It usually happens in a quarter that came in light. Someone who signs off on the spending, a partner, a board, whoever holds the card, looks down the list of marketing costs and stops at the LinkedIn line. It is the biggest one. They ask a reasonable question: what is this actually doing for us?
And the honest answer, the one most owners have, is a shrug and some impressions.
So it gets cut. Notice what did not happen. Nothing about the campaigns changed that morning. No performance threshold was crossed. The channel was not judged and found wanting, because there was nothing to judge it with. It was cut because it could not be defended, which is a different thing entirely, and it is the more common one.
Your Ad Account Is Telling You the Truth About the Wrong Thing
Here is why the defense is so hard to mount.
LinkedIn is not where people buy. It is where they first hear of you, form an impression, and file you away for the month or the year when the problem you solve becomes urgent. Almost nobody sees a LinkedIn ad and books a call that afternoon.
The ad account measures the afternoon. It counts clicks, and conversions that happened inside its own attribution window, and it divides the money by them. When you spend real money reaching people who are not ready yet, that arithmetic returns a small number of expensive leads. Every figure in the report is accurate.
That is the trap. You cannot argue with the numbers, because the numbers are correct. You can only argue that they are answering a question nobody asked, and that argument is very hard to make in a meeting with a spreadsheet open.
Some of the Leads You Did Get Never Reached the Report
Before you defend the number, it is worth knowing the number is understated.
Most conversion tracking still leans on a cookie surviving in the visitor's browser between the moment they click your ad and the moment they fill in your form. On a B2B purchase that gap is rarely minutes. It is days or weeks, across a work laptop, a phone on the train and a browser that clears things out on a schedule nobody chose. When the cookie does not survive, the conversion still happens. It just never gets matched back to the ad that started it, so it lands in your account as though it came from nowhere.
Practitioners who manage this channel at volume put the share lost this way in the region of a fifth to a third of conversions. Treat that as their experience rather than a published figure, but the direction is not in dispute, and it points the same way every time: your ad account is crediting LinkedIn with less than it earned even on the narrow, same-session definition it uses.
The fix is a server-side conversion feed, which LinkedIn calls its conversions API. Instead of relying on the browser to remember, your site or your CRM tells LinkedIn directly that a conversion happened, and LinkedIn matches it to the ad it served. It is more setup than a tracking tag and less than an integration project, and it is worth doing before you conclude the channel is not working.
You Are Holding One Channel Responsible for a Job It Shares
The most useful reframe here comes from the practitioners who manage this channel at volume, and it is worth stating plainly.
Suppose you run the numbers on LinkedIn alone and it comes out slightly underwater. Slightly more money going out than can be traced coming back. The obvious move is to cut it, and the obvious move is often wrong, because you have measured one part of a system as though it were the whole system.
If LinkedIn is the reason a portion of your search traffic already knows your name, then some of the profit sitting in the search column was manufactured in the LinkedIn column. Cut LinkedIn, and search does not stay where it is. It gets slowly worse, over a period long enough that nobody connects the two events. The channel that looked expensive was subsidizing the channel that looked efficient, and now neither of them is what it was.
This is not an argument that LinkedIn always pays for itself. Sometimes it genuinely does not, and it should go. It is an argument that the number you cut it on was never the right number, and almost nobody calculates the combined one before deciding.
LinkedIn Knows Something No Other Platform Knows
There is one report that speaks directly to this, and most owners running LinkedIn ads have never opened it.
When a company in your pipeline becomes a qualified lead or closes, LinkedIn can look backwards and tell you whether it served ads to people at that company beforehand. Not whether someone clicked. Whether the company was reached at all, in the months before the deal appeared.
Think about why only LinkedIn can do this. Every ad platform knows something about the person it showed an ad to. Only LinkedIn reliably knows where that person works, because employer is the first thing everybody puts on their profile and keeps current. Google cannot tell you that one person saw your ad, mentioned you to a colleague, and that the colleague is the one who filled in your form six weeks later. LinkedIn can see the shape of that, because it knows the two of them work at the same place.
The specifics matter, because this is the step where people give up. The report lives in Business Manager, not in Campaign Manager, which is where most people look and conclude it does not exist. You need admin access on Business Manager, and you connect Salesforce, Dynamics 365 or HubSpot through a standard authorization handshake rather than by handing over a password. Expect to wait: LinkedIn's own documentation says it can take up to seventy-two hours for the data to appear, and another day or two after that before it shows in Campaign Manager. It is not a five-minute job, but it is a one-time job, and it does not cost anything.
Read the List of Accounts. Ignore the Number Beside It.
Now the caveat, because this report has a real flaw and you should know about it before you put it in front of anyone.
It takes the deal values sitting in your CRM and reports them next to LinkedIn, as though LinkedIn produced them. It did not. Your salesperson produced them, probably after a referral, three emails and a call, in a process where a LinkedIn impression was one early moment among many. If you walk into a budget conversation and claim that revenue figure, you are claiming something untrue, and the first person who checks will be right to distrust everything else you said.
So use the part that is solid. The report can give you the list of which accounts LinkedIn actually reached before they entered your pipeline. That list is a fact. It is checkable, it is unglamorous, and it is far more persuasive than a headline number that overstates your case.
The claim you want is narrow and defensible: of the companies that turned into real conversations this quarter, this many of them had already been reached on LinkedIn first. That sentence survives scrutiny. "LinkedIn generated this much revenue" does not.
No CRM to Connect? Two Free Ways to Get the Same Evidence
Plenty of small businesses do not run a CRM that any of this plugs into, and the advice usually stops there. It should not, because the two cheapest methods cost nothing at all.
Ask. Put "how did you hear about us?" as an open field on your form, or have whoever answers the phone ask it and write the answer down. Owners consistently find LinkedIn named in those answers while their LinkedIn dashboard shows almost no conversions, which tells you the dashboard was never going to catch it. People do not remember which ad they clicked. They do remember where they first came across you.
There is a persistent worry that adding a field will hurt your conversion rate. One optional question, placed after the fields that matter, is not what stops somebody who wants to talk to you.
Cross-reference by hand. Take the companies that reached a real conversation in the last quarter. Then open the company view in your LinkedIn campaign reporting, look back over a longer window than the quarter itself, and check how many of those companies you had already been serving ads to. It takes an afternoon and no budget, and it produces exactly the sentence from the section above. It is not perfect attribution. It is evidence, and it is more than most people bring.
Stop Paying for the Clicks That Make Your Numbers Look Worse
One practical thing before the conversation, because some of what you are defending is self-inflicted.
By default you are paying to show ads to your current customers, your past customers, and your competitors. All three click. Your customers click because they know you, and your competitors click because they are checking on you. None of them will ever become the lead you are counting, and every one of those clicks is priced into the cost per lead you are about to have to justify.
LinkedIn lets you exclude all three, and it takes minutes. Upload your customer list as an audience and exclude it. Build a list of the competitors you know by name and exclude that too. If you have a list of people who already bought, that is the same job again.
There is a version of this worth keeping: a campaign aimed deliberately at existing customers, showing them what you have added since they signed, is a genuinely good use of a small budget. The problem is not reaching them. The problem is reaching them accidentally, inside a campaign whose cost per lead you are about to be asked to justify, where their clicks arrive as cost and never as leads.
None of this transforms your results. It stops you paying to make your own case harder to argue, which is a smaller claim and an easier one to act on this afternoon.
Decide the Sentence Before the Quarter Turns
The mistake is treating this as a reporting problem to solve later. It is a conversation you can predict, and you already know roughly when it is coming.
So decide now what you will say when someone points at the line item. Not "brand awareness," which sounds like an admission that you cannot count it. Something with a number in it that you can stand behind: this many of the companies that became real conversations this quarter had already seen us on LinkedIn, and here is the list.
Connect the report if you have something to connect. Ask the question on your form if you do not. Either way, gather the evidence in a quarter that is going fine, because the quarter where you need it is not the quarter where you will have time to build it.
We spend a lot of our time on exactly this problem, so if you would rather hand the whole thing to someone who already knows where the report lives, that is the work we do. And if you would rather run it yourself, everything above is the actual method. The channel is easier to defend than most owners think. It just has to be defended with the right evidence, gathered before anybody asks.
