Think of a chest freezer in a garage that runs all year and the only things in it are packages of meat and ice cream you pull from once in a while. Nobody actually bought this freezer or this food, but it came with the house, the freezer still works, and unplugging it or moving it feels more like a hassle than just letting it sit where it is.
A marketing channel can end up in exactly the same spot. It produces something, occasionally. You are still posting to it, still paying for it, still carving out time in your week to nurture it, and if somebody asked you point blank if that effort is worth it, you would have to admit you haven't taken the time to sit down and work it out.
Most small businesses are quick to add marketing channels and slow to let them go. Adding one feels like progress, and dropping one feels like defeat, so the decision to drop one isn't made on that channel's merit. Let's talk about what that merit is and how you can decide whether it is time to close a marketing channel.
What is a marketing channel?
A marketing channel is any distinct, repeatable process that brings you leads or customers: A paid platform that you promote yourself on like Google or Meta; An organic channel like Instagram with a stream of followers; A newsletter you place ads in; A marketplace like Etsy; A directory you offer yours services through; A referral partner who sends you work.
The useful move is to treat each one as its own small business with its own inputs and outputs, rather than as a permanent fixture of how you operate. A channel is something you are currently choosing to engage with to "drum up" new business.
A channel is worth keeping only when it earns its cost, its quality, and your time compared to what else you could do instead. If it does not, and after you have attempted to fix it, it is time to shut it down.
Why do marketing channels stay alive longer than they should?
Three things keep a weak channel alive: the money and time you already sank into it, a goal too vague to ever fail, and noisy numbers you have emotional attachment to rather than an unbiased look at their successfulness.
The first is that you have already put money and months into the marketing channel, and walking away feels like admitting the money was wasted. This is a real and well documented phenomenon of how people think. The 1985 study most often cited for demonstrating the sunk cost effect, by Hal Arkes and Catherine Blumer, found that people who had already sunk money into something rated it more likely to succeed than people looking at the identical prospect from fresh, outsider's perspective. Your judgment about the channel got worse the moment you invested in it.
The second is that the goal was never specific enough to fail. If you started the channel to "build awareness" or "have a presence," there is no result that could ever count as "unsuccessful". Vague goals do not produce vague answers, they produce no answers.
The third is that the numbers are noisy and disorganized, so you fall back on an emotional response to the effort. That feeling is usually generated by the most recent thing that happened (maybe a most recent sale or burst or new leads), but this is a poor way to judge long timelines of activity.
What should you measure before killing a channel?
You do not need a model. Rate each channel 1-5 on four things: cost efficiency, lead quality, operational drag, and strategic value, using your last 60 to 90 days.
| What to measure | The question it answers |
|---|---|
| Cost efficiency | What does a lead or a customer from this channel cost compared to your other channels? |
| Lead quality | Do these leads actually close, stay, and spend well compared to your best customers historically? |
| Operational drag | How much time per week does keeping this channel alive consume? |
| Strategic value | Is this channel helping me grow my business in the long-term? |
Score each one 1 (weak) to 5 (strong) using your last 60 to 90 days.
Lead quality is the one most owners skip, because it is the one that takes real record keeping. It is also the one that changes the answer most often. A channel can look cheap on cost per lead and be your worst performer once you see what those leads actually do, which is the entire argument for scoring your leads in the first place.
Strategic value is the one people abuse. It is an important category, but it is also where many channels goes to hide from scrutiny because there is less hard data involved. If you find yourself defending a channel purely on strategic value year after year, you're hoping for a future vision you don't want to put extra work towards. See our chest freezer analogy above.
How do you set a test window before judging a channel?
Give the channel a specific timeframe, usually 30 to 90 days depending on how long your sales cycle runs, a specific budget in money or hours, and a specific goal that you would describe as successful. These strict boundaries are very specific and unique to your business, but you have to decide what is most important to you and your business's success. Make the goals tough to achieve, but still in the realm of possible.
Here are some examples to get you started:
Break even on spend within 60 days.
Ten booked calls at or under a specified cost per call.
Fifty new subscribers who open at least two emails in the first month.
Be careful that the window is longer than your time to money, meaning how long a lead actually takes to turn into payment. We cover how to measure that number in what to fix when small business marketing stops working.
How do you judge a channel against your alternatives?
A channel does not have to be perfect, no marketing channel is ever perfect. It has to beat what you would get from spending the same money and hours somewhere else. Compare cost per qualified lead, close rate, and average first job value across your channels over the same period.
A channel that is simultaneously more expensive, lower quality, and takes more work than your alternatives is not a tough call to make, but many business owners avoid that decision for the sake of maintaining a routine.
How do you run a fix-it-or-fold-it cycle?
Before you shut anything down, give it one honest attempt at repair. Many channels that get killed were not run properly, and just need some help to improve.
-
Name the single most likely constraint. Some examples are a wrong audience, weak offer, a landing page that does not match the ad, or follow up that never happens. If choosing just one constraint is hard to do, then the whole channel should be re-evaluated.
-
Change one or two things that address it directly. If the constraint is the offer, change the offer and leave the targeting alone, so that the output is limited to only one changing variable.
-
Run it for the full window you agreed to. The most common failure here is panicking at the halfway mark and changing something else, which turns a test into a chaotic mess of changing results.
-
Measure against the original threshold. Not against how it felt, and not against a new threshold you invented after seeing the results.
If it still misses after a deliberate repair attempt, you now have something much better than a hunch. You have evidence, and evidence makes the next part easy.
Should you scale, maintain, or sunset a channel?
Every channel should end the review in one of three buckets:
- Scale is for a channel that is efficient, brings good quality, and does not eat up too much of your week. Put more money or more hours into it, add creative, and build the assets that support it properly.
- Maintain is for a channel that is fine. It reaches a segment you want, or it earns real strategic value, and it does not cost you much to keep running. Keep it, but put it explicitly on maintenance mode so it stops consuming more attention than it should.
- Sunset is for a channel that loses on cost and quality, takes disproportionate work, and offers nothing unique. Wind it down.
The trap is that almost everything drifts into Maintain by default, because Maintain requires no decision. If your review puts five of six channels in Maintain, re-evaluate your channel and make a tough decision.
How to shut a channel down cleanly
-
Turn off the spend first. Pause the ads or the outreach so you stop adding new inputs you will have to handle.
-
Close the loop on everyone already in it. Anyone mid conversation gets a real final message with a clear way to continue if they want to. These people raised their hand and you already paid to reach them.
-
Archive the creative, the copy, and the numbers somewhere you will find them. A channel that does not work today can work in two years when your offer or your capacity is different, and the notes are the expensive part to recreate.
-
Decide where the freed money and hours go before you free them. Into a channel that is working, into a real experiment, or into fixing delivery. Capacity that is not assigned in advance gets absorbed into the week and you will never see the benefit you just bought.
How we think about this at TKBS
TKBS is a small Michigan marketing and web shop, and our thesis is that we don't sell services, we build systems. A system implies choices you can revisit, which is why we think channels should be reviewed on a schedule instead of when somebody gets frustrated.
We cannot tell you which channel to cut without looking at your actual numbers, and anyone who tells you which platform is dead in a sales call is selling a headline. What we can tell you is that the businesses that grow tend to be the ones that decide, in both directions, rather than the ones that accumulate. A channel you consciously keep on maintenance is fine. A channel nobody has evaluated since it started is a chore.
It also helps to remember that a channel is rarely the whole problem. Often what looks like a dead channel is a good channel pointed at a website that cannot convert what it sends, or one more symptom of marketing bought one disconnected piece at a time.
Frequently asked questions
How do I know when to stop running a marketing channel?
Score it on cost efficiency, lead quality, operational drag, and strategic value over the last 60 to 90 days, then compare it to your other channels rather than to perfection. If it is more expensive, lower quality, and more work than your alternatives, and one honest repair attempt did not move it, stop running it.
Should I try to fix a channel before I kill it?
Yes, once. Name the single biggest constraint, whether that is audience, offer, landing page, or follow up, change only what addresses it, and run it for the full window you already set. If a fair repair attempt still misses the threshold, you have evidence instead of a guess, and it is time to sunset it.
How long should you run a channel before deciding to kill it?
30 to 90 days, and always longer than your time to money, the real gap between first contact and payment for your business. A window shorter than that cycle will not tell you the channel failed, only that you did not wait long enough. Set the window, the budget, and the success threshold before you start.
Should I pause a channel or shut it down completely?
Pause first. Turning off the spend stops the bleeding immediately and costs you nothing, and it gives you a clean period to see whether anything actually drops. Shut it down fully once you have closed out the leads already in it and archived the assets.
What should I do with the budget from a channel I cut?
Decide before you cut. Assign it to a channel already performing, to a real experiment with a written hypothesis, or to improving delivery and retention. Freed budget that is not assigned in advance quietly gets absorbed and you never see the gain.
Is it bad to run only one or two marketing channels?
No. Two channels you run well beat five you run distractedly. More channels only help when you have the capacity to run each one properly, and most small businesses hit that limit far sooner than the advice they read suggests.
See which channels are actually earning their place
If you are running channels you have never really evaluated, we will go through them with you against real numbers from your own business and tell you honestly which ones to scale, which to leave on maintenance, and which to unplug.