AI in Advertising: Statistics & Benchmarks — Q3 2026 Roundup

What the AI advertising numbers say in Q3 2026

Two sets of numbers exist in AI advertising. Platforms publish uplift percentages from their own internal tests. Agencies see what happens inside real accounts. The two rarely match. Every figure below is labeled by source so you can tell which one you are reading.

The largest single data point this quarter came from Alphabet’s Q2 2026 earnings call on July 22, 2026. Google chief business officer Philipp Schindler told investors that advertisers adopting AI-powered campaigns like AI Max or Performance Max "see an average of 15% more conversions or value on Search at a similar ROAS." He also confirmed that AI Max had exited beta with 500,000 advertisers using it.

That 15% is Google measuring Google. It is useful context. It is not a forecast for your account. Benchmark tables from iClick-managed accounts appear further down, and they are labeled as first-party portfolio data, not market averages.

What changed between Q2 and Q3 2026 is the framing. A year ago the argument was whether AI campaign types worked at all. Now the argument is about attribution: how much of the reported lift is new revenue, and how much is revenue moved from one campaign to another.

What counts as an AI ad in 2026

The phrase covers five distinct things. Mixing them up is why so many AI advertising statistics look contradictory:

• Automated bidding. Target CPA, Target ROAS, and Maximize Conversions in Google Ads and Meta.
• Keywordless matching. AI Max for Search campaigns, broad match expansion, and Dynamic Search Ads.
• Asset generation. Google’s automatically created assets and Meta’s Advantage+ creative tools.
• Budget and channel allocation. Performance Max and Advantage+ shopping campaigns.
• Audience and ranking systems. Meta’s Andromeda retrieval model and Google’s Demand Gen targeting.

Almost no account is fully manual now. Even a tightly built keyword campaign runs on Smart Bidding. So adoption statistics measure depth of automation, not presence of it. Our AI in PPC playbook covers which of those five layers is worth handing over first.

Adoption statistics: how many advertisers actually use AI

The IAB published "The AI Ad Gap Widens" on January 15, 2026, with Sonata Insights. It found that 83% of ad executives say their company has deployed AI in the creative process. That is up from 60% in 2024, a 23 point jump in two years.

The sample matters. IAB surveyed 104 ad industry executives and 505 US Gen Z and Millennial consumers between October 2025 and January 2026. It is a small executive sample. Treat the direction as solid and the decimal points as soft.

The motivation shifted too. Cost efficiency is now the top cited benefit at 64%, up from fifth place in 2024. Two years ago teams adopted AI tooling for speed and output volume. Now they adopt it to defend margin.

What that 83% does not tell you is depth. An account that uses one AI headline generator counts the same as an account running everything through Performance Max. Survey adoption rates and budget share are different questions.

Platform-side numbers are larger and more concrete. On Meta’s Q2 2026 earnings call, held July 29, 2026, CFO Susan Li said Advantage+ reached "over $75 billion in annual revenue run rate" that quarter. Meta also reported 9 million small businesses using at least one AI ad creative tool, with image generation adoption more than doubling in the quarter.

Google’s comparable figure is the 500,000 advertisers on AI Max. Set against Meta’s 9 million creative-tool users, that gap says something simple. Creative automation spreads faster than targeting automation, because advertisers read it as lower risk.

Performance statistics: what the platforms claim

Below is every quantified AI performance claim the two largest platforms published as of Q3 2026, with its source and date.

Read the fine print on all of it. Google’s own help documentation tags the 14% figure as "Google internal data, 2025" and limits it to non-Retail advertisers. The 27% figure applies only to campaigns where more than 70% of conversions came from exact or phrase match keywords.

There is also a selection problem. Advertisers who test a new AI feature and see it fail usually switch it off and never appear in a case study. Named wins like L’Oréal and AAA Auto Club Enterprises are real results. They are not a random sample.

Meta publishes far less quantified performance data for Advantage+ than Google does for AI Max, which makes side-by-side comparison hard. Meta’s own numbers point the same way: in an April 14, 2023 Meta for Developers post announcing Advantage+ creative standard enhancements in the Marketing API, Meta reported that “ads that were opted-in to standard enhancements delivered a 4% reduction in cost-per-result” for campaigns optimizing for link clicks, landing page views, or offsite conversions, a figure Meta attributes to one month of experiment data across 154,000 ad accounts (May 2 to 30, 2022).

The honest read on the numbers we can verify: platform AI moves performance by low double digits when the starting account was loosely built. It moves performance much less, and sometimes backwards, when the starting account was already tight. We went through that split in detail in Will AI Replace PPC Agencies? An Agency’s Honest Answer.

Reach and spend statistics: where the AI budget goes

Reach is the quiet reason AI campaign types perform at all. Google Ads Help states that Demand Gen reaches 3 billion monthly active users, with 50 billion global daily views on Shorts. Much of that inventory is only reachable through AI-allocated campaign types.

On the money side, eMarketer’s US AI Advertising Forecast 2026, published June 4, 2026, projects US AI ad spending will reach $68.25 billion by 2030. That is more than double the 2026 level.

The same forecast corrects a prediction most 2026 write-ups got wrong. More than 80% of AI advertising in 2026 sits next to AI content, not inside chatbots. Search ads running alongside AI Overviews carry the spend. Ads inside standalone assistants are still a rounding error.

The practical translation: your 2027 AI ad budget is mostly your existing Google and Meta budget, routed through newer campaign types. For how each platform stacks those types, see Google’s AI Ads Stack, Explained: AI Max, Performance Max, and Demand Gen and Meta’s AI Ads Stack, Explained: Advantage+ and the Andromeda Algorithm.

2026 paid media benchmarks from iClick-managed accounts

The figures in this section come from accounts iClick manages, covering 2024 to 2025. They are first-party portfolio data, published on our 2026 paid media benchmarks page. They represent well-managed accounts, not market medians, so they read tighter than a broad survey average.

Google Search

• CPC: $2.32, up 12% year over year
• CTR: 4.23%, up 0.4 percentage points
• Conversion rate: 3.48%, flat
• Quality Score: 6.2 out of 10

CPC up 12% with conversion rate flat is the core 2026 problem in one line. AI matching found more queries. It did not make those queries convert better. Cost per acquisition held roughly steady only because CTR improved slightly.

That pattern also explains why so many advertisers report a conversion lift and a budget squeeze in the same quarter. Volume rose. Efficiency per click did not.

Meta

• CPM: $11.54, up 38% year over year
• CPC: $0.70, up 8%
• CTR: 0.90% and declining, driven by creative fatigue
• Conversion rate: 1.85%, a post-iOS 14 baseline

A 38% CPM increase against a declining CTR is why creative volume now matters more than targeting on Meta. Advantage+ creative tools help with volume. They do not fix a weak offer or a tired hook.

Shopping

• CTR: 0.86%
• Conversion rate: 1.91%, up 0.2 percentage points
• ROAS: 480% in well-managed accounts
• Feed disapproval rate: 12% in unmanaged accounts

That 12% feed disapproval figure is the least interesting statistic in this post and often the most expensive. An AI campaign cannot bid on a product it cannot see.

Average CPC and conversion rate by industry

The spread is the point. A personal injury click at $6.75 costs 8.4x a music and entertainment click at $0.80. One AI uplift percentage means completely different money across those two accounts. Legal converts at 6.98%, which is the only reason that CPC clears.

The Performance Max cannibalization number

One first-party figure from our benchmark work draws the most pushback: Performance Max cannibalizes Shopping in more than 80% of the accounts we review. PMax reports the conversion. Standard Shopping would have captured much of that demand anyway, often at a lower cost.

This is the most common reason a reported AI uplift is not a real business uplift. Campaign-level reporting improves while total account revenue holds flat. If you only read the campaign view, the automation looks brilliant.

Account-level results when AI campaigns are managed properly

Platform statistics describe averages. Named accounts describe outcomes. Every client below is on record in our case studies:

Across the portfolio, iClick has managed more than $61 million in ad spend since 2017 for 250+ clients, with a 41% average lift on primary conversion goals. Those are portfolio numbers, not per-campaign promises.

The pattern in the eCommerce accounts is consistent. Wish Rock Relaxation and We Love Eyes both hold 10x ROAS, roughly double the 480% Shopping benchmark. Neither got there by switching on Advantage+ and waiting for the algorithm to sort it out.

On the legal side, McEldrew Purtell cut CPA 53% and grew conversions 697% in 90 days while doubling call volume. Georgia Probate Law Group cut cost per signed case 40% and grew qualified intakes 85% across two states. Both accounts use automated bidding. Neither uses automated targeting without hard guardrails. Our AI PPC management service is built around that division of work.

The statistic that never makes it into the deck

IAB found a 37 point perception gap. 82% of ad executives believe consumers feel positive about AI-generated ads. Only 45% of consumers actually do. Among Gen Z, 39% feel negative, compared with 20% of Millennials.

The disclosure data is more encouraging. 89% of advertisers using generative AI disclose it, though fewer than half always do. And 73% of younger consumers said disclosure would either increase their purchase likelihood or not affect it at all.

So labeling AI creative costs almost nothing in measured performance, and the downside of getting caught without a label is real. For consumer brands running generated video at scale, that is the cheapest risk reduction available in 2026.

How to read AI advertising statistics without getting burned

1. Check who measured it. A 15% uplift published by Google is Google grading its own homework.
2. Check the baseline. Uplift claims compare against the advertiser’s previous setup, which may have been poorly built.
3. Check the denominator. A 27% lift on a $500 monthly budget is not evidence for a $50,000 one.
4. Check the metric. "Conversions or conversion value" lets a platform report whichever number moved.
5. Check for cannibalization. Compare total account revenue against last year, not campaign-level reporting.

Run those five checks on any AI advertising statistic, including every one in this post. The figures that survive all five are the ones worth building a budget around.

Find out what your own numbers say

Benchmarks tell you whether your CPC and ROAS sit inside a normal range. They cannot tell you whether Performance Max is quietly eating your Shopping revenue, or whether 12% of your product feed is disapproved right now. That takes someone looking inside the account.

iClick runs a 47-point written PPC audit and delivers it as a PDF within 5 business days. No sales call required. Request your free written audit and we will show you which of your AI campaigns are producing real incremental revenue, and which are reporting someone else’s.

Related on iClick

• Will AI Replace PPC Agencies? An Agency’s Honest Answer
• Meta’s AI Ads Stack, Explained: Advantage+ and the Andromeda Algorithm

Sources

• Alphabet Q2 2026 earnings call transcript, July 22, 2026 (Philipp Schindler on AI Max and Performance Max)
• Meta Platforms Q2 2026 earnings call transcript, July 29, 2026 (Advantage+ run rate, AI creative tool adoption)
• IAB and Sonata Insights, The AI Ad Gap Widens, January 15, 2026
• Google Ads Help, About AI Max for Search campaigns (Google internal data, 2025)
• Google Ads blog, Unlock next-level performance with AI Max for Search campaigns, May 6, 2025
• Google Ads Help, About Demand Gen campaigns (reach figures)
• eMarketer, US AI Advertising Forecast 2026, June 4, 2026
• iClick Advertising, 2026 paid media benchmarks (first-party data from iClick-managed accounts, 2024 to 2025)
• iClick Advertising case studies (named client results)
• Meta for Developers: “Advantage+ creative standard enhancements now available in Marketing API” (April 14, 2023)

Frequently asked questions

What is the most reliable AI advertising statistic in 2026?

Google’s own figure, stated on the Alphabet Q2 2026 earnings call on July 22, 2026: advertisers using AI Max or Performance Max see an average of 15% more conversions or value on Search at a similar ROAS. It is a vendor number, but it is dated, on the record, and consistent with Google’s 14% help-center claim.

Do AI ad campaigns actually reduce cost per acquisition?

Sometimes, and mostly in accounts that were loosely built beforehand. Google reports MyConnect at 13% lower CPA and L’Oreal at 31% lower cost per conversion. In iClick-managed accounts, Google Search CPC rose 12% year over year while conversion rate stayed flat at 3.48%, so CPA held rather than dropped.

What is a good ROAS for an AI-managed eCommerce account in 2026?

Shopping ROAS in well-managed iClick accounts runs at 480%, roughly 4.8x. By industry, fashion averages 320% and health and beauty 390%. Top named accounts go far higher: Wish Rock Relaxation and We Love Eyes both hold 10x ROAS. Treat 4x to 5x as competent and anything above 8x as an outlier worth studying.

Does Performance Max cannibalize Shopping campaigns?

In more than 80% of accounts iClick reviews, yes. PMax claims conversions that standard Shopping would have captured anyway, often at lower cost. The campaign report improves while total account revenue stays flat. Always check year over year account revenue before crediting PMax with a lift.

How much AI ad spend goes inside chatbots?

Very little so far. eMarketer’s US AI Advertising Forecast 2026, published June 4, 2026, found more than 80% of AI advertising in 2026 appears next to AI content rather than inside chatbots. The forecast puts US AI ad spending at $68.25 billion by 2030, most of it flowing through search and social placements.