facebook, social media, media, social, internet, network, blog, seo, web, marketing, business, website, design, symbol, icon, online, search, optimization, communication, strategy, computer, service, advertising, information, page, set, sign, document, digital, people, global, group, community, connect, friendship, technology, connection, brown business, brown computer, brown technology, brown laptop, brown marketing, brown facebook, brown online, brown website, brown network, brown community, brown internet, brown digital, brown communication, brown design, brown group, brown information, brown blog, brown web, brown global, brown social, brown media, brown document, brown service, brown friendship, facebook, facebook, facebook, facebook, facebook. Paid social advertising benchmarks with the caveats attached, 2027 edition
Photo by Firmbee on Pixabay

Rules

Part of Paid social advertising, tested against experience

Paid social advertising benchmarks with the caveats attached, 2027 edition

Paid social advertising benchmarks need definitions, counts, cohorts, maturity, uncertainty, net outcomes, full costs, customer effects, and decision thresholds.

What to take away

  • Typical United States planning ranges sit near $8-$20 CPM on Meta feed, $5-$12 on TikTok and $25-$45 on LinkedIn. Link click rates land near 0.8%-1.5%, 0.7%-1.2% and 0.4%-0.8% in the same order.
  • Cost per acquisition commonly falls between $15 and $60 on Meta conversion campaigns and $60 to $200 on LinkedIn lead generation. A 2x return on ad spend is a common planning floor rather than a rule.
  • Treat every figure above as a typical range for orientation. Use them to decide what to investigate, not what to promise.
  • A figure becomes a benchmark only when the definition, period, cohort size and maturity window travel with it.

Read a published number as a starting range. Paid social advertising rewards that habit, because placements, objectives and auctions change what a click or a purchase means.

Typical ranges by platform and objective

These planning ranges cover the United States, with Canada close on most metrics. The video row shows cost per view rather than cost per click.

Platform and objectiveTypical CPMTypical link click rateTypical cost per clickTypical cost per acquisitionTypical return on ad spend
Meta feed, conversions$8-$200.8%-1.5%$0.60-$2.00$15-$601.5x-3.5x
Meta Reels and Stories, conversions$6-$150.6%-1.2%$0.50-$1.80$18-$701.2x-3x
TikTok in-feed, conversions$5-$120.7%-1.2%$0.40-$1.20$20-$701.2x-2.5x
LinkedIn single image, lead generation$25-$450.4%-0.8%$5-$12$60-$2001x-2x
Retargeting, any platform$10-$301%-2.5%$0.50-$2.00$10-$403x-8x
Awareness or video views, Meta and TikTok$3-$100.3%-0.8%$0.01-$0.05 per viewNot applicableNot applicable

Read the table by row, not by column. A LinkedIn lead at $120 can be cheaper than a Meta lead at $25 once qualification and close rates are applied.

Retargeting looks efficient because it captures demand the prospecting campaign paid for. Judge both together or you will shift budget toward the cheaper-looking line.

Cohort rules come before any comparison, and paid media strategy benchmarks explains how to draw them.

Where published benchmarks come from

Your own ad accounts hold the only figures you can audit. Meta Ads Manager, TikTok Ads Manager, LinkedIn Campaign Manager and Google Ads report impression cost, link click rate and cost per result at the ad set level. Export 90 days and 12 months.

Public reports offer orientation. Frequently cited sources include WordStream benchmark reports, Socialinsider studies, Databox surveys, HubSpot marketing data and eMarketer forecasts. Record the publication date, sample size, geography and metric definition.

Funding shapes the sample. A report paid for by a platform or an agency tends to favour that platform or that service. Check the funder before you quote a figure.

Turning a published figure into your own range

  1. Copy the source metric definition word for word.
  2. Rebuild the same metric in your account across a matching window.
  3. Split by platform, objective, format, market, and new versus returning customers.
  4. Report medians with 25th and 75th percentiles instead of one average.
  5. Rewrite the range when late conversions, refunds or cancellations land.

Build the range in five layers: delivery, attention, response, outcome and economics. Each layer needs its own context before anyone acts on it.

[figure 1]

Benchmark Layers and Context

Layer

Delivery
Impressions, reach, frequency
Attention
Viewability, video starts
Response
Clicks, visits, messages
Outcome
Qualified lead, net order
Economics
Contribution, payback, cash

Measures

Delivery
Population, method, format
Attention
Definition, duration, sound
Response
Unique rules, page, quality
Outcome
Source, identity, delay
Economics
Full cost, incrementality

Context required

Delivery
Attention
Response
Outcome
Economics

Cost per click equals CPM divided by 1,000 multiplied by the click rate. Cost per acquisition equals cost per click divided by the landing conversion rate.

A paid social advertising checklist keeps cohort definitions and maturity windows in one reviewable place, which is what step five depends on.

Example: from CPM to a cost per acquisition decision

Assume a $14 CPM, a 1.2% link click rate and a 4% landing page conversion rate. Cost per click works out to about $1.17, and cost per acquisition to about $29. If your target is $25, you need a 1.4% click rate or a 4.7% landing conversion rate at the same CPM.

Google Ads Target CPA bidding applies the same arithmetic in reverse, setting bids to reach a cost per action you name (Target CPA bidding guidance).

Caveats that move the number

  • Attribution window, where 7-day click, 1-day click and view-through produce different totals
  • Reported conversions versus incremental ones
  • Consent and tracking loss, which affects iOS and Canadian accounts more
  • Seasonality and auction density inside the measured period
  • Creative, audience or landing page changes made during the window

Canadian results need their own note. The Office of the Privacy Commissioner's PIPEDA compliance guidance for digital advertising tracking sets consent and transparency expectations for tracking. Quebec adds French-language advertising requirements under the Charter of the French Language.

Separating an observation from a tested effect is a design question. Design choice starts with the objective and the practical constraint (NIST design selection guidance).

Testing one variable at a time turns a range into a decision, and how to improve paid social advertising walks through the isolation step.

Common questions

What is a good click rate for paid social?
On Meta feed, 0.8% to 1.5% is a workable planning band. Below 0.5% usually points to a creative or audience mismatch. Check your own median before you change anything.
Should I quote industry reports to leadership?
Quote them with the date, sample and definition attached, and show your own cohort range beside the external one. The pair is more convincing than either alone.
How often do these ranges need replacing?
Quarterly for cost metrics, and immediately after a platform changes attribution or placements. A stale range is worse than no range.
Do the figures here come from 2027 data?
No. They are planning ranges for 2026 and 2027. The edition label marks the planning period the ranges serve, and each review replaces them with account data.

More in Rules

Latest from Field Desk