Rules
Part of How to assess digital marketing trend claims before you plan
Two digital marketing trend figures read closely: what they show and what they hide
Two IAB/PwC figures, US internet ad revenue and retail media network revenue, are read as reference points with unit, population, date, geography and method attached. The article separates observed revenue from survey expectations and tests trend figures against matched company evidence.
What to take away
- Every headline number needs its unit, base, date, geography, population, method, sponsor, and limitations.
- Observed industry revenue and survey-reported expectations answer different questions and should not be blended.
- Treat trend figures as planning reference points, then refresh them and compare them with matched company evidence.
Digital marketing trend data is most useful when each number keeps its population, date, geography and method. This article reads two figures from one report family: the IAB Internet Advertising Revenue Report's full-year United States total and the retail media network line inside it. Both are reference points, not benchmarks.
Figure one: United States internet advertising revenue
The IAB Internet Advertising Revenue Report: Full Year 2024 reports $258.6 billion of United States internet advertising revenue for 2024, up 14.9 percent year over year and the highest level recorded. The IAB published it in April 2025, and PwC Advisory Services LLC conducted the work.
Method carries as much weight as the total. According to the report's methodology note, PwC compiles a database of companies selling internet and mobile advertising, surveys leading industry players, adds publicly disclosed corporate data, and applies a conservative revenue estimate to non-participating companies.
Scope is wide: audio, video, CTV and OTT, gaming, search, social, commerce media and programmatic, sold on desktop and mobile. A total that broad cannot size one channel or one advertiser's opportunity, and it says nothing about whether any of that spending worked.
Two quirks hide inside the headline. Programmatic revenue, $134.8 billion in 2024, is a way of selling rather than a separate format. Check the current IAB methodology note before comparing it with format lines, because definitions can overlap. Treat any forward-looking commentary as separate from billed revenue, and check the methodology note before using a projection.
A full-year 2025 edition dated April 2026 is now listed by the IAB. Check current totals and definitions there before quoting a number in a budget paper, because publishers revise series and estimates between editions, and an old figure on a slide still looks current.
US internet ad revenue 2025
- $294.6 billionUS internet advertising revenue
- 13.9%year-over-year growth
- 2025full-year period
Direct link or citation to the IAB/PwC report itself
The Full Year 2024 report landing page is on iab.com at /research/iab-pwc-internet-advertising-revenue-report-full-year-2024/. The full report PDF is at iab.com/wp-content/uploads/2025/04/IAB_PwC-Internet-Ad-Revenue-Report-Full-Year-2024.pdf. IAB's press release announcing the report, which was conducted by PwC, is dated April 17, 2025, according to IAB. Both are hosted on iab.com.
The Full Year 2024 edition is the 29th year of the report, according to IAB. The IAB Internet Advertising Revenue Report was initiated by IAB in 1996 and is conducted by PwC on an ongoing basis, with results released quarterly. The Full Year 2025 edition, the 30th year, is dated April 16, 2026, according to IAB.
The Full Year 2025 PDF is published at iab.com/wp-content/uploads/2026/04/IAB_PwC_Internet_Ad_Revenue_Report_Full_Year_2025_April_2026.pdf, according to IAB. A companion webinar on that report was held April 21, 2026 at 1 pm ET with experts from IAB, PwC and Madison & Wall, according to IAB. Check the IAB site for the session materials.
Figure two: retail media network advertising revenue
The same 2024 report puts retail media network advertising revenue at $53.7 billion, an increase of $10.1 billion on 2023 and 23.0 percent growth. The line sits inside commerce media, which the report says was estimated at $1 billion in 2016. It is the closest thing there to a channel-level trend.
What it hides starts with the payer. The figure is revenue reported by retailers and their networks, not brand profit. Check the current report's definition of a retail media network on the IAB site before comparing it with another network's scope or inventory.
Network definitions, inventory and measurement differ, and the aggregate gives no split by retailer, category or placement. Nor can it tell a brand whether extra spend adds sales or moves demand it would have captured anyway. That needs a matched test.
Definition of commerce media versus retail media
IAB describes commerce media as including retail media networks. IAB reports that commerce media revenues rose 23% to $53.7 billion in 2024. Commerce media advertising revenues increased by $9.7 billion since 2024, reaching $63.4 billion in 2025, an 18.0% year-over-year increase, according to the IAB/PwC Full Year 2025 report.
The IAB/PwC report says commerce is expanding beyond traditional retail media networks into a broader data-driven ecosystem across off-site, in-store and cross-partner environments. Affiverse reports that retail media networks are valued for privacy-compliant targeting based on actual purchasing data, as signal loss affects user-level identifiers.
Retail media went from an estimated $1 billion in 2016 to $53.7 billion in 2024, per the IAB/PwC report. That history helps separate a broad commerce media category from the retail media network line inside it, but the figures still describe seller revenue rather than brand profit.
Keep survey populations separate
A revenue census, a respondent survey, an account log and an expectation question observe different populations and units. Preserve who answered, how respondents were recruited, the countries and roles covered, the exact question, field dates, weighting, missing responses and sponsor before setting two figures side by side.
Bases also change within one report, because not every respondent is eligible for every item, and a subgroup figure can look like the whole sample. Keep the base, the response option and whether the measure is observed, self-reported, modeled, intended or forecast. Never copy a percentage without that context; the common digital marketing campaigns questions page covers the ones readers ask most.
Cross-industry technology surveys carry the same risk. Check the current Stanford AI Index appendix for its adoption question, sample and definition before quoting an organizational use figure as marketing evidence. Those figures describe organizational behavior, not campaign performance, and their samples and wording rarely match a marketing plan.
Build a decision-ready data card
- Exact claim and unitobserved revenue, intention, forecast, or modeled estimate.
- Publisher, sponsor, reporting period or field dates, geography, population, sample, weighting, and question or definition wording.
- Relevant segment, and whether your own customer evidence agrees.
- Decision, threshold, uncertainty, and the date the figure will be refreshed.
A large percentage can still be irrelevant to your market. A small, well-matched operating data set, such as your own order export for one region, can beat a national headline. Keep the method attached whenever a statistic reaches a presentation, a forecast, an article or a budget model.
Decision table
| Data card field | Record | Failure prevented |
|---|---|---|
| Unit | Revenue, respondent, user, action, or forecast | Unlike measures are blended |
| Population | Who or what could be counted | A sample becomes the market |
| Time | Field and measurement dates | Old evidence looks current |
| Method | Source, question, model, and weighting | A result loses its limits |
| Decision | Threshold and refresh date | A statistic becomes decoration |
A worked example: the growth rate losing its meaning
An illustrative direct-to-consumer brand reads the 14.9 percent growth figure and plans a 14.9 percent increase in its own paid spend. The number measures the whole United States market across search, social, video, retail media and every other format. It says nothing about the brand's cost per order or marginal return.
The better test is the brand's own data. Compare the last twelve months of spend and orders by channel, then check whether the planned increase still returns the same contribution at the margin. Only that result belongs in the budget model, with the market figure kept as context.
A worked example: the retail media figure losing its meaning
Another brand sees 23.0 percent growth in retail media network revenue and moves a fifth of its budget onto one network. The figure describes what retailers collectively billed, not what any advertiser gained. A network can grow because more brands joined, because rates rose, or because measurement changed.
A matched test answers the real question. Hold one category budget steady, split spend between the network and the existing mix over the same period, and compare incremental orders or contribution. For a starting point on rates, these retail media network pricing benchmarks show what to ask for.
Check the figures against the page
Before a figure reaches a presentation, check that the table, caption and text state the same unit, date and population. IAB format totals must not be summed with the programmatic line, and the retail media figure must not be read as a brand return.
Re-read the source once a year, or sooner if the publisher changes its method or issues a new edition. When you are comparing survey benchmarks rather than revenue, how to compare digital marketing campaigns benchmarks sets out which bases have to match.
Common questions
What makes digital marketing trend data trustworthy?
Trust rises when the source discloses the unit, population, sample or census coverage, timing, definition, sponsor, uncertainty, limitations and correction process. Revenue reported by sellers of advertising and answers from a recruited panel meet that bar differently, so record which one you are quoting.
Can an industry growth figure set a company budget?
No. It can inform a scenario, but allocation still depends on customers, unit economics, channel role, capacity, measurement, risk and tested contribution.
How often should a trend statistic be refreshed?
Set the refresh date from decision risk, source cadence, market volatility and material platform or method changes, not from a fixed editorial habit. The wider digital marketing trends framework applies the same rule.
