Most LinkedIn reporting we inherit from clients looks healthy and explains nothing. Impressions are up, click-through rate is respectable, cost per click is described as high but acceptable, and nobody in the room can say whether the spend produced revenue. That gap is the whole problem. The linkedin ads metrics the platform surfaces first are the ones it can measure instantly, and instant measurement is rarely the same thing as commercial truth. A considered B2B purchase takes weeks or months to close, involves several people who never click the same ad, and lands in a CRM that the ad account cannot see.
So the job is not to collect more numbers. It is to decide, before the campaign launches, which numbers you would be willing to defend in front of a finance director who does not care about social media. That decision changes what you build: which conversions you define, which fields you pass into the CRM, which reports you refuse to send. We run this exercise with every LinkedIn client in the first week, and it usually removes half the metrics from the dashboard while making the remaining half much harder to argue with.
Start by admitting what LinkedIn costs
LinkedIn is expensive per click compared with most other paid channels. That is not a flaw to be optimised away; it is the price of reaching a filtered professional audience by job function, seniority and company. The practical consequence is that LinkedIn only makes sense when a single won deal is worth enough to absorb a lot of clicks that go nowhere. If you sell a low-value product, a subscription with a small monthly fee, or anything bought on impulse, LinkedIn will almost certainly lose money for you no matter how well the campaigns are built. We say this to prospects regularly, and some spend the budget on search instead. That is the right outcome.
Once you accept the cost structure, the measurement framework follows. A channel with a high entry price must be judged on deal value, not on volume. Cost per qualified opportunity becomes the number you manage, and the other linkedin ads metrics exist to explain why it moved.
The linkedin ads metrics worth reporting
We group reporting into three layers. The delivery layer tells you whether the campaign is reaching the intended audience at a sane price. The engagement layer tells you whether the creative and offer are landing with that audience. The commercial layer tells you whether any of it produced money. A report that mixes the three without labelling them is how agencies hide behind good top-of-funnel numbers for six months.
- Cost per qualified lead, where qualified means a sales team member has confirmed fit, not that a form was submitted.
- Lead to opportunity rate, split by campaign and by audience, because a campaign can produce cheap leads and zero opportunities.
- Pipeline value influenced by LinkedIn, tracked in the CRM with a source field that is set once and never overwritten.
- Frequency against your target audience, since small B2B audiences fatigue quickly and rising frequency usually precedes falling response.
- Cost per click and click-through rate, used only to diagnose creative and targeting problems, never as headline results.
Notice what is missing. Follower growth, reactions and video views are not on the list. They are not worthless, but they belong in an organic content review, not in a paid performance report where they pad the page and distract from the cost per opportunity that is quietly climbing.
Getting the data to survive the journey to your CRM
Attribution breaks in boring, fixable ways. A lead form captures a personal email address while the CRM keys on company domain. A sales rep re-enters a record manually and the source field resets to direct. Someone clicks an ad on their phone, remembers the brand two weeks later, and arrives via a branded search. Each of these is a data plumbing problem, and each one makes LinkedIn look worse than it is.
Our fix is unglamorous. We agree a single source field with the sales team, lock it against overwriting, pass campaign and creative identifiers through the form, and reconcile the ad platform against the CRM monthly rather than trusting either one alone. We also ask sales to record a plain-language answer to how the buyer first heard of the company, because self-reported attribution catches the influence that click tracking never sees. Where a click-based view and a self-reported view disagree badly, that disagreement is itself useful information about how the channel really works for this business.
Reading the numbers over a realistic window
The last failure mode is impatience. Judging linkedin ads metrics after two weeks on a product with a three-month sales cycle guarantees a wrong conclusion. Early on, all you can honestly evaluate is delivery and engagement: is the audience the right one, is the offer getting attention, is frequency sensible. Commercial judgement has to wait until enough leads have had time to move through the pipeline.
We set that window explicitly at the start of an engagement and write it into the reporting schedule, so nobody has to defend a channel against a verdict it could not yet have earned. It also protects the account from the most expensive habit in paid social, which is rebuilding campaigns every fortnight in response to noise. Stable campaigns with clean data and a patient review cycle beat a frequently rebuilt account in almost every B2B case we have measured. That discipline is where defensible results come from.
Keep reading: Linkedin Ads · Social Media Marketing