Paid media benchmarks answer one question: what does a click cost? That’s a useful question for a media buyer. It’s the wrong question for a CEO.

The question a CEO, founder, or CFO actually has to answer is different. What does a dollar of paid growth buy a company like mine, how long does it take to come back, and am I above or below the line?

Gartner's sample is mostly large enterprises, so treat its figures as a ceiling on marketing intensity rather than a peer benchmark for a $5M to $50M ARR company. Benchmarkit and SaaS Capital measure different things (share of revenue versus share of ARR) and should not be combined.

The strongest evidence that acquisition and retention are one system comes from High Alpha's 2025 SaaS Benchmarks (800+ companies). It grouped companies by NRR and CAC payback:

Source: High Alpha with Growth Unhinged, 2025 SaaS Benchmarks Report, published November 2025, survey August to September 2025.

You will notice there is no LTV: CAC benchmark in this report, and no single 'good' CAC payback number. That is deliberate. The familiar 3:1 LTV: CAC rule is widely repeated, but we couldn't find a current, primary, methodology-backed source for it. Many of the CAC payback medians that circulate online trace back to OpenView reports, and OpenView stopped publishing after 2023. We would rather give you fewer numbers you can defend in a board meeting than more numbers you can't.

Paid media is only as efficient as its understanding of how buying actually happens. In 2026, buying happens mostly without you.

This section covers the three channels that carry most B2B SaaS paid budgets: Google Search, LinkedIn, and Meta. For each, we show platform metrics first and revenue outcomes second, because the two often point in opposite directions.

Verdict: still essential for capturing demand, increasingly poor at creating it.

Platform benchmarks, US search campaigns (medians):

Source: Dreamdata, B2B Google Search Ads Benchmark (September 2025). Aggregated, anonymized Dreamdata customer data.

Source: Seer Interactive, AIO Impact on Google CTR, September 2025 update (published November 2025). 3,119 informational queries and about 1.1M paid impressions. Informational queries aren't the same as commercial B2B buying terms, so treat this as directional.

Across the period, paid CTR on queries showing an AI Overview fell by roughly two-thirds. Separately, Ahrefs found position-one organic CTR is 58% lower when an AI Overview appears (300,000 keywords, December 2023 versus December 2025).

Verdict: the channel most worth scaling for B2B SaaS, and the one most often wasted.

Revenue outcomes. This is where LinkedIn differs from the other channels.

Source: Dreamdata, LinkedIn Ads Benchmarks Report 2026 (March 2026). Data-driven attribution on closed-won deals over 12 months. Dreamdata changed its company identification method in 2025, which weakens year-on-year cost comparisons.

LinkedIn also takes about 41% of B2B paid budgets among Dreamdata customers. And it's slow: it takes about 281 days on average from first LinkedIn impression to closed revenue. A quarterly review will usually undervalue it.

Verdict: useful in a supporting role, risky as a primary B2B channel.

Source: Metadata.io, B2B Advertising Benchmarks 2026 (CY2025). Facebook n=71, Instagram n=49 advertisers.

If paid growth is a capital allocation decision, waste is capital destroyed. The evidence suggests it's substantial, and that most of it comes from setup and measurement choices rather than from the channels themselves.

An agency audit of 43 enterprise B2B SaaS Google Ads accounts found 36.1% of spend went to queries or form fills that never reached SQL within 90 days. Of that waste, the largest causes were: broad match without negative keywords (32%), Performance Max without offline conversions (24%), and default 7-day click attribution (15%). In the same audit, among the 21 accounts with broad geographic targeting and no exclusions, 23% of clicks came from regions the advertiser doesn't serve.

On LinkedIn, an audit of 56 B2B SaaS accounts put average waste at 32.0%, ranging from 13.5% for the best-managed accounts to 52.4% for the worst-managed. Non-ICP job-function targeting caused 30% of that waste, and only about 22% of job-function spend reached roles in the actual buying group.

In monitor-only accounts analyzed for Lunio's IT and Security 2026 report: invalid traffic averaged 15.34% of LinkedIn clicks, 11.63% on Bing, 5.37% on Meta, and 3.90% on Google over nine months to Q1 2026, with LinkedIn's rate rising every quarter (Lunio, via PPC Land). Treat this as a reason to check, not a reason to panic: the source sells click-fraud protection, and the sample is one vertical.

Cost per lead is the most common target in B2B paid media and one of the least predictive of revenue. LinkedIn Lead Gen Forms cut CPL by about 44%, Meta's CPL is only about 28% lower than LinkedIn's, yet the revenue returns differ by a factor of more than two. If the number your team reports weekly is CPL, your budget will drift toward the channels that produce the cheapest form fills.

As Section 3.1 showed, paid CTR is meaningfully higher when your brand is cited in the AI Overview above the ads. Companies that treat AI visibility as purely an SEO concern are leaving paid performance on the table.

B2B fintech companies sell into some of the most scrutinized buyers in the economy, through ad platforms that restrict financial advertising, under rules that govern what they can say. The result is a premium on every stage of paid growth. We call it the FinTech Tax. It has four parts.

A warning about 'finance' benchmarks. Most published Google Ads benchmarks for 'Finance and Insurance' are dominated by consumer advertisers: insurance, personal loans, and credit cards. That category shows a high CTR (9.83%), a low CPC ($3.39), one of the lowest conversion rates of any industry (2.64%), and a CPL of $74.44 (WordStream/LocaliQ, 2026). Those numbers don't describe a B2B fintech selling to CFOs, banks, or credit unions. Benchmark against B2B SaaS data, and expect to sit above it on cost.

The platforms that work best for B2B restrict financial advertisers.

Every claim, testimonial, and badge on a paid landing page carries more risk for a fintech than for a typical SaaS company. Several rules are changing at once, in both directions.

Selling into banks and credit unions means selling through third-party risk management. The 2023 interagency guidance from the OCC, Federal Reserve and FDIC expects banks to manage vendor risk in proportion to risk and criticality, which in practice adds security reviews, due diligence questionnaires and procurement stages to every deal.

That may be about to ease. On September 11, 2026, the Federal Reserve, OCC, FDIC and, for the first time, the NCUA proposed replacing the 2023 guidance with a lighter, more risk-based approach that is friendlier to bank-fintech partnerships, including letting certifications cover parts of due diligence. Comments close November 16, 2026; until a final version is adopted, the 2023 guidance stays in force.

Meanwhile, demand from financial institutions is strong. In a survey of 416 US bank and credit union executives, more than 80% planned to increase technology spending in 2026, and about three-quarters were raising fraud budgets (Cornerstone Advisors, What's Going On in Banking 2026).

Benchmarks tell you where others are. They don't tell you what to do next. The Scale Test is the framework we use to decide whether a channel, campaign, or audience has earned more budget. It has five gates. A channel should pass all five before it gets scaled.

This benchmark combines publicly available, third-party research published between 2024 and October 2026, selected for relevance to US B2B SaaS and B2B fintech companies. We prioritized sources that disclose their sample size and data period, and sources that connect advertising spend to pipeline or revenue. Where a figure comes from a secondary write-up rather than the original report, or from a vendor with a commercial interest in the finding, we say so in the text.

We didn't combine figures from sources that measure different things, or calculate our own averages across sources. All figures are in US dollars unless marked in euros, which is how Dreamdata reports cost data.