Speed to Lead: What Happens to Conversion Between Minute 1 and Minute 30

Speed to lead is the time between a prospect raising their hand and your business making first contact. The research on it is blunter than the phrase suggests: the odds of qualifying a lead fall away inside the first half hour, not over the first few days.

What almost no one publishes is a clean minute-by-minute curve. The studies everyone quotes measure brackets, not minutes, so the honest version of the story is a set of thresholds.

This post lays out what the research actually says between minute 1 and minute 30, where the delay usually comes from, and what changes when the first response is a phone call that gets answered in seconds.

One thread runs through all of it: whoever owns the response loop owns the number. Speed to lead does not improve because a vendor tuned something last quarter. It improves because the team that hears the failed call can change the script, the routing and the retry cadence that week.

TL;DR

• Speed to lead measures the gap between an inbound inquiry and your first meaningful response, on any channel.
• The Harvard Business Review study behind most speed-to-lead claims found that contacting a lead within an hour made a firm nearly 7 times more likely to qualify it than waiting one more hour, and more than 60 times more likely than waiting a day.
• A separate 2007 study led by the same researcher for InsideSales.com, covering three years of data at six companies, more than 15,000 leads and over 100,000 call attempts, found the odds of qualifying a lead were 21 times greater at 5 minutes than at 30 minutes. That one is vendor platform data analyzed by an academic, not a peer-reviewed paper, and it is often mislabeled as an MIT study.
• Nobody has published a per-minute curve, so treat 5 minutes and 30 minutes as thresholds rather than points on a line.
• Most of the delay is structural: routing rules, business hours, and the gap between a form submission and a human noticing it.
• An AI voice agent that calls the lead back in seconds removes the queue, then hands the qualified ones to a rep.

What is speed to lead?

Speed to lead is the elapsed time between a prospect becoming a lead and your team making first contact with them.

The clock starts when the prospect acts: submits a form, requests a quote, books a demo, or calls and does not get through.

It stops at the first meaningful response, which means a conversation or an attempt at one. An automated confirmation email does not stop the clock, because nothing in the sales process has moved.

The metric is channel-agnostic. A call, a text, and an email all count, which is why teams that measure only email response time usually report a number better than the one their prospects experience.

What happens to conversion between minute 1 and minute 30?

Conversion odds drop sharply across that half hour, and the sharpest documented fall sits between 5 and 30 minutes.

The foundational work here is the 2011 Harvard Business Review study The Short Life of Online Sales Leads by James Oldroyd, Kristina McElheran and David Elkington. It audited 2,241 US companies by submitting a test lead and timing the reply.

Their companion analysis of roughly 1.25 million leads found that firms contacting a prospect within an hour of the inquiry were nearly 7 times more likely to have a qualifying conversation with a decision maker than firms that waited one additional hour, and more than 60 times more likely than firms that waited 24 hours or longer.

A second, earlier body of work puts a finer point on the window inside the first half hour. The 2007 Lead Response Management research, led by Oldroyd for InsideSales.com across three years of data at six companies, more than 15,000 leads and over 100,000 call attempts, found the odds of qualifying a lead were 21 times greater when contact happened at 5 minutes rather than at 30 minutes.

Know where that number comes from before you quote it. It is vendor platform data analyzed by an academic, it was presented at a 2007 summit rather than published in a journal, and it is widely miscredited to MIT. Treat it as directional and treat the 2011 HBR paper as the load-bearing citation.

Set the 2011 findings against that same study's audit of actual behavior and the gap is the whole problem.

So the research supports three claims and not a fourth.

• Supported: the 5-minute threshold matters, the 30-minute threshold matters, and a day-long delay is close to a write-off.
• Supported: a large share of inbound leads never get a response, which makes coverage a bigger lever than polish.
• Supported: the average firm is an order of magnitude slower than the threshold that predicts conversion.
• Not supported: any specific per-minute decay figure. If you see a chart showing conversion at minute 7 versus minute 8, it was modeled, not measured.

Why the five-minute rule is a ceiling, not a target

Five minutes is the point at which the research stops being kind to you, which makes it a poor goal.

A team that targets five minutes builds for five minutes: an alert, a queue, a rep who checks between calls. That design produces a median inside the window and a long tail well outside it, because the tail is made of nights, weekends, and busy hours.

The tail is where the money is. A lead that arrives at 7pm on a Friday and gets a call on Monday morning has crossed every threshold in the study.

The other problem with a five-minute target is that your competitors share it. When three vendors call inside the window, order of arrival decides who has the conversation, and the prospect explains their problem once.

Where the delay actually comes from

Almost none of the delay is a rep being slow. It is structural, and each piece is separately fixable.

1. Coverage gaps. Leads arrive around the clock and staffing does not, so anything outside business hours waits for the next shift.
2. Routing rules. Territory, round robin, and product-line assignment each add a hop, and every hop adds a chance the lead sits with someone who is busy or away.
3. Channel mismatch. The prospect submitted a web form and the first response is an email, which starts a slow asynchronous thread instead of a conversation.
4. Notification decay. An alert lands in a feed the rep checks between meetings, so the clock runs while nobody is aware the lead exists.
5. Volume spikes. A campaign, a media mention, or a seasonal peak produces more leads in an hour than the team can call in a day.
6. Unreached callbacks. The first attempt often fails to connect, and the retry cadence, not the first call, decides whether contact ever happens.

Read that list and the pattern is clear. Speed to lead is a capacity and coverage problem wearing the costume of a discipline problem.

What changes when a call is answered in seconds

The queue disappears, so the thresholds stop mattering.

This is where an AI voice agent fits. Retell is a Customer Experience AI Platform for Autonomous Customer Relations. You build, test and run the voice agents yourself, so an inbound inquiry can be answered live and an inbound form can trigger an outbound call within seconds of submission. The specific job is lead qualification: confirm who the prospect is, what they need, budget or timeline if that is how you qualify, and whether they are worth a rep's hour.

Three things follow from removing the wait.

• Every lead gets a first attempt. Nights and weekends included. The 23% of companies that never responded at all in the 2011 audit is a coverage number, and an alert sent to someone who is asleep does not change it.
• Qualified leads reach a person while the intent is live. The agent can transfer the call to a rep with the context already gathered, or book the appointment straight into the calendar when no rep is free.
• Spikes stop being a capacity ceiling. Concurrent calls are a setting, not a hiring plan, so a campaign that triples volume does not triple response time.

Outbound callbacks have their own failure mode, which is that people do not answer unknown numbers. Branded call ID and verified phone numbers address the pickup rate rather than the dial rate, and they matter more than the speed of the first attempt if nobody picks up.

Boatzon's agent "Anna" runs this pattern in production. It covers outbound financing follow-up, inbound application support and SMS, and it reaches applicants in about one minute.

For recruiting at volume, high volume hiring covers screening 500 applicants a week by phone.

Two honest caveats. An agent that reaches a prospect quickly and handles the conversation badly has not improved anything, so conversation quality is the constraint, not the speed of the dial. And results vary with how the qualification script is written and how often it gets revised.

How to measure speed to lead without flattering yourself

Measure the distribution, not the average, and measure from the prospect's action rather than from the moment the lead appeared in your CRM.

• Use the median and the 90th percentile. An average of 11 minutes can hide a quarter of leads waiting overnight.
• Start the clock at submission. Integration lag between a form, an automation tool, and the CRM is invisible in CRM-based reporting and can be minutes on its own.
• Count attempts, not touches. An automated acknowledgment is not a response. A dial that rings out is an attempt, and it belongs in the count.
• Segment by arrival hour. If in-hours and out-of-hours medians differ by more than a few minutes, the problem is coverage.
• Track connect rate alongside speed. Speed with a low pickup rate produces fast failures, which look like progress on the dashboard.

If calls are part of the motion, the post-call analysis on each call is where the follow-on questions get answered: what the prospect asked, where the conversation stalled, and which qualified leads never reached a rep.

How to cut speed to lead this quarter

Pick the structural fixes in order of how much delay they remove per hour of work.

1. Instrument the real number first. Median, 90th percentile, and in-hours versus out-of-hours, measured from submission.
2. Close the out-of-hours gap. Leads that arrive outside staffed hours wait for the next shift no matter how urgent the team feels, so this one needs coverage rather than urgency.
3. Collapse the routing hops. Assign after qualification rather than before, so no lead waits on a rule.
4. Make the first response a call. Voice gets a decision in one exchange where email starts a thread.
5. Set a retry cadence and run it. Several attempts across the first day, then tapering, beats one fast attempt.
6. Review recordings weekly. The script that qualified well in week one usually needs changes by week three, so whoever owns the number should be able to make those changes without filing a ticket.

That last point is worth being explicit about. The team closest to the prospect should be able to change the agent's questions, its handoff rules and its retry logic directly, the same afternoon they hear the problem on a call. Waiting on a vendor to edit a script is its own kind of response-time failure.

Unlike managed AI vendors and BPOs, a change does not become a ticket, a queue, or another SOW. That is the difference that decides whether the number you instrument in month one is still true in month six.

For law firms specifically, personal injury intake covers what has to happen in the first 90 seconds of the call.

Frequently asked questions

What is a good speed to lead?

Under five minutes is the widely used benchmark, because the research shows a steep fall between 5 and 30 minutes. Teams with round-the-clock inbound should aim for a first attempt in seconds, since the median matters less than the leads that arrive when nobody is working.

Is speed to lead the same as lead response time?

They are used interchangeably. Both measure the gap between an inbound inquiry and first contact. Speed to lead is the more common term in sales operations, lead response time in marketing reporting.

How do you calculate speed to lead?

Take the time of each prospect's action, subtract it from the time of your first contact attempt, and report the median and 90th percentile across all leads in the period. Averages hide the overnight tail.

Does responding faster actually increase conversion?

The published research shows a strong association, not a controlled causal effect. Faster contact is linked to far higher odds of qualifying a lead, but the leads you reach quickly may also be the ones with higher intent. Speed is a lever worth pulling, and it is not the only one.

Can an AI voice agent qualify leads on its own?

It can run the qualifying conversation, capture the answers, and route or book the outcome. Anything needing judgment, negotiation, or authority should go to a person, which is why the transfer path matters as much as the script.

What breaks speed to lead most often?

Out-of-hours arrivals and routing hops. Both are structural, so they persist through any amount of rep coaching.

Answer every lead in seconds, not shifts.

Build a Retell voice agent that calls new leads back the moment they arrive, qualifies them on the call, and transfers the ones worth a rep's time. Run it on your own inbound before you sign anything, and measure it against the median and 90th percentile you instrumented in step one. Start a pilot on your own calls.