01
Train an agent on your business
Homepage, pricing, notes — two minutes of setup so the reply engages the lead’s message instead of confirming receipt.
Research
Every speed-to-lead pitch quotes the same numbers: five minutes, 21×, 78%. Some of those numbers are real research. One of them is untraceable. And nearly everyone citing the real ones skips what the study actually measured.
Here’s the honest version — what was studied, what holds up, and what it means if your leads arrive through a website form.
The “MIT study” is the Lead Response Management study by James Oldroyd, built on data from InsideSales.com — an analysis of when companies called back leads who had filled out a web form, and what happened when they did.
Its headline findings: the odds of making contact with a lead were roughly 100× higher when the call came within 5 minutes versus 30 minutes, and the odds of qualifying that lead were roughly 21× higher. Note both halves: the leads came from web forms, and the responses being measured were phone dials. The famous decay curve is about how fast a form fill goes cold — the clock starts at the submission.
The follow-up, published in Harvard Business Review (“The Short Life of Online Sales Leads,” 2011), audited 2,241 US companies by submitting a test lead through each one’s own website and timing the response.
The results: average first response time was 42 hours. Only 37% responded within an hour. About a quarter took more than 24 hours — and 23% never responded at all. Against that baseline, responding in seconds isn’t an optimization; it’s a different sport.
“78% of customers buy from the first company to respond” appears in thousands of sales decks, attributed to a rotating cast of sources — none of which lead to a methodology, a dataset, or an original publication anyone can produce. We don’t use it, and we’d suggest you don’t either. The traceable numbers above are more than strong enough.
Today the 5-minute study is quoted mostly to sell AI phone callers and dialers: respond to every web lead with an instant call. That’s one reading. But the lead in the study chose to type a message into a form — and 2007’s phone habits are not 2026’s, where an unknown number calling seconds after you hit submit is as likely to be declined as answered.
The durable lesson of the research isn’t “call people.” It’s that the value of a form lead collapses within minutes, so the first meaningful response should arrive within seconds — in a channel the lead will actually receive. For a form fill, the channel they demonstrably use is email: it’s how they just contacted you.
Your competitors, statistically, will respond in hours-to-never. A reply that lands in seconds — engaging the lead’s message and asking one qualifying question — captures the window the research describes without staffing anything. That’s the product we built, and every stat on this site links back to the sources on this page.
Homepage, pricing, notes — two minutes of setup so the reply engages the lead’s message instead of confirming receipt.
One extra notification recipient in any form builder. The clock the study describes starts at the submission; this stops it in seconds.
Send a test lead through your own form, watch the reply arrive, and compare that to the 42-hour average the HBR audit found.
The full walkthrough — with what to upload and how to test before paying — is in Connect your form & go live.
Is the 5-minute study still valid in 2026?
The dataset is from 2007, and no one has replicated it at scale since — treat the exact multipliers with care. But the HBR audit’s finding that most companies respond slowly or never has been re-confirmed by vendor audits repeatedly, and the decay logic of a lead comparing multiple providers hasn’t changed.
Did the study measure email response at all?
No — it measured phone dials to web-form leads. That’s precisely why we cite it carefully: the timing evidence is about how fast form leads go cold, not about which channel must answer them.
Where does the 78% first-responder stat come from?
Nobody can say — which is the problem. It circulates with shifting attributions and no locatable original study. We recommend dropping it from your own materials.
Stop it in seconds. Test it free on your own form and time the reply yourself.
