Speed to Lead: how fast you actually have to reply, and which numbers are folklore

Speed to lead is how long your business takes to make first contact after someone submits an enquiry. The evidence for why it matters is real but narrower than the internet suggests. Two studies carry almost all of it, both are United States data, and the single most quoted figure in the category has no traceable source at all.

That last sentence is the reason this article exists. If you have read anything about response times in the last five years you have met four numbers. We went and found the studies behind them, read the primary documents, and one of the four does not have one.

The four numbers you have already seen

Here they are as they usually appear, stripped of the blog posts they usually appear in:

1. The average business takes 42 hours to reply to a web enquiry.
2. Replying in 5 minutes instead of 30 makes you 21 times more likely to qualify the lead.
3. 78 percent of buyers purchase from whoever responds first.
4. 77 percent of leads never get a response at all.

Three of those trace to real research. One does not. And of the three that do, two are routinely restated in a way the source does not support.

Number one: the 42 hours is real, and it is not the headline

This comes from The Short Life of Online Sales Leads, by James B. Oldroyd, Kristina McElheran and David Elkington, published in Harvard Business Review in March 2011. The researchers audited 2,241 United States companies by submitting a test enquiry to each one and timing the reply.

The full distribution is more useful than the average, and almost nobody quotes it:

Time to first response	Share of the 2,241 companies
Within one hour	37 percent
One to 24 hours	16 percent
More than 24 hours	24 percent
Never responded	23 percent

The 42 hours is the average among the companies that responded within 30 days. It is not the average across all of them, because a quarter of them never replied and cannot be averaged in. Quoted without that qualifier it sounds like a typical business, when it is actually the mean of the slow tail.

The more interesting figure is the 37 percent. More than a third of those businesses already replied within the hour in 2011. Being fast was not rare then and it is not a differentiator now on its own.

Number two: the 21 times is real, and it is odds of qualifying, not sales

This one comes from a 2007 paper titled How Much Time Do You Have Before Web-Generated Leads Go Cold?, presented at MarketingSherpa's fourth annual Business-to-Business Demand Generation Summit on 16 October 2007. Its exact words: "The odds of contacting a lead if called in 5 minutes versus 30 minutes drop 100 times. The odds of qualifying a lead if called in 5 minutes versus 30 minutes drop 21 times."

Three things about that paper are worth knowing before you repeat its numbers.

It is a vendor study. It is copyright InsideSales.com, built on InsideSales.com's own platform data, and co-presented by that company's chief executive. The analysis was done by Dr James Oldroyd, who was a Faculty Fellow at MIT's Sloan School of Management at the time, which is why it circulates as "the MIT study". It was never an MIT publication. Say "a 2007 vendor study analysed by Oldroyd" and you are on solid ground.

The dataset is smaller than the headline implies. In the paper's own words: three years of data across six companies, over fifteen thousand leads and over one hundred thousand call attempts. Six companies is a lot of calls and not many businesses.

It measures odds of qualifying, not closing. The paper states plainly that it "did not address close ratios". The version you usually see, that a 5 minute response "converts 21 times better", changes the outcome being measured. Qualifying a lead and selling to one are different events.

The genuinely useful finding in that paper is one almost nobody quotes, because it is not in the summary. The decay is front-loaded. From 5 minutes to 10 minutes, the odds of making contact drop 5 times and the odds of qualifying drop 4 times. Most of the damage is done inside the first ten minutes, not somewhere out at the hour mark. If you take one operational instruction from this whole article, take that one.

Number three: the 78 percent has no source we could find

"78 percent of buyers purchase from the company that responds first" is the most repeated claim in this category. It is attributed almost everywhere to a "Lead Connect survey". We went looking for that survey: no published report, no methodology, no sample size, no date, no author. Every citation we followed led to another blog citing another blog.

We are not claiming the survey does not exist. We searched for it and could not reach it, which is a different and more honest statement. But a number you cannot trace to a document is a number you should not put in a sales deck, and we have removed it from ours.

It is also, on inspection, a strange claim. Taken literally it says response order beats price, product, reputation and referral in 78 percent of purchases, across every category. Nobody who repeats it seems to have asked what population that was measured in.

Number four: the 77 percent is real, and it says something narrower

This one traces. In its Lead Response Report announcement of 17 February 2021, XANT (the company formerly called InsideSales.com) reported on 14,000 companies and 55 million sales interactions.

What the announcement actually says is: "When teams lead with marketing automation, only 23% of leads get touched by sales reps." The 77 percent is the complement, and the release never prints it. Two things get lost in the restatement. The condition, which is teams that lead with marketing automation rather than all businesses everywhere. And the meaning, which is not touched by a sales rep, not never responded to by any means.

There is a coincidence here that makes the folklore look self-confirming, and it is worth naming. The Harvard Business Review audit found 23 percent never responded. XANT found 23 percent do get touched. Same number, opposite meaning, different decade, different dataset. Anyone assembling a statistics roundup from summaries rather than sources will find those two agreeing with each other, and they do not.

So what is actually true

Strip the folklore out and a short, defensible list remains:

• Responding fast measurably improves your odds of reaching and qualifying an enquiry, and the effect is strongest in the first ten minutes.
• A meaningful share of businesses do not respond at all, which is a larger and cheaper problem to fix than shaving minutes off the ones you do answer.
• The evidence base is United States data from 2007, 2011 and 2021. We have not found an equivalent South African study, and we are not going to relabel American numbers as local ones.

That last point matters for how you should read all of this. Nothing above tells you what happens in Cape Town or Gqeberha. What it tells you is the shape of the effect, and the shape is almost certainly directional here too, because the mechanism is human: a person who has just filled in three forms talks to whoever calls while the intent is still fresh.

Why property feels this hardest

Estate agencies run into this before most other businesses, for a structural reason rather than a cultural one. A property enquiry is usually made on a portal that sends the same lead to several agencies at once. The enquirer is not choosing between replying to you and replying to nobody. They are choosing between you and the three agents who also got it.

That changes the maths. In most industries a slow reply costs you some probability. In a syndicated-lead market it costs you the conversation entirely, because by the time you call, someone has already booked the viewing. This is also why agencies feel the effect at a scale the studies above were never measuring: the studies looked at one company's response to one lead, not four companies racing for the same one.

The other property-specific complication is that the fast reply cannot be the whole job. An agency still has FICA obligations before it can act for a client, and speed at the enquiry stage does nothing about the verification work waiting behind it. Fast to first contact, then properly slow where the law requires it, is the right shape.

What changes when this is built properly

Response time is not a discipline problem and telling your team to be faster does not fix it. It is a capture and routing problem, and there are four mechanisms that actually move it.

The enquiry lands in one place, once. Most businesses are slow because the enquiry arrives in four places: a website form to one inbox, a portal to another, a WhatsApp message to somebody's personal phone, a phone call nobody logs. Nobody is watching all four at 19:00. When every channel writes into one record, the clock starts at submission rather than at whenever somebody next opens a tab. Moving WhatsApp off a personal number onto a business number is usually the single biggest change here, because it is the channel most South African enquiries actually arrive on and the one least likely to be recorded anywhere.

The first acknowledgement is automatic and honest. Not a fake personal message. A confirmation that names the property or the service asked about, says who will call and roughly when, and gives the person a way to answer sooner if they want to. This is the only part of the process that can genuinely run in seconds, and it buys back most of the ten minutes the 2007 data says you are losing.

Routing is decided in advance, not in a group chat. Who takes this enquiry is a rule: by area, by property type, by who is on duty. When that rule is written down and applied by the system, the assignment happens at submission. When it is not, the enquiry sits in a WhatsApp group while people work out whose it is, and that discussion is often the entire delay.

Escalation is on a timer, not on somebody remembering. If the assigned person has not made contact within the window you have set, it goes to the next person automatically. This is what stops the one lead that slipped from becoming the one lead nobody ever called, which is the 23 percent problem from the Harvard Business Review audit in miniature.

What does not change: none of this makes the conversation better, and a fast reply to a badly qualified enquiry is still a wasted call. Automation moves the clock, it does not move the skill. Businesses that install the plumbing and expect the close rate to rise on its own are usually disappointed, and that is a reasonable expectation to set before you start rather than after. The same trap catches agencies adopting AI tools in general: the tool does the mechanical part, and the mechanical part was never the hard part.

How to find out where you actually stand

You do not need a study for this. Get a friend to submit an enquiry through your website, your portal listing and your WhatsApp, on a Saturday, and write down what time each one gets a human reply. Most owners are surprised twice: once by the number, and once by discovering one of the three channels goes nowhere at all.

Then measure the same thing for a month before you change anything, so you know whether the fix worked. Doing that properly is the same discipline as ranking any process before automating it: measure, then build, then measure again.

Frequently Asked Questions

What is a good speed to lead time for a South African small business?

There is no local benchmark study to cite. The best-evidenced target is first contact inside ten minutes, because the 2007 data shows contact and qualification odds falling fastest between five and ten minutes. Treat under an hour as the minimum acceptable, not the goal.

Is the 78 percent first responder statistic true?

We could not find a published report, methodology or sample behind it. It is attributed industry-wide to a "Lead Connect survey" that we were unable to locate. We do not use it, and we would not rely on any claim that only traces back to other blog posts.

Does replying faster actually increase sales?

The strongest studies measure contact and qualification odds, not closed sales. The 2007 paper states outright that it did not address close ratios. Faster replies reliably get you more conversations. Turning those into sales is a separate problem.

Does an automatic reply count as responding?

For the customer, partly. An honest acknowledgement that names what they asked about and says when a person will call holds attention and sets expectations. It does not replace the human contact, and the research measures human contact, so keep both clocks.

Why does speed to lead matter more in property than other industries?

Because portal enquiries are usually sent to several agencies simultaneously. The enquirer is choosing between competing agents rather than between you and nobody, so a slow reply loses the conversation outright rather than reducing its odds.

Where to start

If you do not know your current response time across every channel, that is the first thing to find out, and it usually takes an afternoon. If you do know it and it is worse than you would like, the fix is almost always in capture and routing rather than effort.

A Business Autopsy maps how enquiries actually reach you, where they stall and what it would take to close the gap, before anything gets built. If you would rather just talk it through first, book a discovery call.