← All QuickAds reports
Creative Operations Benchmark · US mid-market · September 2026

Creative demand went up. Your team didn't.

Marketing budgets have been flat at 7.7% of revenue for two years while the number of assets a brand must produce has multiplied. This is what in-house creative teams cost, produce and endure, and where the gap actually opens.

62%
Of in-house creative leaders say they are adequately staffed
47%
Of creatives lose about a day a week to administrative work
7.7%
Marketing budget as a share of revenue, flat two years running
27%
Of CMOs say generative AI gave them more capacity to produce
Benchmarks, US in-house creative operations, 2025–2026
ANA members with an in-house agency82%
Same figure in 200842%
In-house teams of ten people or fewer39%
Teams with a dedicated creative operations role53%
Teams that restructured in the last twelve months74%
Reported an increased workload88%
US graphic designer median salary$61,300
Fully loaded cost over base salary+20–35%
Agency cost per static asset$75–300
Typical team output per month50–100 assets
Projects approved within five review rounds83%
Retail media networks worldwide, each with its own specs277
Section 01

Who this report is for

Written for the person being asked to produce more creative next quarter than they produced this one, with the same team.

Read this if you are

  • A Head of Design or Creative Director being asked to justify headcount, or defend why the queue is long
  • A CMO deciding between another hire, an agency retainer and a subscription
  • A creative operations lead who needs external numbers to benchmark against
  • A founder at $5M–$100M working out when to build a creative team at all

Read the caveats carefully

  • Most creative-ops data comes from companies selling creative-ops software or services. Their incentive is to make the pain look large
  • Every vendor figure here is flagged as such, and the neutral sources are named
  • Almost no published study segments by revenue band, so mid-market-specific benchmarks are thin
  • Where a number does not exist, this report says so rather than inventing one

Section 02

The pain, quantified

Start here, because the rest of the report is an explanation of these numbers.

What in-house creative teams report
Reported an increased workload
88%
Restructured in the last twelve months
74%
Experienced burnout in the past year
70%
Feel adequately staffed
62%
Lose about a day a week to admin
47%
Workload and restructuring from ANA (n=162, 2023) and Lytho. Burnout from the 2024 Mentally Healthy survey of over 2,000 media, marketing and creative professionals, note that sample is Australia and New Zealand weighted, against 53% for workers generally. Staffing from Cella, around 300 US in-house leaders. Admin load from inMotionNow, over 600 respondents. Cella, Lytho and inMotionNow all sell into this market.

The queue is the symptom

Nearly four in ten in-house leaders say they are not adequately staffed, and 88% report a heavier workload than the year before. The backlog is not a planning failure. It is arithmetic.

A day a week, gone

47% of creatives spend roughly one full day each week on administrative duties, and that figure has barely moved in years. On a five-person team that is one full-time equivalent spent on coordination.

Burnout is the tell

70% of media, marketing and creative professionals reported burnout in a twelve-month window, against 53% of workers generally. Attrition then removes the institutional knowledge that made the team fast.

The structural problem

The gap is not a hiring problem. It is an order-of-magnitude problem

A typical in-house team produces 50 to 100 assets a month. The asset requirement for a mid-sized omnichannel brand now runs into the thousands per year. One more designer does not close that.

capacity demand not a gap you hire across

Section 03

The squeeze

Two lines moving in opposite directions. Almost everything in this report follows from the gap between them.

Marketing budget as a share of company revenue
Pre-pandemic
~11%
2023
9.1%
2024
7.7%
2025
7.7%
Gartner CMO Spend Survey, 402 CMOs fielded February to March 2025. Note the respondent base skews to large enterprises (median revenue above $5bn) so treat the level as directional for mid-market and the trend as the finding. In the same survey 39% of CMOs planned agency cuts and 39% planned labour cuts.

In-housing is finished, not starting

82% of ANA members now have an in-house agency, against 78% in 2018, 58% in 2013 and 42% in 2008. ANA expects the ceiling at 85 to 90%. Over the prior three years, 65% had moved established work from an external agency to their in-house team.

This matters because the easy lever has already been pulled. The work came in-house to save money; the volume then kept rising and the team did not.

And the agencies did not fully leave

92% of the same respondents still use external agencies. Media planning and buying is the least in-housed function at 54%, described as too complex to bring fully inside.

The result is a hybrid that most teams never designed on purpose: in-house for volume, agency for peaks and specialisms, and a coordination load nobody costed.

What changed in what the in-house team is measured on

Cost savings as the primary KPI fell from 69% in 2018 to 62% in 2023, while business performance rose from 45% to 59%. In-house creative is increasingly judged on whether the work performs, not on whether it was cheap, which raises the standard without raising the headcount.


Section 04

The volume math

Four forces multiply asset count independently of each other. Together they are why the queue grows no matter how fast the team gets.

DriverWhat it doesThe number
AlgorithmMeta's retrieval system rewards creative diversity and collapses near-duplicates, so volume became a delivery lever rather than a nice-to-have11+ variations beat 2–3 on cost per action; Advantage+ supports up to 150 combinations
Placements and formatsEvery placement has its own aspect ratio, safe zone and duration rule, one concept becomes many buildsOne concept × 25 sizes × 7 languages × 25 variations = 4,375 assets
Retail mediaEach network publishes its own spec rulebook. Assets rebuild rather than reuse277 networks worldwide; US retail media $53.7bn in 2024, up 23%
Refresh cadenceCreative fatigue forces replacement on a fixed clock regardless of how good the asset wasTwo to four refreshes a month is standard guidance

The worked example

A ten-SKU brand selling on Amazon, Walmart and Target plus Meta, TikTok and Google is estimated to need somewhere between 2,500 and 30,000 assets a year depending on refresh discipline. The typical in-house team produces 50 to 100 a month.

Where the variants actually land

Account-level data across 1,200+ e-commerce accounts found moving from one or two creatives per ad set to four to six produced a 22% lift in click-through rate. That is a production requirement disguised as a media insight.

Read the source bias

The asset-count estimates come from creative automation vendors, whose interest is in the number being large. The direction is corroborated across sources; the magnitude is not audited. Model your own SKU count against your own placement list.


Section 05

What a team costs, and what it looks like

The numbers to benchmark against before you argue for headcount.

In-house team sizeShare
Ten people or fewer39%
Eleven to thirty34%
Thirty-one to fifty11%
More than fifty16%
US roleRange
Graphic designer$52,000–79,500
Art director$83,250–122,500
Creative director$102,000–162,500
Creative ops manager~$123,000 average
RouteIndicative costSpeedBest for
In-house designerBase salary +20–35% fully loadedFastest once briefedSteady, brand-critical volume
Freelance$25–200 per hourVariableUnpredictable or low volume
Agency retainer$3,000–8,000/month for 20–40 assets5–10 business days per batchPeaks and specialist craft
Agency hourly$150–250 mid-tier, $250–450 seniorProject-dependentConcepting, not versioning
Per static resize$75–1505–10 business daysNothing, if you can avoid it

Motion is the hiring bottleneck

49% of in-house leaders describe motion designers as limited in availability and 22% report real trouble finding them. Graphic designers are the easiest role to fill, which is the opposite of where demand is growing.

Only half have an ops function

53% of in-house teams have a dedicated creative operations role, 87% use a project management tool and 68% use a DAM. Among mature in-house agencies, 95% have a dedicated DAM manager.

The benchmark that does not exist

No published study segments creative team size by revenue band, and no neutral source publishes assets per designer per month. Anyone quoting you a precise ratio is quoting a vendor estimate.


Section 06

The review tax

The most-cited bottleneck in every survey of in-house teams, and the one with the weakest independent evidence behind it.

What is measured

The approval numbers

Review rounds
83% approved in five or fewer
Up from 77% the prior year, improving, but five rounds is still five
Turnaround
78% approved within a week
Up from 65%, on a sample of over 600 creative and marketing professionals
Admin load
47% lose a day a week
Effectively flat year on year, which suggests tooling has not touched it
Where it degrades
Late stakeholders
Feedback arriving after round one is repeatedly cited as the multiplier
How to read it

Three cautions

01
The cleanest figures are vendor telemetry
The best revision data measures one vendor's own customers against a comparison panel, structurally flattering
02
Two widely-quoted stats have no traceable study
The "67% of revision rounds caused by vague feedback" figure could not be traced to a primary source
03
Self-report inflates improvement
Teams surveyed about their own process report faster approvals than their own timestamps usually show

The number worth measuring yourself

First-time approval rate is the single most useful internal metric in this report, and almost nobody tracks it. If fewer than six in ten deliverables clear on the first pass, the constraint is the brief, not the designer, and hiring another designer will produce more rejected work faster.


Section 07

What AI actually changed

Adoption has been fast and the savings are real. They have not turned into more output, because demand absorbed them.

Where CMOs say generative AI returned value
Time efficiency
49%
Cost efficiency
40%
Increased capacity to produce more content
27%
Gartner CMO Spend Survey, 402 CMOs, 2025. The gap between the first two bars and the third is the finding: the saved hours were reinvested into existing work rather than banked as additional output.

Adoption is not the question any more

65% of organisations reported regular generative AI use in a McKinsey survey of 1,363 respondents, roughly double the figure ten months earlier, with marketing and sales the fastest-adopting function.

The savings are plausible

Adobe reports 85% of marketers and creatives saving around four hours a week. Adobe sells the tools, so treat the figure as directional, but four hours against a day a week of admin is the right order of magnitude.

Maturity is rare

Only about 1% of leaders describe their generative AI rollouts as mature. Most teams are using it inside an unchanged process, which is why the time comes back as relief rather than as capacity.

Why the savings disappear

If a team was already over capacity, faster production does not create slack, it clears backlog, and the backlog refills from the volume drivers in Section 04. Capacity only appears when the saved time is deliberately ring-fenced, usually by moving a whole category of work out of the queue rather than speeding each item up.


Section 08

Where the load is heaviest

Two different pressures: SKU count, which multiplies assets, and compliance, which multiplies rounds. The worst position is both.

SectorAsset multiplicationReview burdenThe dominant constraint
Beauty & cosmeticsHighestModerate50–150 new SKUs a year; one foundation line alone can be 40–50 shades
Fashion & apparelVery highLowSize × colour × season multiplies every concept
Consumer electronicsModerateModerateSpec accuracy and compatibility claims slow sign-off
JewelryModerateLowMetal and stone variants, plus high-ticket trust assets
Financial servicesLowHighestDisclosure review; one vendor counted 30,660 compliance comments in a year across its clients
Health & wellnessModerateHighestMandatory medical, legal and regulatory review before anything ships
Food & beverageLowestModerate5–15 new SKUs a year, the least asset multiplication of any consumer category

No published cross-industry benchmark exists

This table is assembled from SKU-launch data and compliance reporting rather than from a single study, because no published source benchmarks creative team size or asset volume by vertical. Use it to locate yourself, not to set a target.


Section 09

What to do about it

Four moves, in the order that actually resolves the constraint.

First

Measure before you hire

Track four numbers for one month: cost per asset, brief-to-live time, first-time approval rate, and the share of designer hours spent on versioning rather than concepting.
If first-time approval is below 60%, the brief is the constraint. Another designer will produce rejected work faster.
Second

Hire operations before you hire craft

Only 53% of teams have a dedicated creative ops role. At roughly $123,000 it typically returns more throughput than an equivalently-priced production designer, because it removes the day a week going to admin.
Fix intake first. Unclear briefs are the most-cited cause of extra rounds in every survey in this report.
Third

Match the route to the rhythm

Steady brand-critical volume belongs in-house. Specialist peaks, motion especially, where half of leaders report limited availability, belong outside it.
Treat retail media as its own pipeline once it passes 10–15% of spend. Each network's specs require rebuilds, not reuse.
Fourth

Move whole categories of work out of the queue

Speeding each task up gets absorbed. Removing a category (resizing, versioning, creator sourcing and chasing) is what actually creates capacity.
Ring-fence the time you recover. If it goes back into the queue by default, you have bought relief rather than capacity.

Where this is heading

The work leaving creative teams first is the coordination layer rather than the craft: resizing and versioning, then sourcing and briefing creators and chasing the output back. That last piece is what we are building Remy for. The pattern worth noticing is not which tool does it. It is that the categories being removed are the ones that never needed a designer, and they are the ones consuming a day a week.

About QuickAds

QuickAds.ai, AI Creative Engine

Research winning ads. Generate high-converting creatives. Ship them at volume. All under one roof. The data in this report was built using the QuickAds ad intelligence library.

32M+

Ads in Library

35+

Languages

23K+

Brands Trust QuickAds

Ad Intelligence

Search and analyse competitor ads across Meta, Google and TikTok. Filter by industry, format, hook type, engagement, brand and market. The 32M+ library is updated continuously.

Creative Production

Generate high-performing ad creatives using AI: static, video, carousel. 1200+ creatives a month per brand, five to seven days from brief to first full batch. The production bottleneck is solved; now strategy is the constraint.

Want to scale creative production without scaling the team?

Visit quickads.ai →