Check the measurement first
Are calls tracked? If not, every channel that mainly produces calls looks weak, and your Google Business Profile is usually the worst affected — the single most valuable local channel appearing to produce nothing.
Does conversion tracking fire correctly? Submit a real enquiry and confirm it records within a day.
Is branded search credited to the right place? People searching your name were already coming, often from a channel that gets no credit.
Do you know which enquiries closed? Without it you are judging on lead volume, and cheap leads that never convert will look like the best performer.
Cut nothing until those four are answered. This step is cheap and it is the one most often skipped.
The number that decides it
Cost per booked job, compared against value per lead — average job value × gross margin × close rate.
A channel producing jobs above your value per lead is losing money. Below it, it is earning before overhead.
Not cost per lead, not return on ad spend, not click-through rate.
Give it enough time
Paid ads: several weeks minimum. The first fortnight is the campaign learning which searches to ignore, and early numbers are always poor.
Search and content: six months. Cutting at three is cutting before the thing you paid for has arrived.
Email: several sends, and after list hygiene has been addressed.
Directories and offline: a full season, because seasonality can swamp everything.
Cutting early is the most common way businesses conclude a channel does not work.
Before cutting, try these
Fix response speed. A channel converting badly may be delivering fine leads to a phone nobody answers.
Check the landing page matches. Paid traffic to a mismatched page fails regardless of the channel.
Narrow rather than stop. Reduce area or service before switching a channel off entirely — the problem is often a segment rather than the whole thing.
Reduce spend and watch. A pause tells you more than a deletion, and it is reversible.
When cutting is right
- Cost per booked job is clearly above your value per lead, over a fair period
- The leads are consistently outside your area or service
- The channel requires a response speed you cannot deliver
- The price has risen year on year while quality has not
- You cannot serve the work it produces
The incrementality question
Worth asking of any channel reporting results that look too good: what would have happened without it?
Branded search advertising is the usual example — excellent numbers, and many of those people were already coming to you. Pausing it for a few weeks in a quiet period, changing nothing else, is the only way to find out.
Deciding
Check the data, then compare cost per booked job to your value per lead. If you want a second pair of eyes, call 832-338-2926.
Frequently asked questions
What should I check before cutting a channel?
Whether calls are tracked, whether conversion tracking fires correctly, whether branded search is credited properly, and whether you know which enquiries closed. Cut nothing until all four are answered.
How long should I give a channel?
Several weeks for paid ads, six months for search and content, several sends for email, and a full season for directories and offline. Cutting early is how businesses conclude a channel does not work.
What should I try before cutting?
Fix response speed, check the landing page matches the promise, narrow the area or service rather than stopping, and reduce spend and watch — a pause is reversible, a deletion is not.
What number decides it?
Cost per booked job compared against value per lead — average job value times gross margin times close rate. Not cost per lead, return on ad spend, or click-through rate.