Kaizen
Sound familiar? · Speed-to-lead

Why am I paying for leads that never turn into jobs?

The invoice from the ad platform arrives every month. The jobs do not. Before you fire the agency or pause the campaign, look at what happens to a lead in the first hour after it lands.

The short answer

Paid leads usually fail to convert because of what happens after they arrive, not because the leads are bad. A lead that waits hours for a reply has already booked a competitor; one followed up once and never again goes cold by default. Before changing the marketing, measure two numbers: median time to first response and how many touches each lead gets. Fixing those two, with instant acknowledgement and an automatic follow-up sequence, typically moves conversion more than any change to the ads.

You know this is you when

  • Your cost per lead looks fine and your cost per job looks terrible.
  • The agency says the leads are good; your team says the leads are tire-kickers; nobody has checked the timestamps.
  • Leads get one call attempt, maybe two, and are then quietly written off.
  • You cannot say which channel produced last month's booked jobs, only which produced the most form fills.
  • Pausing the campaign feels like the only lever you have.
21×
more likely to qualify a lead contacted within 5 minutes than after 30
LEAD RESPONSE MANAGEMENT STUDY / INSIDESALES · MORE DATA →
01

What it is costing you

The arithmetic is unforgiving. If you spend $3,000 a month to generate 40 leads at $75 each and convert 10% into $1,200 jobs, you are paying $750 per job on $1,200 of revenue, and the campaign looks like a loser. The same 40 leads converted at 25% cost $300 per job, and the same campaign is the best thing you did all year. Nothing about the ads changed between those two outcomes; the only variable is what you did in the hours after each lead arrived.

The waste is also invisible in the usual reports. Ad platforms report leads, not response times. Your team remembers the leads that said no, not the ones that never answered a 4pm callback because they booked someone at 10am. So the conclusion "the leads are bad" gets reached with no evidence, and the budget gets cut right as the fix was sitting in the inbox.

02

Why it keeps happening

Paid leads are the most time-sensitive leads you have. Someone who clicked an ad for "emergency plumber" at 9:40 is not comparing brochures; they are contacting the top three results in quick succession. The company that replies first gets a conversation and the others get nothing. If your reply takes an hour, you were never really in the running, and no amount of ad optimisation changes that.

Then follow-up fails. Most paid leads do not book on the first touch; they book on the third or fourth. Studies of home-service estimates find roughly half never receive a single follow-up at all. A lead that got one voicemail is not a bad lead. It is an unworked one.

And attribution is broken in a way that hides the real story. When a lead arrives by form, gets a reply by text, books by phone, and is invoiced from another tool, nothing connects the ad click to the paid job. You end up judging the campaign on form fills, which it is good at, and never on revenue, which is the only thing that matters.

03

How to fix it

  1. 01
    Audit twenty paid leads by timestamp before touching the campaign

    For each, note when it arrived and when the first outbound reply went out, then count the touches it received. Take the median response time. If it is over an hour, or most leads got one or two touches, the marketing is not the problem and pausing it will not fix anything.

  2. 02
    Reply to every paid lead within a minute, automatically

    A text that names what they asked about and asks one qualifying question lands while they are still on the results page. This is the single biggest lever on paid conversion because paid leads are the ones racing you against two competitors.

  3. 03
    Put a follow-up sequence behind every lead that does not book

    Day 1, day 3, day 7, with messages that sound like a person and stop the moment the customer replies. The leads that "went cold" come back in surprising numbers when the fourth touch arrives, because for many of them the first three never did.

  4. 04
    Track the lead from click to invoice

    Each lead should carry its source all the way to the paid job, so you can report cost per booked job by channel rather than cost per form fill. That is when you find out which campaign is actually working, and it is often not the one with the most leads.

Same 40 leads, same adsConverted at 10%Converted at 25%
Jobs booked410
Revenue at $1,200 average$4,800$12,000
Cost per booked job$750$300
Verdict on the campaign"Leads are bad""Best channel we have"

See the fix running, on your business.

A 30-minute walkthrough with Mel, built around slow speed-to-lead. No deck: the product, using scenarios from your trade.

Book a walkthrough
04

Common questions

How do I know whether it is the leads or my follow-up?
Timestamps. If your median first response is under five minutes and every lead got four or more touches and conversion is still poor, the leads may genuinely be weak. In most businesses we see, response is measured in hours and touches in ones, and the leads were never given a fair chance.
Should I pause the ads while I fix the follow-up?
Usually not. The follow-up fix takes days to switch on and the ads take weeks to re-learn after a pause. Turn on instant acknowledgement and a follow-up sequence first, then judge the campaign on cost per booked job over the following month.
Are Local Services Ads leads different?
They are the most urgent leads you will get, because the customer is calling from the listing with a problem right now. They reward speed above everything, and a missed LSA call without a text-back is close to money set on fire. The same applies to any pay-per-lead marketplace.
What does Kaizen do with a paid lead?
It replies by text within seconds, attaches the lead source to the customer record, runs a follow-up sequence until the customer answers or books, and carries that source through to the invoice. Cost per booked job by channel is a report, not a research project.