Google Ads statistics 2026

Key takeaways

• Google advertising was $81.629B in Q2 2026. Search & other $63.271B (+17%). YouTube ads $11.055B (+13%). Network $7.303B. Source: Alphabet Exhibit 99.1.
• US search averages: CTR 6.64%, CPC $5.42, conversion rate 8.18%, cost per lead $66.69. Source: WordStream / LocaliQ, 13,474 campaigns.
• CPC is more than double 2016 ($2.32 → $5.42). CPL finally ticked down versus last year, still up versus 2016 ($59.18 → $66.69). Conversion rate rose in 87% of industries.
• Ecommerce is a different dataset. Triple Whale stores: CPA $28.14 (+9.96%), CVR 3.11% (-7.02%), ROAS 3.27. Do not quote 8.18% conversion on a Shopify brand.
• Free spy: Google Ads Transparency Center. Paid YouTube: AdPlexity YouTube. Cheap scan: BigSpy. Display publishers: AdBeat. Meta tools do not cover this auction.

What Alphabet actually reported

Most “Google Ads statistics 2026” pages still quote 2024 10-K totals or a blogger’s $340B guess. The filing that matters right now is Alphabet’s Q2 2026 earnings release, dated 2026-07-22. Consolidated revenue was $119.8 billion, up 24% year over year. Google Services — Search, YouTube, subscriptions, devices — was $94.5 billion, up 15%.

Inside Services, Alphabet splits advertising three ways. Search & other grew 17% to $63.271 billion. YouTube ads grew 13% to $11.055 billion. Google Network — the old AdSense / AdMob-class line — printed $7.303 billion, a touch below the year-ago $7.354 billion. Add those three and you get Google advertising of $81.629 billion for the quarter.

That mix is the briefing. Search is still the gravity well. YouTube is the line that grew like a product, not a side tab. Network is not where a new spy budget should start. If your stack is only Facebook and TikTok, you are researching the auctions that do not include $63 billion of intent inventory.

Source: Alphabet Inc. Q2 2026 earnings release, Exhibit 99.1. We cite the Exhibit, not a recap blog.

Q2 2026 vs Q2 2025 — same exhibit

Year-over-year is the only comparison that belongs next to a filing. Alphabet printed the prior-year advertising lines in the same table: Search & other $54.190B, YouTube ads $9.796B, Network $7.354B, Google advertising $71.340B. This quarter those four became $63.271B, $11.055B, $7.303B, and $81.629B.

Google Cloud is not in that advertising total. Cloud printed $24.8 billion, up 82%. Subscriptions, platforms, and devices sit inside Services next to ads. Mixing Cloud into “Google Ads revenue” is how a $340B listicle gets written. We do not do that here.

Why a spy desk cares

Revenue is not a trivia slide. It tells you which creative index is even allowed to be incomplete. A Facebook spy can miss YouTube pre-rolls and still be honest about its job. A “best ad spy tools” list that never opens Transparency Center is lying about coverage.

Search copy is not a Meta primary text. It is a query, a sitelink, a landing page that has already survived an intent auction. YouTube sequences are not Spark Ads. Duration still matters, but the unit is skippable in-stream, in-feed, and Shorts — labelled, or you are looking at a Facebook library with a play button sticker.

Network / AdMob is the third bucket. There is no Meta Ad Library for apps. AdPlexity Mobile and similar modules catch some in-app and mobile-web campaigns. They do not dump Google’s official AdMob network. We killed the old AdMob how-to on this site because it promised a library that does not exist. The honest path is Play/App Store research plus a mobile spy module, not a fake “AdMob Ads Spying” H1.

Search benchmarks you can actually use

WordStream by LocaliQ published the 2026 Google Ads benchmark study from 13,474 US-based search campaigns running 2025-04-01 through 2026-03-31. US-based search campaigns on Google Ads and Microsoft Ads. LocaliQ/WordStream, not Google’s own average. Search Engine Journal recapped the same dataset. Cross-industry averages:

Those four numbers are a guardrail, not a KPI. Legal clicks cost $9.87 and a lead costs $131.63. Arts & Entertainment clicks cost $1.63 with a 12.75% CTR. Animals & Pets converts at 16.22%. Finance & Insurance converts at 2.64%. If you brief a media buyer with “aim for 6.64% CTR,” you have not briefed them.

Year over year, WordStream says conversion rate rose in 87% of industries and average CPL fell for the first time in five years. That is the useful sentence. Costs are not “down.” Click prices are still historically high. The lead got slightly cheaper because more clicks converted, not because Google got generous.

The ten-year stretch is harsher. WordStream’s 2016 search averages were CTR 1.91%, CPC $2.32, conversion rate 2.7%, CPL $59.18. 2026 sits at 6.64%, $5.42, 8.18%, $66.69. CTR and conversion rate more than doubled. CPC more than doubled. CPL is still up versus 2016 even after this year’s dip. If a 2019 playbook is still open on your desk, the auction moved.

Source: WordStream by LocaliQ, 2026 Google Ads Benchmarks. Also: SEJ write-up of the same study.

How to read the industry table

Sort by the metric that matches the job, not the one that looks impressive in a screenshot. CPC high-to-low is the legal/home-services auction. CTR high-to-low is Arts & Entertainment and Travel — people click because the query is browsing, not because the landing page is a machine. Conversion rate high-to-low is Animals & Pets and auto repair: intent is already there. If you sell a $29 gadget, none of those three sorts is your brief.

Two rows to keep in your head. Attorneys & Legal: CTR 5.87%, CPC $9.87, CVR 5.55%, CPL $131.63. Arts & Entertainment: CTR 12.75%, CPC $1.63, CVR 5.91%, CPL $26.84. Same platform. Different sport. Finance & Insurance converts at 2.64% with a $3.39 click — cheap clicks that do not become leads. Animals & Pets converts at 16.22% with a $31.50 CPL. Quote the row. Never quote the average as a KPI.

The chart only plots the first twelve after sort so the labels stay readable. The table under it is the full 23-industry WordStream set. If your vertical is missing, you are closer to “Business Services” or “Shopping, Collectibles & Gifts” than to a made-up 24th row. Do not interpolate a number Google did not publish.

Ecommerce is not 8% conversion

Affiliates and DTC operators keep quoting lead-gen averages on product funnels. Triple Whale’s Google Ads ecommerce set (Aug 2025 – Jul 2026 vs prior 12 months, 21,000+ brands) is the correction: median CPA $28.14 (+9.96% YoY), CPM $15.35 (+13.34%), conversion rate 3.11% (-7.02%), ROAS 3.27 (-3.54%), CTR 1.65% (+13.78%), AOV $87.65 (+4.31%). Clicks got cheaper to notice and more expensive to buy. Efficiency slipped. Customers clicked more ads and converted less.

Use WordStream when the conversion is a form. Use Triple Whale when the conversion is a cart. Mixing them is how a $29 gadget store decides it is “underperforming” a dentist.

Source: Triple Whale Google Ads ecommerce benchmarks. Ecommerce stores only (21,000+ Triple Whale brands). Do not mix with WordStream’s all-industry search set.

How to spy Google ads without a fake library

There is no Facebook Ad Library for Search. Google Ads Transparency Center is the official surface: advertiser, region, whether ads ran. It will not give you a duration sort, a lander download, or a saved tracker. Paid tools earn the seat after that tab, not instead of it.

How to actually use it: search the legal entity, not the Facebook page name. Filter the country you buy. Open a creative and note the format — Search, YouTube, Display, Shopping, Demand Gen. If the advertiser is dark in your GEO, stop. A YouTube spy that returns last year’s pre-roll is a history lesson, not a brief. If they are live, write down the destination domain before you open a paid index. That domain is the query you take into AdPlexity YouTube, BigSpy, or AdBeat.

1. 1. Transparency Center. Search the legal name. Note the countries. If nothing is active, a paid YouTube spy will not invent inventory.
2. 2. YouTube paid index. AdPlexity YouTube when YouTube sits next to native or display. BigSpy for a cheap multi-network scan. Sort by duration, not view count.
3. 3. Display publishers. AdBeat if the question is “which sites actually run this advertiser.” That is not a social spy job.
4. 4. Do not buy Meta for Google. Tyver, AdSpy, and AdPlexity Social are Facebook/Instagram desks. PiPiADS is TikTok. The coupon on those pages will not make them index Search.

Live codes live on the deal pages, not in this article. YouTube module coupons are not the Social first-month code. See AdPlexity pricing and the coupon log.

Why the AdMob spy page is gone

The old URL promised a library Google does not run. AdMob inventory lives inside Google Network — $7.303 billion this quarter, -1% versus $7.354 billion a year ago. That line is AdSense, AdMob, and third-party publisher demand. It is not a public creative archive with a duration filter.

You can still research mobile ads. Play Store / App Store listing plus a mobile spy module (AdPlexity Mobile, sometimes BigSpy’s app tab) is the honest stack. Pretending AdMob has a Meta-style library is how this site used to waste a ranking URL. That permalink now 301s here. If you landed from an old “AdMob ads spying” backlink, this is the page that replaced it.

A spy brief that uses these numbers

Do not paste “average CPC $5.42” into a Slack. Write the industry row, the dataset, and the job. Example: “US search, Attorneys row, WordStream 2026 (Google + Microsoft mix). Target CTR above 5.87%, expect CPC near $9.87, CPL near $131.63. That is lead-gen. If we switch the offer to a Shopify product, Triple Whale ecommerce medians apply instead: CPA $28.14, CVR 3.11%, ROAS 3.27.”

Then name the index. Search copy: Transparency Center, then the destination. YouTube: AdPlexity YouTube or BigSpy, sort by duration. Display: AdBeat. Meta funnels stay on AdPlexity Social / Tyver / AdSpy — different auction, different coupon. The coupon log is dated so you can cite the check, not a screenshot from 2024.

If the brief is native or push, leave this page. Anstrex and AdPlexity Native do not care what Search CPC did in Q2. Using Google Ads statistics to pick a Taboola spy is the same error as using Facebook Ad Library to judge YouTube.

What a “good” number means on this desk

A good CTR is above your industry row in the table, with the same match types you actually buy. A good CPC is the one that still clears CPA after the landing page does its job. WordStream’s $5.42 is the mean of a US search set that includes attorneys at $9.87 and restaurants at $2.05. Publishing “average CPC $5.42” as a target is how amateur media plans get written.

A good conversion rate is closer to 8% on lead-gen search and closer to 3% on ecommerce, if Triple Whale’s median is the peer set. If your offer is nutra on native, neither number applies — that traffic is Anstrex / AdPlexity Native, not this page.

A good spy workflow for Google is: confirm the advertiser is live, collect the YouTube cuts that have actually been paying, follow the destination, rebuild. Same discipline as Meta. Different index.

Sources

We do not blend analyst “$318B FY2026” forecasts into these tables. Forecasts go stale the next quarter. Filings and named benchmark studies stay.

FAQ