Five minutes versus thirty. The research is older than you think
Almost every article about lead response time cites the wrong source. Here is where the numbers come from, what they do and do not prove, and how to measure the one figure that matters in your own shop.
Where the famous numbers come from
The statistic you have seen — contact a lead within five minutes and you are 100 times more likely to reach them, and 21 times more likely to qualify them, than if you wait thirty minutes — comes from research by Dr James Oldroyd with MIT and InsideSales.com, published in 2007.
It is almost always attributed to Harvard Business Review. That is wrong, and the confusion matters because the HBR work is a different study with different findings: a 2011 analysis of 2,241 companies that found an average first-response time of 42 hours, and roughly a 7× improvement in qualification odds when responding within an hour.
So there are two findings, not one. If you are going to quote a number at your team, quote it correctly — someone will check.
What the research does not prove
Two honest caveats, because they change how you should act on it.
It is B2B inbound data, not restoration emergency data. Nobody has published an equivalent study on 2 a.m. water losses. The direction of the effect is almost certainly stronger in an emergency vertical — a homeowner standing in water has less patience than a software buyer — but the specific multipliers are not yours to claim.
Correlation is doing some of the work. Firms that answer in five minutes tend to be better run in other ways too. Speed is not the only variable that moved.
What survives both caveats: the ordering is not in dispute. Faster is better, the drop-off is steep and early, and the cost of a slow pickup is paid in jobs you never hear about.
The number to measure in your own shop
Forget benchmarks. Measure these four for two weeks and you will know more about your funnel than any industry report can tell you.
- Time to first human contact, by source and by hour of day. Not time to auto-reply — time to a person. Median, not average: one 6-hour outlier ruins a mean.
- Missed-call rate after hours. Pull the call log. Count rings that ended without a human. This is usually the number that shocks people.
- Contact rate by response bucket — under 5 minutes, 5 to 30, 30+. Your own version of the study, on your own leads.
- Booked rate within each bucket. Contact is not the goal. A booked job is.
If your median after-hours response is measured in hours, you do not have a marketing problem yet. Buying more leads on top of a slow pickup is paying retail to widen a leak.
What to fix first, in order
- Escalation on missed calls. Ring the tech, then the lead tech, then the owner. A single voicemail box is where mitigation jobs go to die.
- Missed-call text-back, automatic, within seconds. Cheapest fix on this list. It converts a missed ring into a live conversation surprisingly often.
- An answering service that can dispatch, not just take a message. A message taken at 2 a.m. and read at 8 a.m. is a lost job with extra steps.
- Only then, more demand. Fix the pickup, then buy the traffic.