Marketing is working when you can write one sentence about last month without opening a tab. Something like: we spent this much, it produced this many conversations, and this is what they had in common.
Most founders cannot write that sentence. Not because the data is missing. Because it is spread across six tools that each describe their own slice, and the slice each one describes is always fine.
So here is the test. It takes ten minutes, it needs no software, and it will tell you more than the dashboards did.
The five-line test
If you cannot fill in all five lines, marketing is not working. It might be happening, but it is not working.
Write these on paper, from memory, for last month.
- How much we spent, in money and in people’s hours.
- How many real conversations that produced: calls booked, replies, demos, trials with a human on the other end.
- Which channel those conversations came from, by name, not by feeling.
- How many of them became customers, and how long it took.
- What the ones that converted had in common that the rest did not.
Line one is easy and everyone stops there. Line two is where the honesty starts, because clicks and opens do not count. Line three is where the dashboards fail you, because every tool claims the same conversation. Lines four and five are the business.
If you got to line three and wrote “probably LinkedIn”, that is the answer to the question in the title. Probably is not working. Probably is the sound of a budget being spent on a hunch.
Probably is the sound of a budget being spent on a hunch.
Why the dashboards cannot take the test for you
Every tool measures its own slice, and every slice is fine.
Your analytics tool sees a visit. Your email tool sees an open. Your calendar sees a booking. Your billing tool sees a payment. Not one of them can see that those four events happened to the same person, in that order, across three weeks.
That person is the only thing you want to know about, and they are invisible to everything you pay for. The tools are not broken. They are honest about their slice and silent about the seam.
Which is why the test is on paper. It forces the join the tools refuse to make. If you can make it in your head for one month, you know your marketing. If you cannot, you know what to fix, and it is not the ad copy.
The one number to write down
Cost per conversation, by channel, with the sample size next to it.
Not cost per click. Not cost per lead. Cost per real conversation, because that is the first point where a human on their side chose to spend time on you. Divide what you spent on a channel by the conversations it produced, and write the count of conversations next to it.
The count is the honesty. A channel with two conversations has a cost per conversation that means nothing yet, and the number should say so. A channel with forty has a cost you can act on.
Do this for three months and a pattern will show up that no dashboard showed you: one channel that produces conversations that close, and one that produces conversations that do not. Move the budget toward the first. That is what working looks like.
Marketing is working when the sentence writes itself. Until then, the test is the work.
See the use case for law, accounting and consulting firms: How does a law, accounting or consulting firm get named when a founder asks who to call?
Questions people ask next
What is the simplest sign that marketing is working?
You can say, from memory, what last month’s spend produced in real conversations and customers, by channel. If the answer starts with “probably”, it is not working yet.
Why is cost per conversation better than cost per lead?
A lead is a form fill; a conversation is a human who chose to spend time on you. Conversations predict revenue; form fills predict form fills.
How much data do I need before trusting a channel number?
Enough that the number would survive being asked twice. A channel with two conversations has no rate yet; write the count next to every number and act only on the ones with dozens.
Can Throughline tell me if my marketing is working?
It reads your CRM, calendar, email and analytics tools, joins them on the email address, and reports which channels started real conversations and which of those became customers, each with its sample size. It reports what moved together and never claims cause.