Your CFO just asked why LinkedIn's 0.67% click-through rate is worth celebrating when Google Ads delivers 6.33%. The question sounds reasonable. The comparison is not. CTR is the most misused metric in B2B paid media because teams treat it as a cross-channel scorecard when it is actually a within-channel diagnostic. Get this wrong and you will reallocate budget toward the wrong channels, optimize for the wrong signals, and defend the wrong numbers in your next pipeline review.
The Numbers That Actually Matter
Metadata's 2026 benchmark report, drawn from 153 B2B advertisers and $57.6M in 2025 spend, provides the clearest picture of what "good" looks like across channels:
- LinkedIn: 0.67% CTR, $9.39 CPC, $202 cost per lead
- Facebook: 0.79% CTR, $1.95 CPC, $145 cost per lead
- Instagram: 0.65% CTR, $2.82 CPC, $138 cost per lead
- Google Ads: 6.33% CTR, $9.76 CPC, $524 cost per lead
The 9x CTR gap between Google and LinkedIn is arithmetic, not performance. A search impression follows stated intent; a feed impression interrupts someone scrolling through their network. Comparing them is like comparing a warm inbound call to a cold email open rate.
Why CTR Fails as a Cross-Channel Metric
The practical implication is that CTR is close to useless as a cross-channel scorecard. A team moving budget from LinkedIn to Google because "the CTR is 9x better" is comparing a feed interruption to a search result. The right comparison is cost per lead, and even that requires a quality adjustment. Google Ads produces the most expensive leads in the Metadata dataset ($524) and the most intent-qualified ones. LinkedIn's $202 leads come from a different buying stage entirely.
42 Agency's B2B Google Ads benchmarks show even wider variance by vertical: construction tech runs $150 to $300 per lead, while logistics and supply chain can hit $1,500 to $3,000. If your CFO is comparing your numbers to a single industry average, you are defending the wrong baseline.
What "Good" Looks Like Within Each Channel
CTR becomes useful when you compare creative performance within a single channel. Here are the thresholds that separate average from strong:
LinkedIn Sponsored Content
The B2B House reports a global average of 0.44% to 0.65%, with single-image ads at 0.56%, video at 0.44%, and carousel at 0.40%. Getuplead's 2026 benchmarks suggest aiming for 0.7% to 1.0% when launching a new campaign; hitting that range improves your Quality Score and can lower CPC over time.
Facebook (B2B SaaS)
Metadata's B2B customer survey puts the average at 0.65%. Anything above 0.8% signals strong creative-audience fit.
Google Ads (Search)
Usermaven's 2026 industry benchmarks show that a CTR above 3% is considered strong in most industries, with high-intent niches exceeding 5%. B2B search typically runs lower than consumer categories because the audience is narrower and the keywords more competitive.
Google Ads (Display)
Expect 0.35% to 0.7% as average, with 0.5% to 1.0% considered good. Display is awareness, not intent capture; judge it on view-through conversions and assisted pipeline, not raw CTR.
The Format and Targeting Levers
CTR is not fixed. It responds to format, targeting, and creative decisions you can change in a single sprint.
Format matters more than most teams realize. DigitalApplied's 2026 LinkedIn benchmarks show Document Ads generating 3.4x more dwell time and 2.6x more leads per dollar than static image ads in B2B technology, professional services, and financial services. The format lets the audience consume content without leaving the feed, which is closer to organic consumption behavior than any other LinkedIn ad format.
Targeting precision changes the math. ABM-targeted campaigns combining LinkedIn Matched Audiences with persona filters convert 2.7x higher than industry-plus-seniority targeting alone, according to the same DigitalApplied data. The audience-size sweet spot is 50 to 500 companies: large enough to deliver media efficiency, narrow enough to maintain message-market fit. ABM campaigns also produce 38% lower CPLs than broad targeting once the audience reaches statistical maturity.
Predictive Audiences deliver a 19% CTR lift on average. LinkedIn's Predictive Audiences feature, rolled out in 2025, builds lookalikes from your existing converters using engagement signals across the platform. The AI-bidding integration also reduces manual budget pacing work by an estimated six to eight hours per campaign manager per week.
The Conversion Rate That Matters More
CTR gets you to the landing page. Conversion rate determines whether that click becomes a lead. LinkedIn Lead Gen Forms convert at 6.1%, roughly 5x the CVR of off-platform landing pages. Pre-filled profile data and the in-feed checkout-style experience reduce drop-off to 28%, compared to 65% drop-off when sending traffic to external forms.
If your CTR is strong but your cost per lead is high, the leak is probably on the landing page, not in the ad. Run a holdout test: send half your traffic to a Lead Gen Form and half to your existing landing page. Measure cost per lead, not CTR. The answer will be obvious within two weeks.
A Two-Week Pilot to Pressure-Test Your Benchmarks
Before your next pipeline review, run this diagnostic:
- Pull CTR by channel and by creative within each channel. Flag any creative running below the channel median.
- Calculate cost per lead by channel. Compare to the Metadata benchmarks above, adjusted for your vertical.
- Identify the format gap. If you are running zero Document Ads on LinkedIn, you are leaving the highest-engagement format on the table.
- Test one ABM audience. Upload a list of 100 to 200 target accounts, layer on persona filters, and compare CPL to your broad-targeting campaigns after 500 clicks.
The goal is not to hit a benchmark. The goal is to know which benchmark applies to your channel, your format, and your audience, and to defend that number with data your CFO can verify.
CTR is a diagnostic, not a destination. Use it to compare creative within a channel, not to compare channels against each other. The metric that matters is cost per qualified lead, and the only benchmark that matters is the one you can trace back to pipeline.
