Two parts. First, a test you can run tonight to see how long a customer really waits. Then a calculator that turns your own numbers into what slow replies might be costing you, and how many extra jobs a fix would need to pay for itself.
Written by Sean Solano Paul, founder of AutonomicsAI LLP. Published .
Part 1: the 10-minute test
Do this from a phone that isn't yours, after 6 pm, the way a homeowner would.
Search for your business name the way a stranger would, and tap through to your site. Note how many taps it takes to reach a form.
Fill in your own quote form with a real email address and mobile number you can check. Say the problem is a leak.
Write down the time you pressed send.
Check that phone's email and texts every few minutes until something arrives. Write down what arrived and when.
The next morning, note when anyone at your business actually got in touch.
Three things to look for:
Did anything arrive at all? A form that goes to an old inbox is more common than you'd think. If nothing came, fix that before anything else.
Did the reply say what happens next? "Thanks for your message" isn't much use to someone with water coming through the ceiling. "Your request is in, Mike will call by 8 am, and here's what to do tonight if it's getting worse" is.
How long until a person called? That's the gap a competitor can step into.
In 2011 the Harvard Business Review published a study of how fast companies answer web leads, and its headline finding was that most weren't answering anywhere near fast enough. Your test tells you whether that's true of you, which matters more.
Part 2: what the wait might cost
The calculator below does four sums. They're shown underneath so you can check them, or do them on paper.
The sums, in plain words
Requests that wait = requests a month × the share nobody can answer.
Jobs lost to the wait = requests that wait × how many of those go elsewhere × your win rate. We multiply by your win rate because you wouldn't have won all of them anyway.
Revenue at stake a year = jobs lost × average job value × 12. Gross profit at stake = that × your margin.
Break-even = what a fix costs a year ÷ the gross profit on one job. For our roofer setup that's $350 for the pilot plus $397 a month. Twelve months of the plan plus the pilot is $5,114, so we use that as the yearly cost even though the first month is part of the pilot.
A worked example with made-up numbers
A roofer gets 40 requests a month. A third arrive when nobody can answer (about 13). They win 30% of the requests they answer. Average job $9,000, gross margin 35%, so about $3,150 gross profit a job. They guess 1 in 5 of the waiting requests go elsewhere.
13.3 waiting × 0.2 × 0.3 = 0.8 jobs a month, or about 9.6 a year. That's roughly $86,000 of revenue and $30,000 of gross profit. Break-even on $5,114 a year is 2 jobs.
Change the guess to 1 in 10 and the answer halves. That's why the calculator asks you for it instead of hiding an assumption in the maths.
What this doesn't tell you
It doesn't know your town, your competitors or your reviews. A fast reply won't save a quote that's double everyone else's.
It counts only requests that wait. Slow replies during the day cost jobs too.
The fix doesn't have to be us. The same sums tell you whether an answering service, a part-time hire or a cheaper tool is worth it.