How often are these numbers updated?

When the publisher issues a new edition. Each table shows the data period it covers, and datasets older than roughly two years are flagged as dated rather than quietly refreshed with newer figures from a different source.

Why do two tables for the same channel disagree?

Because they measure different advertisers. An agency's US small-business sample and a DTC ecommerce platform's sample can differ by a factor of four on the same metric. Read the universe line above each table before comparing anything.

Can I use these as targets?

As context, not as targets. A median is where the middle of someone else's sample sits, not where your account should be. The number worth hitting is your own break-even, which depends on your margin and conversion value.

Marketing benchmarks, with their sources attached

Advertising, ecommerce, email and SMS benchmarks by industry — every figure carries its publisher, sample, data period and a link to the report it came from.

Most benchmark posts give you a number with nothing attached to it. A median cost per click means very little until you know whose accounts it came from, in which country, over which months, and whether it is a median or a mean. Those four facts change the number more than the industry label does.

So every table here states them, and no table mixes publishers. Where two reputable reports disagree — Meta CPM at $8.19 in one and $15.06 in another — both are shown with their samples, because the gap is the advertiser mix, not an error to average away.

Pick the channel you are measuring. Each page pairs its tables with the calculator that produces the same metric, so you can put your own number next to the benchmark rather than reading it in isolation.

Benchmarks by channel

Google Ads benchmarksSearch advertising benchmarks by industry: average CTR, cost per click, conversion rate and cost per lead, from named reports with their samples and data periods.
Facebook Ads benchmarksMeta advertising benchmarks by industry and objective: click-through rate, cost per click, CPM, conversion rate, cost per lead and ROAS, each with its source and sample.
TikTok Ads benchmarksTikTok advertising benchmarks for ecommerce brands: CPM, click-through rate, conversion rate, cost per acquisition and ROAS by category, plus social CPM across platforms.
B2B advertising benchmarksB2B paid social benchmarks by platform: cost per lead, cost per click, CPM, click-through and click-to-lead rates from 153 advertisers and $57.6M of spend.
Ecommerce conversion benchmarksEcommerce conversion rate, average order value, add-to-cart and cart abandonment benchmarks, each with the sample, market and period the publisher states.
Email marketing benchmarksEmail benchmarks by industry for campaigns and automated flows: open rate, click rate, click-to-open, unsubscribe rate and placed-order rate, with samples and caveats.
SMS marketing benchmarksEcommerce SMS benchmarks from 321 million messages: click-through and conversion rates for campaigns against automation, plus revenue per message and deliverability.

How to use these numbers

A benchmark is a description of someone else's sample, not a target. The number worth hitting is your own break-even: the cost per acquisition your margin allows, or the return on ad spend that leaves a profit after the product is paid for. Use the calculators to work that out first, then use the benchmark to ask whether the market is even priced for it.

When your number sits far from the table, check the sample before checking your account. A store compared with a lead-generation sample, or a brand-heavy account compared with broad prospecting, will look broken while being perfectly healthy. The universe line above each table is there for exactly that check.

Frequently asked questions

How often are these numbers updated?: When the publisher issues a new edition. Each table shows the data period it covers, and datasets older than roughly two years are flagged as dated rather than quietly refreshed with newer figures from a different source.
Why do two tables for the same channel disagree?: Because they measure different advertisers. An agency's US small-business sample and a DTC ecommerce platform's sample can differ by a factor of four on the same metric. Read the universe line above each table before comparing anything.
Can I use these as targets?: As context, not as targets. A median is where the middle of someone else's sample sits, not where your account should be. The number worth hitting is your own break-even, which depends on your margin and conversion value.

How this data is handled

Every figure is transcribed from a public report and carries its publisher, sample, data period and a link. Nothing is estimated or rounded to “about”.
Publishers are never blended. A click-through rate from one report and a CPM from another never appear in the same row.
Each table states who is in its sample, because advertiser mix — not disagreement — explains most of the spread between reports.
Publisher errors and dated datasets are footnoted rather than quietly corrected or refreshed from another source.

Datasets we left out

These sources rank well for benchmark queries and are not used here. The reason is listed so the omission can be checked rather than taken on trust.

Source	Why it is not used
Store Growers	Its industry tables are an older WordStream dataset, so it is not a second opinion.
Mailchimp	The live page now carries five rows and the data is from December 2023.
Campaign Monitor	2021 data.
Dynamic Yield / Mastercard	The benchmark index redirects to a product page; the figures are only available second-hand.
Contentsquare	The device split is behind a form.
Varos, Emplifi, Social Insider, Hootsuite	Respectively: benchmark content discontinued, paid figures gated, organic-only, and derivative of other reports.
Statista	Paywalled beyond the single global conversion figure.
“LinkedIn Ads Benchmarks 2026” listicles	No named study, no sample, no methodology.