One finding in lead generation is replicated often enough to plan around, and it is about speed: the odds of reaching an online lead collapse within minutes, not hours. Almost every other number you will read on this subject is a vendor benchmark with a product attached. This page separates the two, and is deliberate about which claims it will and will not make.
What does the research actually say about response time?
The most-cited work is The Short Life of Online Sales Leads (Oldroyd, McElheran and Elkington, Harvard Business Review, March 2011). Its actual findings, stated correctly:
- Firms responding within one hour were nearly seven times more likely to qualify a lead than firms responding later.
- Responding within five minutes rather than thirty made a firm roughly 21 times more likely to qualify the lead.
⚠ You will see a "100× more likely to convert within five minutes" figure attributed to HBR all over the internet, including on pages selling chatbots. That number is about making contact, not converting, and it comes from a 2007 MIT / InsideSales dataset rather than from HBR. We had it wrong on this page ourselves until August 2026. If a vendor quotes you the 100× as an HBR conversion figure, they have copied it from somebody who copied it.
The corrected numbers are less dramatic and still decisive. Seven times is an enormous effect for something that costs nothing but a mechanism.
Why does speed matter so much?
Because an online enquiry is almost never exclusive. Somebody researching a supplier, a guesthouse or a plumber contacts several at once, and the first useful reply frames everything that follows. You are not competing on quality at that moment. You are competing on being present.
That is also why the effect is so much larger than most marketing interventions: it is not persuading anybody of anything. It is arriving before the decision has already been made elsewhere.
What is different about South Africa?
Two things, and both are measurable rather than anecdotal.
Messaging is the default channel. DataReportal's Digital 2026 South Africa report puts internet penetration at about 79.6% of the population, with WhatsApp overwhelmingly the most-used platform among internet users aged 16 to 64. A South African enquirer expects a messaging-speed reply, not an email-speed one, and the gap between those two expectations is where leads are lost.
Business hours do not match enquiry hours. This is the part nobody publishes a statistic for, and you do not need one — you can count it. Export last month's enquiries, sort by timestamp, and count how many arrived outside 08:00–17:00 SAST. In most small businesses it is between a third and a half.
How do I work out what this is worth to my business?
Do not use a benchmark. Use four numbers you already have:
| Number | Where to get it |
|---|---|
| Enquiries last month | Your inbox, your forms, your phone log |
| How many arrived outside working hours | Sort by timestamp. Count them |
| How many of those you never replied to | The uncomfortable one. Count it anyway |
| Your average job or order value | You know this |
Multiply the third by the fourth, then apply whatever conversion rate you think is honest. That is your ceiling — the most an instant-response system could possibly be worth to you. If that number is smaller than the monthly fee, stop reading and keep your money. That happens, and a vendor who tells you it never happens is selling rather than advising.
Which benchmark numbers should I distrust?
- Any conversion-rate benchmark without an industry and a sample. "The average landing page converts at 6.6%" is a blended figure across wildly different businesses and tells you nothing about yours.
- Any chatbot statistic published by a chatbot company. Including ours. We do not quote one on this page for exactly that reason.
- Any figure that has lost its original source. The 100× above is the standard example: repeated for fifteen years, attributed to the wrong publication, describing the wrong outcome.
- Round numbers in headlines. "AI increases conversion by 50%" is a claim shaped for a slide, not for a business.
What actually fixes the response-time problem?
Something that answers when nobody is there. That can be a person on shift, an answering service, or an automated system — the mechanism matters less than the fact that one exists. A website chatbot is the cheapest of the three at R299 a month plus R250 once-off, and it replies in about a second at any hour. If your problem is a shared inbox rather than a website, an AI email assistant starts at R349 a month.
Seven times more likely to qualify a lead, for responding within the hour. That is not a marketing effect. It is arriving before somebody else already answered.
What is the shortest version?
Speed is the one lever with real evidence behind it, and the honest numbers are smaller than the ones you will be quoted. Count your own after-hours enquiries rather than trusting a benchmark, and put something — anything — in front of them.
Where we do this
The same build runs for businesses in Bloemfontein, Polokwane and Rustenburg, and the city pages set out what changes locally — the languages, the trading hours and the sectors that dominate. Everything is deployed and supported remotely from Vanderbijlpark.