where Monty fits

Monty is software that watches the inbox after a lead replies and drafts the response for a human to approve.

Everything above says response time is not a typing-speed problem. It is a context problem wearing a stopwatch. Monty is built on that reading, which is why it makes no attempt to help anyone type faster. It removes the two steps where the hours actually go: something watches the channel during the hours you are not, and the thread history, the previous promise and today's facts are assembled at the moment the draft is written instead of hunted across three tabs.

If you are uneasy about software emailing your prospects in your name, you should be — a fast generic reply costs more than a slow specific one, and that is the failure mode this whole post warns about. So you decide how much it does. The autonomy tiers are a ramp, not a menu:

• Manual. Sending is locked. Monty organises the inbox; you write and send everything yourself.
• Assisted. Every AI-drafted reply and follow-up waits in a review queue. Nothing reaches a prospect until you approve it, and you choose which situations always need your nod.
• Automated. Once you have read enough drafts to recognise the pattern, Monty sends routine replies inside rules you set and holds anything sensitive for approval. Available on the Professional plan.

You never have to trust it before you have watched it work. The inbox and review pages show what each tier looks like in practice.

What it does not fix: a business-hours clock, an unowned escalation rule, or a tiering decision nobody has made. Those cost nothing, they take an afternoon, and they are what tell you whether anything else you try is working. Do them first — genuinely, before you buy anything, ours included.

**Connect your inbox and read the first draft it writes for you.** Monty is in early access, so every team gets 1-on-1 onboarding from the people building it.

The full argument, including where response time actually goes, is in The Speed-to-Lead Handbook. More in the speed-to-lead collection.