METHOD, NOT MARKETING

What a Missed Lead Costs

Written by Susana López, founder of Angelica. Sources linked inline with dates.

The short version

Nobody can tell you what a missed lead costs your business, and every company quoting you a percentage is guessing with someone else’s arithmetic. Your own figure takes four minutes and four numbers you already have: inquiries per month, the share that sits unanswered too long, your close rate, average job value. Below: why the statistics in your industry do not survive a fact check, the two studies that do, and the math.

The number your industry quotes you traces back to nothing

The most-repeated statistic in home services marketing is that 78% of customers buy from the first business to respond. Don’t take our word that it is hollow. Take any page that quotes it and read the whole page instead of the headline.

We did that in July 2026 with a lead-response statistics article ranking for these exact searches. Its summary table lists the 2007 MIT / InsideSales study by Dr. James Oldroyd along with what that study found: contact 100 times more likely, qualification 21 times more likely. Two rows down, the same table credits “78% of customers buy from the first responder” to that same study. The published findings of the 2007 study are the 100x and 21x odds. The 78% is not among them. The article contradicts itself without you having to leave it.

The same article puts the average business response time at 47 hours. The Harvard study it cites two paragraphs earlier measured 42.

We are not pointing at anyone in particular, because this is the norm, not the exception. We quoted the 78% on this site too, until we went looking for the original document and could not find one. Neither could a 2026 review that traced these figures back one at a time (AInora, 2026). Our own audit is on the research page where we check these numbers one by one.

So when a page tells you roofers lose 40% of their leads to slow response, ask the only question that matters: lost by whom, measured how, in what year? If nobody can answer it, the number is decoration.

The two studies that hold up, and what they do not cover

Two are real and correctly attributable. Both are old.

The 2007 Lead Response Management study. Dr. James Oldroyd, then a visiting professor at MIT, analyzed data with InsideSales.com covering six companies, more than 15,000 leads and over 100,000 call attempts. Replying within five minutes rather than thirty made contacting a lead about 100 times more likely and qualifying it about 21 times more likely.

The 2011 Harvard Business Review audit. Oldroyd and Kristina McElheran sent a web inquiry to 2,241 US companies and timed the replies: 37% answered within an hour, 23% never answered, and the average reply among those who did came 42 hours later (HBR, March 2011).

The limits, plainly: both measured phone replies to web leads, at companies mostly larger than yours, before anyone texted a contractor about a roof. Neither measured a roofer in Broward County in 2026. The durable finding is the shape of the curve, not the multiplier. Which is why the number you need is your own.

Your number, from four things you already know

Write these four down. Estimates are fine on the first pass.

  • Inquiries per month. Every first contact from someone who might buy: forms, chats, WhatsApp, email, DMs, voicemails. Not traffic. People who reached out.
  • The share that goes cold before you reply. Not the ones you never answer. The ones where your answer lands after they stopped caring.
  • Your close rate on the ones you work. Quotes sent divided into jobs won.
  • Average value of a closed job. Revenue divided by jobs closed, last six months.

Then: inquiries × cold share × close rate × job value = your monthly exposure.

Worked example with invented inputs, so you can see the arithmetic:

InputExample figure
Inquiries per month40
Share that goes cold before you reply25% (10 inquiries)
Close rate on inquiries you work20% (2 jobs)
Average job value$6,500
Monthly exposure$13,000

Those inputs are placeholders, not a claim about anybody’s business. Put your own in and the answer changes completely, which is the point.

Then take the honest haircut. That figure is a ceiling, not a bill. Some of those people call you back, some were shopping six companies and were never yours, some were never real. Cut it by the share that comes back anyway. If a third returns, the example drops to about $8,700 a month. Still yours, not ours.

Do not guess number two. Measure it for ten days.

The second input is the one everybody fudges, and the only one you can measure this week without buying anything.

Open a note on your phone. For ten business days, log two timestamps per inquiry: when it came in, and when a real human reply went out. Not the auto-responder. Then count how many took longer than an hour, and watch where they cluster: on a roof between 10am and 2pm, in peak season when the phone never stops, on Sunday nights when med spa inquiries pile up for Monday. What that looks like in your trade is in the industry-by-industry view.

Ten days of timestamps will tell you more than any statistic on any vendor’s site, including this one.

What to do with the result, cheapest option first

Stop at the first step that closes your gap. Do not skip to the bottom because a page told you to.

  • Free, today. Set business hours and an away message on WhatsApp Business that says what you do, when you will reply, and one question they can answer meanwhile. “What’s the address and what’s going on?” gets you a usable reply while you are on the ladder.
  • Free, this week. Put your inquiries in one place. A lead you never saw and a lead you answered late cost the same, and five scattered inboxes produce the first kind.
  • Cheap. Give one person the first-reply job with a defined window, then check it against your timestamps a month later.
  • Paid, only if the gap survives all of that. Coverage you do not have to remember: a hire, an answering service, or automation. The tradeoffs and the monthly costs are in the decision guide to AI receptionists.

If you land there, here is our product, flatly. Angelica answers website chat, email and WhatsApp in about 60 seconds, 24/7, in English or Spanish. WhatsApp comes with the Growth and Elite plans, on a dedicated business number. Plans are $197, $347 and $597 a month. She does not answer voice calls. If your inquiries arrive as ringing phones and nothing else, this is the wrong tool and an answering service is the right one. Setup takes 1 to 3 business days from the start page.

Who this page is not for

If you get eight inquiries a month and answer them all from your truck within minutes, your exposure is near zero and your problem is lead volume. Spend the money on getting found instead.

If your average job is under a few hundred dollars, the arithmetic rarely justifies paid coverage. An away message and one consolidated inbox get you most of the way.

And if your exposure comes out smaller than the cost of fixing it, the right move is to do nothing. We would rather you know that now than churn in month three.

FAQ

Why do customers stop replying to me?

Usually because the conversation went quiet on your end first and they moved on. A 2011 Harvard Business Review audit of 2,241 US companies found that 23% never responded to a web inquiry at all, and the average reply among those who did came 42 hours later (HBR, March 2011). It is old and it measured phone replies to web leads, so treat it as directional. The pattern holds: most businesses are slow, and the buyer only needs one that is not.

How fast should I reply to a new inquiry?

Fast enough that the person has not contacted anyone else yet, which in practice means minutes rather than hours. The 2007 Lead Response Management study by Dr. James Oldroyd found that replying within five minutes rather than thirty made contact about 100 times more likely. Before you optimize for any specific target, measure what you do now: log the arrival time and the real reply time for every inquiry for ten business days.

Is it my prices, or is it my response time?

There is a test. Look at the deals you lost and split them into two piles: the ones where you sent a quote and got beaten, and the ones where the conversation died before you ever quoted. Pile one is a pricing or positioning problem. Pile two is a response problem, and no discount will fix it. Count both piles before you touch your prices.

What if my leads come in by phone, not by message?

Then automated messaging will not help you and you should not buy it. Angelica answers website chat, email and WhatsApp. She does not answer voice calls. If your inquiries arrive almost entirely as ringing phones, a live answering service is the right purchase. Count your own mix for a week before you decide.

What if my customers write in Spanish?

Angelica replies in English or Spanish, matching whichever language the customer used, without you setting anything up per conversation.

How do I cancel if it does not work?

Plans are billed monthly in advance and there is no annual contract. You can cancel any time and it takes effect at the end of the billing period you already paid for.