When a prospect fills in your contact form, the clock starts. They've just told you they're interested, and they're probably telling two or three of your competitors at the same time. Most businesses know speed matters. Fewer have looked at how large the gap actually is, or why their own replies run late.
The best-known study here is from Harvard Business Review (2011). It looked at 1.25 million sales leads across 42 US companies. Firms that tried to contact a lead within an hour of the inquiry were nearly seven times as likely to qualify it as firms that tried even one hour later, and more than 60 times as likely as firms that waited 24 hours or longer.
The same research audited 2,241 companies on how fast they actually replied. The average response time was 42 hours, and 23% never replied at all.
An earlier study from MIT and InsideSales.com (2007) looked at a shorter window. Across more than 15,000 leads at six companies, contacting a lead within five minutes instead of thirty made it about 21 times more likely to qualify. That sample is small, and both studies are old. They point the same way, though: the first hour matters most, and the sooner the better.
In most small businesses, a new inquiry lands in an inbox next to everything else. A serious buyer who wants a quote this week and someone who is "just looking" arrive looking identical. Nothing says "this one first." So everything waits until somebody happens to check, and on a busy day that can be hours.
It's a sorting problem more than a speed problem. If the right message is obvious, people reply quickly. If it's buried, they don't.
We had exactly this problem on our own website: every inquiry arrives as a plain email. So we built a triage layer in n8n. A new lead hits a webhook, Google Gemini scores how urgent it looks and drafts a first reply, and anything flagged high-intent triggers an instant alert to the team. Everything else waits behind it.
Two honest notes. It is built and tested on our own lead data, which is what the case study describes, so we're not claiming production numbers. And it doesn't reply on its own. The AI sorts and drafts; a person reads the draft, edits it, and decides what goes out. We think that's the right split for anything a customer will read.
The triage idea works with plain rules too. Three changes that cost almost nothing:
Add automation once these rules are stable. Automating a process nobody has defined just makes the confusion faster.
If you don't know the answer to the first one, that's the finding. Check the timestamps on your last ten inquiries and compare them with the timestamps on your replies. Most owners are surprised.
Harvard Business Review, 2011: study of 1.25 million leads at 42 US companies, and a separate audit of 2,241 companies. MIT Lead Response Management Study, 2007 (Dr. James Oldroyd, with InsideSales.com): more than 15,000 leads at six companies.
