A B2B software founder showed me two numbers and asked an obvious question. Meta was producing leads at about $19. LinkedIn was producing them at about $81. Why, he wanted to know, were we still running LinkedIn at all?
The answer was not in either report. We pulled the closed deals for the previous two quarters and looked at where they had originated. The Meta leads had converted to a customer roughly one time in ninety. The LinkedIn leads had converted about one in twelve.
At those rates the Meta cost per customer was far higher. The cheap channel was the expensive one, and it had looked cheap for two quarters because nobody had measured past the form.
Why LinkedIn costs what it costs
You are paying for certainty about who you are reaching. LinkedIn knows job title, company size, industry and seniority because its users maintain that data themselves for their own reasons. No other platform has that at the same reliability.
For most consumer advertising that precision is worth very little — it is cheaper to reach everybody and let the algorithm sort it out. For a product sold to, say, finance directors at manufacturing companies over two hundred staff, it is worth a great deal, because the alternative is paying to reach a hundred wrong people to find one right one.
The price is high. Whether it is expensive depends entirely on what a customer is worth to you.
When LinkedIn earns its price
Four conditions. The more of them you meet, the more defensible the cost.
Your customer is definable by job attributes. If you can describe your buyer as a title at a type of company, LinkedIn can target it. If your buyer is "anyone with a dog", it cannot, and you are paying a premium for nothing.
Your deal size is substantial. As a rough test, if a customer is worth less than about $3,000 in lifetime value, LinkedIn's cost per lead rarely works. Above $10,000 it usually does, and above $50,000 the question is why you are not spending more.
Your sales cycle involves several people. LinkedIn is the only channel where you can reach a buying committee deliberately — the person who signs, the person who uses it, the person who objects.
You can measure past the form. Without closed-revenue data you will be comparing cost per lead across channels and reaching the wrong conclusion, exactly as that founder was about to.
Structure it for a long cycle
The most common mistake is running LinkedIn like Meta: one campaign, a lead form, a demo request, done. B2B does not buy that way.
Separate cold from warm properly. Cold audiences get something useful with no demo ask — a genuine piece of research, a teardown, a comparison. Warm audiences, people who engaged or visited, get the demo offer. Asking a stranger for a sales call is what produces the $200 leads people complain about.
Use lead gen forms for the top, your own landing page lower down. Native forms convert far better because the fields are pre-filled, but that is also why their leads are weaker — the friction that disqualifies people is gone. Use them where volume matters and send higher-intent traffic to a page you control, built properly: a dedicated page beats your homepage more here than anywhere.
Expect retargeting to do the closing. In a four-month cycle, the ad that produces the demo is rarely the first one someone saw. Budget for the follow-up, not just the first touch.
Keep audiences above roughly 50,000. Tighter than that and delivery gets expensive and erratic. If your perfect segment is 4,000 people, LinkedIn is not where to reach them — that is a job for a list and an email.
What to measure instead of cost per lead
Three columns, in this order.
Qualified lead rate. What share of leads match your actual buyer definition? This exposes channel quality differences immediately and it is the number most reports omit.
Cost per qualified lead. Cost per lead divided by that rate. This is the first figure that compares channels honestly.
Pipeline and closed revenue by source. The only one that settles the argument. It needs your CRM to record origin and keep it through to close, which is a one-afternoon setup that most B2B companies have not done.
This is the same failure that makes cheap Google Ads leads useless in high-ticket categories: an account optimised toward form fills will reliably find you more of the cheapest, least committed ones.
When to spend the money elsewhere
To be useful rather than merely enthusiastic.
If your deal size is small, your buyer is not definable by job title, or you cannot yet measure closed revenue by source, spend on search first. Someone typing your category into Google has declared intent, and intent is cheaper than precision.
If your LinkedIn leads are arriving but nobody is following up within a day, fix that before raising budget. In B2B the speed of the first response moves the close rate more than the channel does.
And if you have never run the qualified-lead analysis at all, do that before adding any channel. Most companies that think they have a traffic problem have a qualification problem, and it is far cheaper to find out.
The short version
LinkedIn is expensive per lead and frequently cheaper per customer. Whether that holds for you depends on deal size, how definable your buyer is, and whether you can measure past the form — which is the condition most companies fail, and the reason so many conclude LinkedIn does not work.
Measure cost per qualified lead and closed revenue by source. Then the decision makes itself, in whichever direction.
If you are weighing channels more broadly, where a first serious budget should go covers the same reasoning for e-commerce, and a brand that reads like every competitor raises your cost per lead on every channel at once.
We run B2B paid media as part of our Google Ads and Meta Ads work, with LinkedIn alongside where it earns its place. The free growth audit includes the qualified-lead analysis, which is usually the first time a B2B account has been looked at that way.