Your leads
Your questions
Times can be minutes (12), 1h 30m, 1:30 or 2 hours; blank or never means nobody answered. Won can be won/lost, yes/no or 1/0. Leave out leads still open. At least 40 answered leads, 8 won and 8 lost.
Upload or paste your leads, then click Analyse.
Result
Win rate by how fast you replied
| Reply time | Leads | Won | Win rate | 95% range |
|---|
What the fitted curve says
How it is calculated
Leads with a reply time are grouped into six bands and each band's win rate gets a Wilson 95% interval. A logistic regression of won on ln(1 + minutes) gives a smooth curve (fitted by Newton steps with a tiny ridge so perfectly separated data cannot run away). The what-if takes every answered lead slower than your target and replaces its fitted win chance with the chance at the target, summing the difference; it is scaled up for unanswered leads and never counts a negative gain. The 90% range comes from 300 seeded bootstrap refits that keep the same number of wins and losses, so a result is reproducible. Revenue per month = extra wins per lead × your leads per month × average job value (taken from won leads with a value when you do not enter one, and only with at least 3).
Not modelled: why a lead was answered fast or slowly (easy leads may get quick replies), lead source and job size differences, time of day, whether you replied by phone or text, how many competitors the customer contacted, and capacity to actually answer faster. Unverified: the arithmetic is checked in unit tests against known curves and hand-computed intervals; the estimate is an association in past data and has not been validated against a change in how a business actually answers.
Common questions
Does answering a lead faster really win more jobs?
Often, but this tool measures it from your own log rather than asserting it. It groups your leads by how long the first reply took, shows the win rate in each group with an interval, and fits a curve. If your slope is flat or upward, the page says so and shows no gain.
What reply time should I aim for?
Whatever the curve says is realistic for you. Enter a target in minutes and the tool estimates the extra wins if every slower lead had been answered at that target. Try 5, 15 and 60 minutes to see how much of the gain sits in the first few minutes and how much in the first hour.
Is the extra revenue a promise?
No. It is an association in your past leads. Fast replies can go to easy leads, and slow replies to hard ones, which would make speed look more valuable than it is. The page gives a 90% range from a seeded bootstrap and labels the figure a ceiling when the trend is not clearly different from noise.
What goes in the file?
One row per inbound lead: minutes until your first reply (blank or never if nobody answered), and whether you won it. Optional job value. A header row may name the columns in any order. Leads still open should be left out.
How many leads do I need?
At least 40 answered leads with 8 won and 8 lost. With fewer, the tool refuses rather than draw a curve through luck. A few hundred leads across different reply times gives a much tighter range.
Why are unanswered leads kept separate?
Never answering is a different event from answering slowly, so they do not enter the curve. They are listed on their own line so you can see how many you lost without a reply.
Is my lead list uploaded anywhere?
No. The file is read and analysed in this browser tab. Nothing is sent to a server, there are no accounts or cookies, and the tool never stores your leads. It works offline once the page has loaded.