Creating demand from cold audiences costs $187 per lead. Retargeting people who already know you costs $196. That nine-dollar difference just blew up every budget assumption you've been defending in quarterly reviews.
Metadata's 2026 B2B Paid Media Benchmark Report analyzed $57.6 million in calendar-2025 ad spend across 153 advertisers and 211,000 leads. The headline finding is almost too clean to believe: the "premium" everyone budgets for demand creation barely exists. The real gap worth acting on isn't between cold and warm audiences. It's between ad formats nobody runs and ad formats everybody defaults to.
The Demand Creation Myth, Quantified
For years, the conventional wisdom has been simple: creating demand is expensive, capturing it is efficient. Budget accordingly. Except the data says otherwise.
Inside cold audiences, document ads delivered leads at $148, while video ads came in at $358. That's a 2.4x difference hiding in plain sight. Meanwhile, the typical B2B advertiser puts just 25% of budget into creating demand, 7% into retargeting, and 16% into buying clicks.
That last number should make you wince. Sixteen percent of budget going to click or traffic objectives. And here's the kicker: 99.4% of those click-objective campaigns recorded no lead at all. Not a low conversion rate. Effectively zero.
If marketing is like dating, as I like to say, then running traffic campaigns in B2B is like paying for a billboard on a highway where nobody ever exits. You're buying impressions from people who will never convert, because the algorithm is optimizing for exactly what you asked for: clicks from clickers.
The Format Gap Nobody Talks About
The report's most actionable finding isn't about channels. It's about formats.
Document ads, the carousel-style PDFs that let prospects swipe through content without leaving their feed, outperform video by a wide margin in cold audiences. LinkedIn data shows document ads generate 3.4x more dwell time and 2.6x more leads per dollar than static image ads in B2B technology and professional services.
Yet almost nobody runs them. The typical B2B advertiser defaults to video because video feels premium, or to single-image ads because they're easy to produce. The format that actually works sits unused in the campaign builder.
This is shiny object syndrome in action. We chase the format that looks impressive in a creative review instead of the format that performs in the data. The report doesn't just suggest you test document ads. It suggests you're leaving money on the table every week you don't.
LinkedIn's Quiet Dominance
Dreamdata's 2026 LinkedIn Ads B2B Benchmarks Report found that LinkedIn now accounts for 41% of B2B ad budgets, ahead of any individual Google channel. It's also the only major ad platform with a positive ROAS: 121%, up from 113% last year.
The cost per company influenced on LinkedIn dropped 54% year over year, from €154 to €70. That's not a rounding error. That's a structural shift in platform economics.
But here's what makes the Metadata data even more interesting when you layer it on top: the cheapest channel isn't always the best-converting one. The report explicitly notes that channel selection should be evaluated on cost-per-pipeline-dollar, not cost-per-lead. A $362 webinar lead that converts to opportunity at 14.2% beats a $178 paid social lead that converts at 4.1%. The math isn't complicated, but it requires you to track past the form fill.
The 272-Day Reality
The average B2B customer journey now lasts 272 days, with 88 touchpoints across four channels and ten stakeholders involved, according to Dreamdata's analysis. That's up from 211 days last year.
Eighty-one percent of that journey happens before a deal even hits the sales pipeline. If your brand isn't visible during the research phase, you're not on the shortlist. And if you're not on the shortlist, you're not in the deal.
This is why the Metadata finding about demand creation costs matters so much. The old model assumed you could afford to underspend on cold audiences because retargeting would do the heavy lifting. But when 94% of buying groups have already ranked their preferred vendors before contacting any of them, according to 6sense's buyer research, the retargeting window is too late. You're not harvesting demand. You're harvesting the scraps.
What to Do Monday
The report includes a "Monday list" of seven moves, each tied to a specific finding. Five are settings changes or an afternoon of work. Two are planning decisions. Here's the distilled version:
Kill your traffic campaigns. If you're running click-objective campaigns, stop. The data is unambiguous: 99.4% of them produce nothing. Reallocate that 16% of budget to lead-generation objectives with proper conversion tracking.
Test document ads in cold audiences. The $148 CPL versus $358 for video isn't a marginal improvement. It's a different category of performance. Run the test for two weeks and let the data decide.
Rebalance toward demand creation. If you're at 25% of budget on cold audiences, you're at the median. The report publishes the full range, including the boldest quartile. Consider whether your current mix reflects strategy or inertia.
Connect your CRM. The pipeline metrics in the report use a $29.4 million subset of lead-generation campaigns with CRM opportunity attribution. If you can't trace a lead to closed-won revenue, you're optimizing for the wrong thing.
The Benchmark Trap
One caveat worth noting: benchmarks are useful for context, not for targets. 42 Agency's 2026 B2B Paid Media Benchmarks show LinkedIn CPL ranging from $207 to $345, with industry-specific variation that can swing costs by 3x or more. DevOps and engineering leads run $400 to $600 for content, $1,500 to $2,500 for direct lead gen. Legal tech sits at $97 to $110.
Your number is your number. The value of the Metadata report isn't that it tells you what your CPL should be. It's that it tells you what 153 other advertisers actually paid, with the methodology visible and the dataset filterable. That's rare. Most benchmarks are surveys of what marketers think they paid, not what they actually spent.
The Uncomfortable Question
Here's what the report doesn't say, but implies: most B2B paid media programs are running on assumptions that haven't been tested against real data in years. The demand-creation premium that justified your budget split? Doesn't exist. The video-first creative strategy your agency recommended? Underperforming document ads by 2.4x. The traffic campaigns you kept running because they "build awareness"? Producing nothing measurable.
The data is public. The methodology is transparent. The only question left is whether you're willing to let it change what you do on Monday.
