LinkedIn Ads for B2B: What Actually Works

A $15 CPC that closes a $40,000 deal is cheap. A $2 CPC that produces nothing but form-fill spam is expensive at any price. That’s the trap most B2B teams fall into on LinkedIn: they benchmark cost-per-click against Google or Meta numbers, panic at how high it looks, and either abandon the platform or pressure the account into cheaper tactics that quietly destroy the thing that made LinkedIn worth using in the first place — its targeting precision.

Start from minimum viable budget, not maximum viable reach

LinkedIn’s ad auction behaves differently below a certain daily spend threshold. Campaigns running under roughly $50–75/day often get stuck showing to a narrow, repetitive slice of the audience because the algorithm doesn’t have enough signal volume to optimize delivery. In practice, this means a $20/day test budget frequently produces worse cost-per-result than a $100/day budget on the identical audience and creative — not because more spend is inherently better, but because the platform needs enough daily volume to learn who within your target list is actually responding.

If the budget available can’t support at least $50–75/day per active campaign, it’s usually better to run one well-funded campaign for two weeks than to spread that same money thin across four underfunded ones that never escape the platform’s learning phase.

Single Image and Document ads outperform video for most B2B offers

Video gets pitched as the higher-engagement format, and it can be for brand awareness plays. But for B2B demand generation specifically — where the goal is a demo request, a content download, or a webinar registration — static Single Image ads and Document ads (the carousel-style native document format) consistently produce lower cost-per-lead than video in accounts running comparable spend. Two reasons show up repeatedly:

• Video requires a longer watch time before the message lands, and LinkedIn’s feed environment doesn’t reward that patience the way TikTok’s does — most B2B buyers are scrolling LinkedIn between meetings, not settling in.
• Document ads let the prospect flip through a short, native-feeling carousel (3–5 slides making one clear argument) without leaving the feed, which produces a much softer, lower-friction path to the click than a video that requires sound or full attention.

This doesn’t mean video has no place — a founder or customer testimonial video can work well for retargeting warm audiences who already know the brand. But as the first-touch format for cold ABM audiences, image and document formats are the more reliable starting point.

Job title targeting is a blunt instrument — layer it with company attributes

Targeting “VP of Marketing” alone pulls in an enormous range of company sizes, industries, and buying authority levels, and the CPMs reflect that breadth. The accounts getting real efficiency out of LinkedIn layer job title with at least one company-level filter:

• Company size band matched to actual ICP (not just “51–200 employees” as a default — pull it from your closed-won data).
• Industry, when the offer genuinely performs differently across verticals.
• Company list upload, for true ABM — a named list of 200–2,000 target accounts, targeted directly rather than through job title alone.

The narrower, list-based approach typically shows a higher CPM but a meaningfully lower cost-per-qualified-lead, because almost nobody in the audience is a mismatch. Job-title-only targeting is cheaper per click and more expensive per outcome — the classic B2B paid social trap.

Retargeting website visitors is the highest-leverage campaign most accounts underfund

Cold prospecting on LinkedIn is expensive by nature — you’re paying to interrupt someone who wasn’t looking for you. Retargeting website visitors who didn’t convert is dramatically cheaper per result and gets underfunded constantly, usually because the audience size looks small in Campaign Manager and teams assume it’s not worth a dedicated campaign.

A retargeting audience of even 500–1,000 monthly website visitors, shown a specific offer (case study, ROI calculator, a limited-time consultation slot) rather than a generic “learn more” ad, routinely produces the lowest cost-per-lead of any campaign in the account. The fix is simple and often skipped: install the LinkedIn Insight Tag before launching any cold campaigns, let it build audience for 2–3 weeks, and budget a standing retargeting campaign alongside prospecting from day one rather than as an afterthought once cold campaigns feel expensive.

Conversation Ads work for high-intent, narrow offers — and nowhere else

Conversation Ads (the message-style format that opens a chat-like sequence in LinkedIn inbox) get a lot of attention because open rates look excellent compared to email. But they only perform when the offer is genuinely narrow and high-intent — a specific event invite, a personalized audit offer, a named-account outreach sequence. Used for generic top-of-funnel content downloads, they produce high open rates and near-zero downstream conversion, because the format itself signals “sales outreach” to the recipient, and people engage with it accordingly — clicking to see what it is, then bouncing once they realize it’s not personal.

Reserve this format for offers you’d be comfortable sending as a genuinely personalized 1:1 message, aimed at a tightly defined account list, not as a scaled substitute for standard ad formats.

Creative refresh cadence matters more on LinkedIn’s smaller audiences

B2B target audiences on LinkedIn are inherently smaller than consumer audiences — a well-targeted ABM list might only be 1,500–5,000 people. That audience sees the same ad repeatedly much faster than a Meta consumer audience would, and frequency fatigue sets in within 2–3 weeks rather than the 6–8 weeks common on larger consumer platforms. The practical implication: budget for creative refresh every 3 weeks for cold campaigns targeting narrow lists, even if performance hasn’t visibly dropped yet — waiting for the metrics to decline means you’ve already burned budget on fatigued impressions before noticing.

Measure against a sales-cycle-adjusted window, not a 7-day click window

LinkedIn’s default reporting window undercounts B2B results because the sales cycle for most mid-market and enterprise deals runs 30–90+ days, far longer than any ad platform’s attribution window. A campaign that looks mediocre on day 14 based on form fills alone might be feeding a pipeline that closes in month three. The corrective isn’t complicated: tag leads by source at the CRM level, and review LinkedIn’s contribution to closed-won revenue on a quarterly cadence in addition to the weekly lead-volume check in Campaign Manager. Judging LinkedIn purely on platform-native, short-window metrics is the single most common reason B2B teams conclude the platform “doesn’t work” when the real issue is a measurement window mismatched to how long their deals actually take.

Put the budget where the sales team already sees quality, not where CPL is lowest

The cheapest cost-per-lead campaign on LinkedIn is often the one producing the least sales-qualified pipeline — broad job title targeting on a generic offer will always win on CPL and lose on close rate. Before scaling any campaign, get a monthly read from sales on lead quality by campaign, not just volume from marketing’s own dashboard. The campaigns worth scaling are the ones sales asks for more of, even if their CPL looks worse on the surface — that’s usually the narrow ABM list campaign, not the cheap broad one.

A Worked Example: Two Campaigns, Same Budget, Opposite Verdicts

Put actual numbers on the CPL trap. A broad campaign — “VP of Marketing,” no company filters, $100/day — might produce 40 leads a month at $75 CPL, which looks excellent on a dashboard. Sales works the list and finds 3 are sales-qualified, a 7.5% qualification rate, and none close within the quarter. Effective cost per qualified lead: $1,000. Cost per closed deal: unmeasurable this quarter, possibly infinite.

A narrow ABM campaign — company list of 800 named accounts matched to closed-won firmographics, layered with job title, same $100/day — might produce only 12 leads a month at $250 CPL, which looks three times worse on the surface. But sales qualifies 8 of those 12, a 67% qualification rate, and 2 close within the quarter at an average contract value of $28,000. Effective cost per qualified lead: $375 — nearly a third of the broad campaign’s number, despite the sticker-shock CPL being higher. This is the exact reversal that makes CPL, taken alone, a misleading metric for judging LinkedIn campaigns, and it’s why the sales-quality check in the previous section has to happen before any budget reallocation decision, not after.

The Failure Mode: Optimizing the Campaign Instead of Fixing the Offer

A recurring pattern in underperforming LinkedIn accounts: the team iterates endlessly on targeting, bidding strategy, and creative variations while the actual offer behind the ad — the thing being promised on the landing page — never gets questioned. If the offer is a generic “request a demo” ad with no specific value proposition, no format or targeting change will fix the fundamental problem, because the ad’s job is only to get the click; the offer is what has to convert curiosity into a form fill worth having.

Before spending another month tuning campaign settings, audit the offer itself against a simple test: would a specific, named person at a specific target account, seeing this exact offer with no other context, understand within three seconds what they’d get and why it’s worth the click. Offers that pass this test — “See how [competitor-adjacent company] cut onboarding time 40% in 60 days” beats “Request a Demo” — consistently outperform targeting refinements of any sophistication, because the audience-offer match matters more than audience precision alone once targeting is already reasonably tight.

Sequencing the Fixes: What to Change First When an Account Underperforms

When a LinkedIn account isn’t producing pipeline, the instinct is to change everything at once — new creative, new targeting, new budget, new offer — which makes it impossible to know afterward what actually fixed it. Work through fixes in this order instead, since each step depends on the one before it being solid:

1. Budget floor first. Confirm every active campaign is running at least $50–75/day; anything below that is fighting the algorithm’s learning phase before creative or targeting even get a fair test.
2. Targeting precision second. Check whether campaigns are running job-title-only with no company-level filter — this is the single most common source of wasted spend and should be fixed before touching creative.
3. Offer quality third. With budget and targeting reasonably sound, audit whether the offer itself is specific and compelling using the three-second test above.
4. Creative format and refresh cadence last. Only once the first three are solid does it make sense to test video versus image versus document formats, or to worry about refresh cadence — optimizing creative on top of a weak offer or bad targeting wastes effort polishing the wrong layer.

Accounts that jump straight to step 4 — endless creative testing — while steps 1 through 3 remain unaddressed are the ones stuck concluding “LinkedIn doesn’t work for us” after months of activity that never touched the actual problem.

Measuring Whether Any of This Actually Worked

Because of the attribution window problem discussed above, weekly platform metrics alone will not answer whether a LinkedIn program is working — they need to be paired with a small set of downstream numbers reviewed monthly and quarterly. Track, at minimum: cost per sales-qualified lead by campaign (not cost per lead), pipeline generated by campaign at the 60- and 90-day mark after the lead entered the CRM, and win rate for LinkedIn-sourced opportunities compared to the company’s other channels. A LinkedIn program that’s genuinely working shows pipeline and win-rate numbers holding up or improving over two to three quarters, even if weekly CPL fluctuates — and a program that looks fine on CPL but shows declining pipeline contribution over the same window is the clearest sign that budget is being spent on the wrong campaigns, regardless of how good the weekly dashboard looks.