I Moved 100% of Our Budget Into One Channel
by Chad de Lisle
I made a huge bet a while back.
The data was clear. Paid search was outperforming every other channel we were running, by enough that the decision felt obvious. So I consolidated. Shut the others down, cut sponsored listings entirely, and put the full monthly media budget into Google Ads.
I wasn’t being reckless. I had the numbers in front of me and I read them correctly. I had never felt more sure of any move in my marketing career. Any marketer looking at that same report would have reached the same conclusion.
Then our numbers tanked for a full month.
The Problem: Accurate Data Can Tell A Misleading Story
What our last-lick attribution couldn’t show me was the relationship.
Sponsored listings weren’t generating leads that showed up tagged as sponsored listings. It was producing something harder to measure. Somebody would find us in a listing, read the reviews, and then go search our name directly or type the URL in a week later.
That lead arrived as organic. Or direct. Or branded paid search, where it made Google Ads look even better than it already did.
So several listings sat in my report looking like an expensive line item that produced almost nothing, while quietly feeding the three channels I was crediting for everything.
The moment I turned it off, the top of that chain disappeared. And because the effect was downstream and delayed, it took weeks before the drop was obvious and longer before I understood what had caused it.
Here’s the part that I hate the most: My marketing dashboard stayed green. Paid search efficiency actually looked better for a bit, because I’d stopped spending on something that wasn’t claiming credit for anything.
I had made the business worse and improved my metrics doing it.
It isn’t that the data was wrong. The data was right. It just couldn’t show me a relationship our marketing measurement had never been built to see—I treated a gap in measurement as evidence of no value.
Anything that creates demand rather than captures it has this problem. Review platforms and directories. Brand content. Events. Podcasts. PR. All of them look weak under last-click attribution, and all of them are usually feeding the channels that look strong.
The Marketing Measurement Framework: Three Layers Of Truth
What we built afterward comes down to one question. Is our data telling the truth, and can leadership see it?
Layer One Is Infrastructure
Is the marketing plumbing actually working? This is the most overlooked layer because failures are invisible. The practice is a channel-by-channel audit twice a month with a named owner. They submit all the forms. Call all the numbers. Buy all the things. At every step, note if tracking fires, does the record land in the CRM, does a notification reach a human?
Tracking breaks constantly. It’s close to a law of the universe. If nobody is checking on a rhythm, something is broken right now and you don’t know.
Two specific things to look out for: Redirects that strip tracking parameters, which dump paid traffic into direct and get you killing campaigns that are working. And the gap between marketing data and sales data, where you know how many leads came in and nobody can tell you which campaigns produced customers.
Layer Two Is Diagnostic
When something moves, can you find the cause in minutes rather than days? Revenue is down is a symptom. Conversions down 10%, why? Google flat, Meta down 20%, why? Meta prospecting down, retargeting holding, why? One creative fatigued, and cost per result spiked.
That requires consistent naming conventions and reporting that drills from total to channel to campaign to asset in one view. Build it before you need it.
Layer Three Is Interpretive
Layer three is interpretive, and it usually takes character rather than tooling. It’s the layer my sponsored listing decision lived in. The numbers were accurate. The story I told about them left out everything I couldn’t measure, and I didn’t know I was leaving it out.
Before any channel gets cut, ask what else would move if this disappeared. If the honest answer is nobody knows, that’s not permission to cut it. That’s a reason to test it with a holdout (incrementally testing it) before you commit the whole budget.
The Proof: Owning the Outcome Beats Hiding Behind the Data
The decision wasn’t the worst part. The worst part was the meeting afterward.
Because the easiest sentence available to me was that I’d done what the numbers told me. It’s defensible. It’s even true. The data said consolidate and I consolidated.
And it’s the weakest move a marketer can make.
Hiding behind the data is how you tell your CEO you were a conduit rather than a decision-maker. If the numbers get the credit when it works, the numbers take the blame when it doesn’t, and you’ve just argued yourself out of the job.
So the version I’ve used ever since is: the data informed me, I made the call, I own the outcome.
Jake’s reaction to that was not what I expected. Owning a bad month built more trust than most of my good months had, because for the first time he knew what he was getting from me when things went wrong.
Owning the outcome is terrifying and it’s the cheapest trust you will ever buy.
Take This And Use It Monday
🎯 For Marketing Leaders: Find the channel in your reporting that looks worst, then ask what would stop happening if it vanished. If you can’t answer, don’t cut it. Run a holdout for 30 days and watch what moves in organic, direct, and branded search. Then put a twice-monthly channel audit on the calendar with your name on it. And at your next review, say the words “I made the call” about something that went badly.
📊 For CEOs: When your marketing leader proposes cutting a channel that looks inefficient, ask one question: what else would move if this went away? If nobody knows, you’re about to make a decision in the dark and the bill will arrive two quarters later. And when they own a bad outcome without reaching for the data as cover, pay attention. That’s the person you want making calls.
The Truth
The data told me the truth. It just didn’t tell me everything, and I couldn’t tell the difference.
So here’s the question I actually want an argument about, because I don’t think there’s a clean answer. Every company has at least one channel that can’t prove its value. Do you keep funding it on judgment, or do you cut what you can’t measure?
I’ve now been wrong in both directions. If you’re a CEO you probably have a strong answer. If you run marketing you probably have the opposite one.
Our episode on this is up now. The decision, what it cost, and what we built so it couldn’t happen the same way twice.
— Chad de Lisle
Head of Marketing, Disruptive Advertising
P.S. The next Indispensable Marketer cohort starts January 8th. Eight weeks, live, small group, building VSET (Vision, Strategy, Execution, Team) on your own business rather than reading about ours. The founding cohort just kicked off, so January is the next one. Start the free assessment.
