2026 Headline Numbers at a Glance
Before diving into platform and industry breakdowns, here are the cross-platform averages every performance marketer should have in their head going into 2026.
Sources: WordStream 2026 Google Ads Benchmarks (13,000+ campaigns), Meta Ads data aggregated from OwlClaw/AdEspresso, Triple Whale e-commerce panel (18,000+ brands). Numbers represent cross-industry medians — see per-industry breakdowns below.
What Changed in 2026
2026 is not a copy of 2025. The cost environment, platform dynamics, and measurement landscape have all shifted in ways that change how you should read benchmark data.
CPC stabilized after two years of steep increases. Google CPC rose only modestly in 2026 after 12.88% increases in 2025 and 10% in 2024. Meta CPC remains low at $1.72 but CPL for lead campaigns rose 20% year-over-year. LinkedIn now captures 41% of B2B paid social budgets and posts ROAS 121% above industry median for B2B advertisers. TikTok remains cheapest on CPC but intent variability makes CPA comparison difficult. AI-driven search is beginning to suppress Google CTRs in informational categories — performance-intent queries are less affected so far.
📈 Rising costs
Google CPL up ~5% YoY on average. Meta CPL for lead campaigns up 20%. LinkedIn CPM up as B2B budgets concentrate on the platform. E-commerce median CPA rose 12.35% to $23.74.
📉 Declining CVR pockets
Meta lead campaign CVR fell for 80% of industries, partly economic, partly audience fatigue. Google CVR held or improved for 65% of industries. TikTok CVR highly variable by creative quality.
🔍 AI search impact
Gartner projects traditional search volume down 25% by end of 2026 as buyers shift to AI tools. Informational content losing traffic fastest. Performance/transactional intent queries holding well.
📊 Measurement fragmentation
Platform-reported ROAS increasingly diverges from true incrementality. More advertisers running holdout tests. Attribution gaps between platforms mean benchmark ROAS should be treated as directional, not definitive.
CPC Benchmarks by Platform — 2026
Cost per click varies by an order of magnitude across platforms. LinkedIn at $5.26 and Google Search at $5.42 sit at the top. TikTok and Pinterest remain the cheapest on a raw CPC basis — but cheap clicks don't always mean cheap conversions.
Google Search Ads Highest intent
| Industry | Avg. CPC | YoY Trend |
|---|---|---|
| Legal | $6.75 | ↑ +8% |
| Finance & Insurance | $5.90 | ↑ +6% |
| B2B / SaaS | $3.80 | ↑ +5% |
| Healthcare | $4.10 | ↑ +4% |
| E-commerce | $1.85 | → Flat |
| Travel | $1.75 | ↓ −3% |
| Consumer Goods | $1.30 | → Flat |
| All industries avg. | $5.42 | ↑ Modest |
Meta Ads (Facebook + Instagram) Volume + retargeting
Meta CPC of $1.72 makes it 3× cheaper than Google Search on a per-click basis. But conversion rates differ — use CPA, not CPC, as your primary comparison metric across platforms.
| Objective | Avg. CPC | Note |
|---|---|---|
| Traffic campaigns | $0.94 | Lowest CPC, lower intent |
| Lead generation | $1.92 | CPL up 20% YoY |
| Conversions / Sales | $1.72 | Global average |
| India average | ₹8–18 | ~85% cheaper than global avg |
LinkedIn Ads B2B premium
LinkedIn's $5.26 average CPC is high, but the platform now captures 41% of B2B paid social budgets and posts ROAS 121% above industry median for B2B deals. Expensive per click, but often the right audience.
| Targeting type | Avg. CPC |
|---|---|
| Job title targeting | $5–$8 |
| Company size targeting | $6–$10 |
| Skills targeting | $4–$7 |
| Enterprise / C-suite | $10–$15+ |
TikTok & Pinterest Low CPC, variable intent
| Platform | Avg. CPC | Best use case |
|---|---|---|
| TikTok Ads | $0.90 | D2C, impulse, younger demos |
| Pinterest Ads | $0.50 | Home, lifestyle, high-intent browse |
| Microsoft Ads | $1.54 | Older/professional, less competition |
CPM Benchmarks by Platform — 2026
CPM (cost per thousand impressions) is the primary metric for awareness campaigns and programmatic buying. Rates vary dramatically by platform, audience quality, and ad format.
| Platform | Avg. CPM | High-End CPM | Best for |
|---|---|---|---|
| Google Display | $2.80 | $6–$8 | Reach, remarketing |
| YouTube | $10–$18 | $25+ | Brand, consideration |
| TikTok | $4–$8 | $12 | Awareness, D2C |
| Meta (Facebook/Instagram) | $7.19 | $14+ | Retargeting, broad reach |
| $20–$45 | $80+ | B2B awareness | |
| X (Twitter) | $5–$10 | $18 | News, trending topics |
| $3–$6 | $10 | Lifestyle, home, fashion |
LinkedIn's CPM looks alarming until you calculate effective CPA. A $40 LinkedIn CPM at 0.5% CTR and 5% CVR = $160 CPA — expensive, but not unreasonable for enterprise B2B deals worth $50K+.
CPM Calculator → Platform CPM breakdown →CPA Benchmarks by Industry — 2026
CPA (cost per acquisition) is the metric that ties everything together. It's what you actually pay per conversion after all your CPC, CTR, and CVR factors compound. The 2026 picture: median e-commerce CPA rose 12.35% to $23.74 as CPMs climbed and CVR fell.
Google Ads CPA by Industry
| Industry | Avg. CPA (Search) | Avg. CPA (Display) | Conv. value range |
|---|---|---|---|
| Legal Services | $86 | $65 | $1,000–$10,000+ |
| B2B / SaaS | $116 | $88 | $500–$50,000 LTV |
| Finance & Insurance | $78 | $56 | $500–$5,000 |
| Healthcare | $78 | $60 | $200–$2,000 |
| Real Estate | $116 | $74 | $5,000–$50,000 |
| Education | $72 | $143 | $500–$20,000 |
| E-commerce | $45 | $65 | $50–$300 |
| Travel | $44 | $60 | $300–$3,000 |
| Entertainment | $21 | $60 | $10–$100 |
E-commerce CPA by Category (Google + Meta, 2026)
| Category | Median CPA | YoY Change |
|---|---|---|
| Pets & Animals | $25.15 | ↓ −4% |
| Health & Wellness | $26.80 | ↑ +8% |
| Home & Garden | $31.50 | ↑ +11% |
| Electronics | $33.92 | ↑ +16% |
| Travel Accessories | $28.23 | ↑ +41% |
| All e-commerce median | $23.74 | ↑ +12.35% |
ROAS Benchmarks by Platform and Objective — 2026
ROAS (return on ad spend) is the revenue-side metric that completes the performance picture. A 3× ROAS means $3 returned for every $1 spent. What counts as "good" ROAS depends entirely on your gross margin — a 3× ROAS on a 70% margin business is profitable; on a 20% margin business it isn't.
| Platform / Context | Average ROAS | Top Quartile | Note |
|---|---|---|---|
| Google Search (e-comm) | 3.5–4.5× | 7–10× | High intent = high close rate |
| Meta (e-comm, prospecting) | 2–3× | 5–6× | Retargeting: 6–10× |
| Meta (D2C blended) | 3–4× | 6–8× | Prospecting + retargeting combined |
| LinkedIn (B2B) | 2.21× median | 121% above industry median | Highest B2B ROAS in paid social |
| TikTok (D2C) | 1.5–3× | 5×+ | Highly creative-dependent |
| YouTube (brand + retargeting) | 1.5–2.5× | 4× | Long attribution windows needed |
CTR Benchmarks by Platform — 2026
Click-through rate tells you how compelling your ads are to the people who see them. In Google Ads, CTR also drives Quality Score, which in turn lowers your CPC. In Meta, higher CTR means lower CPM through the auction.
| Platform / Format | Avg. CTR | Good CTR | Trend |
|---|---|---|---|
| Google Search | 3.17% | 5%+ | → Stable |
| Google Display | 0.46% | 0.75%+ | → Stable |
| Meta (all formats) | 0.90% | 1.2%+ | ↑ Traffic improving |
| LinkedIn (sponsored content) | 0.39% | 0.6%+ | → Stable |
| TikTok | 0.84% | 1.5%+ | ↑ Format-dependent |
| Display/Programmatic | 0.35% | 0.5%+ | ↓ Banner blindness |
The False Efficiency Trap — When Good-Looking Benchmarks Signal a Problem
One of the most dangerous patterns in performance marketing: a campaign where every metric looks healthy — CPC is low, CPA is below benchmark, ROAS is above target — but actual business outcomes are deteriorating. This is the False Efficiency Trap.
It occurs when platform optimisation achieves metric efficiency by sacrificing reach, volume, or incrementality. The account wins the benchmark race while losing the revenue race.
| What Looks Good | What Is Actually Happening | The Hidden Cost |
|---|---|---|
| CPA drops 30% month-on-month | Algorithm concentrating spend on lowest-funnel, highest-intent audiences only | New customer acquisition has stopped — only retargeting converting |
| ROAS increases to 6× | Branded search capturing conversions that would have happened organically | You are paying to intercept your own organic demand |
| CPC falls 40% after switch to broad match | Traffic mix has shifted to low-intent adjacent queries | Volume increased, conversion rate collapsed, true CPA higher |
| Impressions up 3× with CPM flat | Frequency on existing audiences exceeds 12× — diminishing marginal returns | Impression count is vanity metric; incremental reach is near zero |
If CPA is improving while revenue volume is flat or declining, you are likely in the False Efficiency Trap. The correct response is to expand audience targeting or increase budget — not to optimise further. Tightening a well-optimised campaign often accelerates the trap.
How Benchmarks Actively Mislead — The Five Distortions
Benchmarks are indispensable as calibration tools and actively dangerous as targets. The distinction matters because most advertisers use them as targets. Understanding the five systematic ways benchmarks distort reality turns a blunt instrument into a precise one.
1. They blend incompatible businesses. The $45 ecommerce CPA average includes brands with 20% margins where $45 CPA is catastrophic, and brands with 65% margins where $45 CPA is excellent. The businesses are in the same vertical but different universes economically. The average of two opposite situations describes neither. Use break-even CPA from your own margin — not the average.
2. They use last-click attribution by default. Most published benchmarks reflect last-click measurement, which overvalues intent-capture channels (Search, retargeting) and undervalues intent-creation channels (YouTube, Display, social prospecting). A benchmark comparison using last-click attribution will consistently make your awareness campaigns look expensive and your branded search look cheap — regardless of which is actually driving more business value. Check which attribution model the benchmark uses before applying it.
3. They don't adjust for account maturity. A new campaign targeting cold audiences has structurally higher CPA than a mature campaign with two years of conversion data feeding Smart Bidding. Both may be in the same vertical with the same offer. Published benchmarks blend both — which makes new campaigns look "above average" and mature campaigns look "at average" even when both are performing correctly for their stage. Compare against vertical benchmarks for accounts of similar maturity.
4. They compress geographic variance. US CPMs and CPAs are the highest globally. A benchmark table that reports "global average" is dominated by US data because US ad spend represents the majority of global digital advertising volume. Advertisers in Germany, Australia, or Turkey running against "global benchmarks" are comparing against a number heavily weighted toward the most expensive advertising market in the world. Use region-specific benchmarks when available.
5. They don't distinguish campaign objective. A Google Ads campaign with a Conversions objective and a campaign with a Traffic objective will produce very different CPMs and CTRs for the same audience — because one is paying for conversion probability and one is paying for click probability. Benchmark tables rarely specify which objective the data represents. Applying a conversion-campaign CPM benchmark to a traffic campaign (or vice versa) produces incorrect expectations and incorrect diagnosis when results diverge.
The Five Benchmark Distortions: (1) Economic incompatibility — different margins, same vertical. (2) Attribution model mismatch — last-click default understates demand creation. (3) Account maturity variance — new vs mature account baselines. (4) Geographic compression — US-weighted averages for non-US advertisers. (5) Objective confusion — conversion vs traffic campaign data blended. Applying benchmarks without accounting for these produces confident decisions based on irrelevant comparisons.
Five Benchmark Rules Every Operator Should Apply
Most practitioners compare their metrics to industry averages and draw the wrong conclusions. These five rules prevent the most common benchmark misreads.
Rule 1 — Your Break-Even Overrides the Industry Average
If the industry average CPA is $45 and your break-even is $32, the benchmark is irrelevant — your economics are the constraint. Conversely, if your break-even is $90, a CPA of $60 is excellent regardless of what the average says. Calculate break-even CPA before looking at any benchmark table.
Rule 2 — Compare Same Periods, Not Calendar Averages
Q4 Meta CPMs average 50–80% higher than Q1. A January campaign running above the annual average CPM is not underperforming — it is operating in a different market. Always compare your metrics against the same calendar period from the prior year, not against annual averages.
Rule 3 — Platform CPC Benchmarks Are Meaningless Without CVR
LinkedIn CPC averaging $6 versus Meta CPC at $1.80 does not make LinkedIn three times more expensive. If LinkedIn converts at 9% and Meta at 2.5%, LinkedIn CPA is actually lower. Compare platforms at the CPA level — never at CPC in isolation.
Rule 4 — Branded Search Distorts Every Account-Level Metric
A Google Ads account running both branded and non-branded search will show a blended CPA that understates true non-brand acquisition cost by 30–60%. Always segment branded vs. non-branded before benchmarking. An account with $12 branded CPA and $75 non-branded CPA has an average of $35 — which is meaningless for acquisition planning.
Rule 5 — Platform-Reported ROAS Is Not Incremental ROAS
Every platform measures ROAS using its own attribution model, which systematically overcredits its own channel. Meta Ads Manager ROAS at 4.2× may represent 2.0–2.5× true incrementality when verified via holdout test. Use platform ROAS for relative performance tracking — not for absolute efficiency judgements or cross-platform budget allocation.
What Actually Changed in Performance Marketing in 2026
Beyond headline benchmark numbers, four structural shifts in 2026 are changing how benchmarks should be interpreted — not just what the numbers are.
| Structural Shift | Mechanism | Benchmark Impact | Operator Response |
|---|---|---|---|
| AI Search fragmenting intent signals | ChatGPT, Gemini, Perplexity intercepting top-of-funnel queries before Google | Google Search impression volume down for informational queries; intent density of remaining traffic higher | CPC benchmarks trending up for remaining high-intent queries; update CAC models accordingly |
| Performance Max displacing Search | Google defaulting new campaigns to PMax, mixing branded search, Shopping, Display, and YouTube | Blended PMax ROAS looks strong; Search-only CPA benchmarks less representative of most accounts | Segment PMax asset group performance; isolate branded vs. non-branded conversion contribution |
| Meta Advantage+ compressing audience control | Advantage+ Shopping removing manual audience controls; algorithm self-selecting audiences | CPM variance compressed — fewer outlier low-CPM campaigns; floor prices elevated | Monitor frequency carefully; cold and warm audiences now harder to separate in reporting |
| LinkedIn Accelerate campaigns | AI-managed campaigns with automated bidding and audience expansion | Accelerate CPC often 20–35% below Manual campaigns initially; long-term CPA convergence observed | Test Accelerate vs. Manual CPC in parallel; evaluate on CPA not CPC over 6-week window |
How to Interpret Benchmark Data — Without Getting Misled
Benchmarks are a reference point, not a target. Most practitioners misuse them. Here's how to read them correctly.
Cross-industry averages hide the real picture
A $5.42 Google CPC average blends legal ($6.75) with entertainment ($0.78). If you're in legal, the average is meaningless — your actual range is what matters. Always pull your industry-specific row.
Platform CPC comparisons without CVR are useless
LinkedIn at $5.26 looks 3× more expensive than Meta at $1.72. But if LinkedIn converts at 8% and Meta at 2%, LinkedIn CPA is actually lower. Always compare at the CPA level, not CPC.
ROAS benchmarks assume nothing about your margins
A 3.7× ROAS benchmark tells you nothing unless you know whether 3.7× is above or below your break-even point. Calculate your break-even ROAS first (1 ÷ gross margin), then evaluate your actual ROAS against that.
Seasonality creates false benchmarks
Q4 CPMs on Meta can be 40–80% higher than Q1 averages due to advertiser competition. A January CPA that looks "bad" against a Q4 benchmark may actually be excellent. Compare same periods year-over-year, not quarter to quarter.
Platform Cost Inflation: Why Every Benchmark Is Rising
2026 benchmarks are higher than 2024 across almost every platform and metric. This isn't random — it reflects three structural forces that are unlikely to reverse.
1. Privacy-driven budget concentration
Cookie deprecation and iOS tracking restrictions since 2021 forced advertisers away from third-party-data-dependent display and toward first-party-data-rich platforms: Google Search (intent), Meta (logged-in identity), LinkedIn (professional identity). The same advertiser dollars now compete in smaller, higher-quality inventory pools. Prices rise when demand concentrates and supply doesn't expand proportionally.
2. AI bidding driving up floor prices
Every major platform now defaults to AI-powered bidding: Google's tCPA/tROAS, Meta's Advantage+ Shopping, LinkedIn's Accelerate. These systems are designed to win auctions, not minimize cost — they bid up to the maximum value the algorithm believes a click is worth. When most advertisers use AI bidding simultaneously, the auction clears at the highest price most bidders will accept. Manual CPC is often 20–40% cheaper, but requires more active management and data.
3. B2B budget reallocation to LinkedIn
LinkedIn's 41% share of B2B paid social budgets in 2026 represents a massive concentration compared to 2019–2020. As more B2B advertisers discovered that LinkedIn's first-party professional data was immune to cookie deprecation, they shifted budgets onto the platform. More B2B budget on LinkedIn = structurally higher LinkedIn CPCs. This reallocation is largely complete — LinkedIn CPCs will likely stabilize at current elevated levels rather than rising sharply again.
Attribution Fragmentation: Why Your ROAS Number Is Probably Wrong
The single most underappreciated problem in 2026 performance marketing is attribution. Every platform reports ROAS based on its own measurement model — and they all overclaim.
The multi-touch problem
A customer sees a LinkedIn ad on Monday, searches Google on Wednesday, clicks a Meta retargeting ad on Friday, and converts via direct traffic on Saturday. Google attributes the conversion to Wednesday's search. Meta attributes it to Friday's click. LinkedIn attributes it to Monday's impression view. Your total reported conversions across platforms may be 2–3× your actual sales — a phenomenon called attribution overlap.
What this means for benchmarks
Platform-reported ROAS benchmarks — including most published industry averages — are based on each platform's own attribution model. A "3.7× ROAS" figure from Meta's Ads Manager may represent 2.0× true incrementality when measured via holdout test. This doesn't mean the campaign isn't valuable — it means benchmark ROAS should be treated as directional signals, not absolute truths.
The gap is widening in 2026
iOS 14+ restrictions, cookie deprecation, and the rise of AI search are all degrading signal quality. More conversions are happening in environments where platform attribution cannot track them — offline, via AI assistant, or after long consideration windows. The practical result: reported ROAS is increasingly optimistic relative to true incrementality, and the gap is larger in 2026 than in 2023.
Apply These Benchmarks to Your Numbers
Every metric above has a free calculator. Put in your actuals and see exactly where you stand.
CPC Calculator CPM Calculator CPA Calculator ROAS Calculator CTR Calculator Research keywords & competitors with Mangools →Benchmark Deep Dives
Each metric and platform has its own dedicated benchmark page with full industry and geography breakdowns:
- Average CPC by Platform — Google, Meta, LinkedIn, TikTok, Pinterest compared
- Average CPM by Platform — Programmatic, social, and video CPM benchmarks
- Average ROAS by Platform — E-commerce and lead gen ROAS by channel
- Average CPA for Google Ads — 2026 Search, Shopping and Display CPA by industry
- Average CPA by Industry — Legal, SaaS, e-commerce, healthcare and more
- Average CPA in the USA — Google Search benchmarks by vertical
- Average CPC in the USA — Google $2.50–$6, Meta $0.80–$2.20
- Average CPC in India — 80–90% cheaper than US benchmarks
- Average LinkedIn CPC — $5.26 global, up to $15+ for enterprise targeting
- How to Improve ROAS — Tactical guide with 34 internal links
- Best Tools to Improve ROAS — Attribution, creative, CRO stack guide
- LinkedIn CPC Reduction ChecklistAverage CPC Consumer GoodsAverage CPC EntertainmentHow to Calculate CPMHow to Calculate CACHow to Calculate ROASCPM vs CPAWhy Is My CPM High?Google Ads Cost 2026Google Ads vs Meta Ads — 7 fast fixes
What Actually Changed in 2026 — Three Structural Shifts
Most benchmark updates adjust numbers slightly. Three things genuinely changed the landscape in 2025–2026 in ways that make prior-year comparisons misleading.
1. Performance Max became the default for most Google advertisers
A 2026 "Google CPA" without knowing the campaign type mix is an unreliable benchmark. PMax blends Shopping, Display, YouTube, and Search — its CPA is a different number from Search-only or Shopping-only CPA. When comparing against benchmarks here, verify whether your account's CPA is Search-only or PMax-blended. Brand-excluded PMax CPA runs 30–50% higher than reported PMax CPA for most accounts.
2. Meta's tracking gap is now measurable — and large
The benchmark gap between fully CAPI-implemented accounts and browser-pixel-only accounts is 20–40% CPA difference for the same audience and creative. The 2026 Meta benchmarks here reflect accounts with adequate tracking. If your CAPI is not implemented, your real CPA will be above these benchmarks.
3. TikTok's CPM arbitrage closed
TikTok CPMs in competitive categories now sit within 20–30% of Meta CPMs — compared to 50–60% cheaper in 2022. The 2026 CPA advantage TikTok has over Meta is now primarily creative-driven, not auction-driven. Advertisers repurposing Meta creative for TikTok see converging CPAs; those producing TikTok-native UGC still see meaningful efficiency advantage.