Paid Ads & Creative
7 Checks Before You Hire an Ad Creative Agency

Joshua Jackai
19 min

Published 18 November 2026 ยท Last updated and fact-checked November 2026
Short answer: An ad creative agency makes the images, videos and copy for paid ads. Hire one only after you can name your bottleneck: volume, craft or testing. Then check the pricing model, brief, judging and exit terms. Published plans run from $99 a creator video to a $15,000 monthly minimum.
Picture a founder of a skincare brand with three proposals open (a composite, not a client). One is a $15,000 a month retainer from a big name. One is a design subscription. One is a creator platform that charges per video. All three promise "scroll-stopping creative". None of them says what happens when the first ten ads lose, who owns the files, or how a winner will be called.
I'm Joshua Jackai, and Demand Designer is my studio. I make Meta and TikTok ad creative for direct-to-consumer (DTC) brands and B2B teams, so I sit on the other side of this decision every month. These are the seven checks I would run if I were the buyer. Each one rests on published research about how agencies and clients work together, and each says where that research came from: mostly agency surveys, experiments and models, not Meta ads by name.
Key takeaways
Rosengren, Eisend, Koslow and Dahlen (2020) pooled 878 effect sizes from 67 papers in the Journal of Marketing: creative ads show positive effects, and the authors stress weighing both originality and appropriateness.
Koslow, Sasser and Riordan (2006) surveyed agency staff on 1,011 campaigns: a client willing to explore new strategy got more creative work than one with a strategy line in the brief.
Horsky (2006) found in Marketing Science that firms use internal agencies when they have internal creative abilities, or low-level needs in both creative and media.
Lewis and Rao (2015) ran 25 large ad field experiments: the median confidence interval on return on investment was over 100 percentage points wide.
Superside's pricing page (vendor, read 18 November 2026) starts subscriptions at a $15,000 monthly minimum on an annual term; Billo's creator videos start at $99.
Check | What it protects you from | Key number | Source type |
|---|---|---|---|
1. Name the bottleneck | Paying for the wrong kind of help | In-house fits firms with creative ability or basic creative and media needs | Peer reviewed |
2. Price the model, not the ad | Retainers that buy hours, not ideas | $99 per video to $15,000 a month | Vendor pages |
3. Ask for original and on-brief | Clever ads that sell nothing | 878 effect sizes, 67 papers | Peer reviewed |
4. Write a brief that leaves room | Agencies that play it safe | 1,011 campaigns | Peer reviewed |
5. Plan how you will judge risky work | Killing the idea that would have won | 1,125 campaigns, 408 agency staff | Peer reviewed |
6. Agree how a winner is called | Arguing about one ad's ROAS | ROI intervals over 100 points wide | Peer reviewed |
7. Pay for a trial, not a free pitch | Spec work and hard exits | Full pitch refunds never optimal in the model | Peer reviewed |
1. How to choose an ad creative agency: name your bottleneck first
The first check is naming what is actually short: new ideas, finished assets, or a way to learn which ads work. An ad creative agency fixes one or two of those, rarely all three. Horsky (2006) found firms keep creative work inside when they have the ability or need little.
Volume problem: you have winning angles but not enough versions; a creator platform or design subscription fits
Craft problem: your ads look cheap next to competitors; a studio with strong art direction fits
Testing problem: you cannot tell what is working; an agency that writes hypotheses and reads the account fits
In-house signal: you already have a designer who can ship weekly
Skip the agency if: one editor or designer already fixes your named bottleneck
Most founders I speak to describe all three problems at once, because a weak account feels weak everywhere. Pick the one that hurts most this quarter. If the ads you have are good but the account is starved of new versions, you do not need strategy decks. You need more frames, faster. If your ads are plentiful but look like everyone else's, volume makes it worse.
Horsky's study looked at how US firms split creative and media work between in-house teams, full-service agencies and specialist shops. It was written about the big agency world, not DTC brands on Meta, so I use it for the logic, not the market share. The logic holds: buying outside help only pays when it brings a skill you lack. A founder who shoots good product video on a phone may only need an editor, not an agency. A brand with a designer who already ships weekly may only need fresh concepts once a quarter.
Write the bottleneck at the top of your brief. Then judge every proposal by whether it fixes that one thing.
"Internal agencies are used when firms have internal creative abilities, or have low-level requirements in both creative and media," Horsky reported in Marketing Science.

2. Ad creative agency cost: price the model, not the ad
Ad creative is sold in three models: per asset, by subscription and by retainer. Published prices I checked on 18 November 2026 run from $99 for a 15-second creator video at Billo to a $15,000 monthly minimum at Superside. Compare what each model buys per tested idea, not per file.
Per asset: Billo videos start at $99; its $1,000 pack covers up to 10 videos (vendor, read 18 November 2026)
Subscription: Design Pickle's Always-on team is from $2,200 a month, priced by design time a day (vendor, read 18 November 2026)
Retainer or flex budget: Superside starts at a $15,000 monthly minimum on an annual term, plus a $1,000 monthly software fee (vendor, read 18 November 2026)
What to ask: how many new concepts, not files, ship each month
Hidden cost: your own time writing briefs and giving feedback
A file is not an idea. Ten resizes of one ad count as ten assets on some invoices and one concept in your account. Ask every agency the same question: how many genuinely different concepts will I see each month, and how many versions of each? Then divide the monthly fee by the concepts. That number is the one to compare.
Spake, D'Souza, Crutchfield and Morgan (1999) used agency theory to study how US advertisers pay their agencies, including fees tied to campaign results. My own reading, not theirs: the way you pay shapes what you get. An hourly retainer rewards hours. A per-video price rewards output. Neither rewards learning unless you write it into the scope.
Most agencies on comparison lists publish no rate card at all. In Webtonic's 2026 roundup, none of the 21 picks shows a public rate card. Treat a missing price as a reason to ask for a written range before the first call.
"Advertising agency compensation may change dramatically in coming years as advertisers put pressure on agencies to reduce commissions and tie compensation to performance," Spake and colleagues wrote in the Journal of Advertising.

3. Ask for work that is original and on-brief: best check for portfolio reviews
When you review a portfolio, score every ad on two things: is it unlike what the category runs, and does it say something true about the product? Rosengren and colleagues (2020) pooled 878 effect sizes: creative ads had positive effects, and both parts matter. Award-winning but irrelevant work is a warning sign.
Originality: would you remember this ad tomorrow without the logo?
Appropriateness: does it show the product, the problem or the proof?
Stronger for: high-involvement products, the ones buyers think about
Red flag: a reel full of awards with no product in frame
Funnel stage: cold prospecting most of all
Rosengren, Eisend, Koslow and Dahlen (2020) combined 93 data sets from 67 papers. Creativity had positive effects across them, and the effects were stronger for high-involvement products than low-involvement ones. Their deeper finding is about the two parts of creativity: originality mostly made people like the ad, while appropriateness made the brand look like it had something worth saying.
Smith, Chen and Yang (2008) tested the same two parts, which they called divergence and relevance, in a controlled experiment. They found the expected divergence-by-relevance interaction on 12 of the 13 variables they measured, from brand awareness to brand intentions.
Neither study looked at Meta ads, and the meta-analysis authors call for more research in digital contexts. I apply them this way: in a portfolio review, cover the logo and ask what is being sold. If you cannot tell in a few seconds, the agency is selling you the first half of creativity only.
"Moderation analyses show that the effects of advertising creativity are stronger for high- (vs. low-) involvement products," Rosengren, Eisend, Koslow and Dahlen reported in the Journal of Marketing.

4. Write a brief that leaves room: best check for your first month
The brief you hand over decides much of what comes back. In Koslow, Sasser and Riordan's survey of agency staff on 1,011 campaigns, a strategy line in the brief did little for creativity; a client willing to explore new strategic ideas did a lot. Write the facts tightly and leave the angle open.
Fix in the brief: product facts, price, offer, claims you can prove, banned claims
Leave open: the angle, the hook, the format
Include: your three best and three worst ads with their numbers
Give access: read-only access to the ad account, so the agency sees what you see
Time to allow: one round of concepts before production
Koslow, Sasser and Riordan (2006) asked agency people how their clients shaped the work on 1,011 campaigns. Three things mattered: setting direction, giving resources and judging the work. Better-resourced agencies did more creative work, and one key resource was access to the client's top managers.
The same study carries a warning for founders. Senior, sophisticated clients sometimes chilled the agency, because staff were afraid of them. In a DTC brand, the founder is usually both the top manager and the final judge. If every concept has to survive your personal taste, the agency learns to send you safe work.
Note the limit: this study asked agency staff, not clients, and it is not about Meta ads. I use it as a guide to briefing, not as proof about Meta results. In practice that means two things. Give the agency the numbers you already have, so it does not repeat a loser. And let the people making the ads hear from you directly, not through three layers of approval.
"Merely having strategy in client briefs has little effect on creativity," Koslow, Sasser and Riordan reported in the Journal of Advertising.
"Sophisticated, high-ranking clients appear to send a "chilling" effect throughout the agency, thus dampening creative potential," Koslow, Sasser and Riordan wrote in the Journal of Advertising.

5. Plan how you will judge risky work: best check for founders who approve everything
Decide before the first round how you will treat an idea that makes you nervous. In Sasser, Koslow and Kilgour's study of 1,125 campaigns reported by 408 agency staff, agencies felt clients asked for creativity yet rejected cutting-edge work. A simple rule, test it small before you veto it, keeps the best ideas alive.
Rule to agree: any concept the agency rates as its strongest gets a small paid test
Budget for it: a fixed share of spend kept for new, riskier angles
Who decides: the numbers after the test, not the meeting before it
What to veto outright: only claims you cannot prove and anything off-brand by your written rules
Funnel stage: cold prospecting, where new angles matter most
Sasser, Koslow and Kilgour (2013) examined how a client's openness to new ideas shaped campaigns. Their sample is agency staff reporting on campaigns, so it describes how agencies see clients, not a Meta benchmark. The pattern is familiar from my own client work: a founder approves the ad that looks like last month's winner and parks the strange one.
Kohavi, Longbotham, Sommerfield and Henne (2009) made a related point about website tests. The authors report from their experience that teams learn most when they listen to customers, not the highest paid person's opinion. It is a practitioner guide, not an ad study. In a small brand, that person is you.
The fix is cheap. Agree in writing that the agency's strongest concept each month gets a test at a fixed budget, whatever you think of it. You still veto anything untrue. You stop vetoing on taste. Keep the test budget small enough that a loss does not hurt, and large enough to reach the minimum spend you agreed in check 6. Then write down what each test taught you, win or lose, so the next brief starts smarter.
"Marketers say they want greater creativity, yet their agencies feel that these clients reject cutting-edge work and fail to adopt risky campaigns," Sasser, Koslow and Kilgour wrote in the Journal of Advertising Research.
"Significant learning and return-on-investment (ROI) are seen when development teams listen to their customers, not to the Highest Paid Person's Opinion (HiPPO)," Kohavi and colleagues reported in Data Mining and Knowledge Discovery.

6. Agree how a winner is called: best check before you sign
Before signing, agree in writing what counts as a winning ad, over how much spend, and who reads the result. Ad results are noisy: Lewis and Rao found, across 25 large field experiments, a median return-on-investment confidence interval over 100 percentage points wide. Judge an agency on a quarter of work, not one ad.
Metric: cost per purchase or cost per qualified lead, not clicks alone
Minimum spend: set it before launch so nobody calls a winner early
Review cadence: weekly reads, monthly decisions
Refresh rhythm: plan new concepts before current ones wear out
Agency scorecard: concepts shipped, share tested, share that beat control over the quarter
Lewis and Rao (2015) studied 25 field experiments with major US retailers and brokerages, most reaching millions of customers. Even at that scale, the return on an ad was hard to pin down. A small DTC account has far fewer buyers per ad, so I expect one ad's numbers to be noisier still; that is my inference, not their finding.
That changes what you should hold an agency to. Not "make me a 4x ad" but "ship this many real concepts, test them the agreed way, and show me what we learned". An agency that promises a set ROAS before seeing your account is promising something nobody can measure that precisely.
Wear-out matters too. Schmidt and Eisend (2015) pooled ad repetition studies and found that, in experiments, attitude reached its maximum at about ten exposures and repetition effects decayed over time. They did not test Meta ads, so my refresh rule is my own: plan the next batch before the current one tires, as my notes on creative fatigue on Meta ads explain. Write the refresh dates into the scope, so new concepts arrive before results drop.
"The median confidence interval on return on investment is over 100 percentage points wide," Lewis and Rao reported in the Quarterly Journal of Economics.

7. Pay for a trial, not a free pitch: best check for the contract
Skip the free spec pitch. Offer a short paid trial with a fixed scope, file ownership and a clean exit. Game theory on agency selection contests shows that refunding pitch costs in full is never the advertiser's best move, but a modest stipend to a new agency can be. A paid trial applies that logic.
Trial length: one month or one batch of concepts
Scope: a set number of concepts, versions and sizes, written down
Ownership: you own final files and raw project files once paid
Exit: notice period of 30 days or less after the trial
Account access: the ad account stays in your business manager, never theirs
Horsky, Horsky and Zeithammer (2016) modelled the modern agency pitch, where each agency spends money up front on a custom campaign to win the account. They found a stipend below a new agency's pitch cost can benefit the advertiser under certain conditions. A small paid trial is my practical version of that idea; the study modelled pitches, not trials. You see real work on your real account, and the agency is paid for it.
Exit terms matter because switching is costly. Davies and Prince (2010) surveyed North American clients and found high switching costs went with outcome-based pay and formal evaluation. Keep your switching cost low from day one: your account, your files, your pixel, your brief.
Ask one more question before you sign: what happens to work in progress if either side ends the deal? A good answer names who keeps half-finished concepts and raw footage, and what you pay for them. A vague answer is the start of a dispute. Put the answer in the contract, next to the notice period, and read it again before the trial ends.
"The authors show that reimbursing bid-preparation costs in full is never optimal, nor is reimbursing any part of the incumbent's bid-preparation cost," Horsky, Horsky and Zeithammer wrote in the Journal of Marketing Research.
"With high switching costs, outcome-based compensation and formal evaluation procedures are likely," Davies and Prince reported in the Journal of Current Issues and Research in Advertising.

Ad creative agency vs in-house vs creator platform: matrix by need
Use this matrix to match your bottleneck to a model before you take any sales calls. Each row shows what the model is good at, what it costs on published pricing, and where it breaks. The live page lets you filter the rows by bottleneck and by budget.
Model | Best for | Published starting price | Where it breaks | Who owns strategy |
|---|---|---|---|---|
Creator platform (Billo) | Volume of creator videos | $99 per video | No strategy or testing plan included | You |
Design subscription (Design Pickle) | Steady static and motion output | From $2,200 a month | Briefs must be precise | You |
Performance creative agency | Testing problem, new angles | Usually quoted per scope | Cost per concept can be high | Shared |
Large creative platform (Superside) | High volume across many brands or markets | $15,000 a month minimum, annual | Too big for most early DTC budgets | Shared |
In-house designer | Daily output on a known style | A salary | Ideas run dry without outside input | You |
Freelancer | One clear skill gap | Quoted per job | Single point of failure | You |
Calculator: cost per tested concept from any proposal
The calculator in the live version of this page takes an agency's monthly fee, the number of new concepts it ships and your monthly ad spend. It returns cost per concept and creative fee as a share of spend. Defaults are placeholders, not benchmarks, so type in the numbers from each proposal you hold.
With example inputs: a $4,000 monthly fee for 8 new concepts is $500 a concept. Against $20,000 of monthly ad spend, the fee is 20% of spend. A $2,200 subscription that produces 4 real concepts is $550 a concept, more than the agency in that example, even though its invoice is smaller. This rule is mine, not a study: I want cost per concept to fall, or win rate to rise, by the end of the first quarter.
What does an ad creative agency do?
An ad creative agency makes the images, videos and copy that run in paid ads, and the better ones also decide what to test. It does not buy media unless that is in scope. For DTC brands, the work is mostly Meta and TikTok assets in many sizes, refreshed often.
The scope varies more than the name suggests. Some agencies write hooks and scripts, cast creators, edit and resize. Some only design statics. Some, often called performance creative agencies, also read your ad account each week and plan the next tests. Horsky's 2006 study treats creative and media as separate components a firm can buy from different agents. Ask every agency to list what is in and out of scope, in writing, before you compare prices.
Is an ad design agency the same as an ad creative agency?
Mostly yes, with one difference in emphasis. An ad design agency usually means static and motion design to a brief. An ad creative agency usually also covers concepts, scripts and video. The labels are loose, so judge by the scope list, not the name.
If you need finished statics in many sizes from ideas you already have, a design team or subscription is the cheaper fit. If you need new angles, scripts and creator video, look for the wider scope. Either way, use the cost-per-concept sum from the calculator, and check whether the team will see your account data. Design without data tends to repeat what looked good last time.
Frequently asked questions
How much does an ad creative agency cost?
Published prices I checked on 18 November 2026 range from $99 for a 15-second creator video at Billo to a $15,000 monthly minimum on an annual term at Superside. Design Pickle starts at $2,200 a month. Most agencies quote per scope, so ask for cost per concept.
Is it better to hire an ad creative agency or build in-house?
Build in-house when you already have creative ability or only basic needs; Horsky (2006, Marketing Science) found firms use internal agencies in those cases. Hire an agency when you lack a skill, need new angles, or cannot yet justify a full salary.
What should I look for in an ad creative agency portfolio?
Look for ads that are both original and clearly about the product. A meta-analysis of 878 effect sizes found creativity works when originality and appropriateness come together. Cover the logo on each ad: if you cannot tell what is sold within seconds, keep looking.
How many ad creatives should an agency deliver each month?
There is no published standard; ask how many new concepts, not files, ship monthly. Ten sizes of one idea is one concept. Pick a number your budget can test fairly, plan refreshes before ads wear out, and judge the agency over a quarter.
Should an ad creative agency pitch for free?
I would not ask for one. A game-theory study found refunding pitch costs in full is never optimal for the advertiser, but a modest stipend to a new agency can help. A short paid trial is my recommendation, not a finding.
How I make ad creative at Demand Designer
Demand Designer is my studio, not an agency with account managers. I make Meta and TikTok ad creative for DTC brands and B2B teams on a fixed monthly price, and I run the same seven checks on myself: one bottleneck, cost per concept, a tight brief and an agreed way to call winners.
The concepts I make follow the research in my Meta ads creative strategy rules and get tested with the method in ad creative testing. If you are deciding between still images and video first, start with static vs video ads on Meta.
My pricing is published. The Conversion Creatives plan, my monthly ad creative subscription, is $999 or $2,449 a month (USD), with briefs, scripts and edits included. Whoever makes your ads, you need a weekly way to decide which to keep, rest or kill: my free Kill Sheet gives you five signals, three verdicts and ten minutes every Monday.
Sources
Peer-reviewed research
Davies, M., and Prince, M. (2010). Advertising agency compensation, client evaluation and switching costs: An extension of agency theory. Journal of Current Issues and Research in Advertising, 32(1), 13 to 31.
Horsky, S. (2006). The changing architecture of advertising agencies. Marketing Science, 25(4), 367 to 383.
Horsky, D., Horsky, S., and Zeithammer, R. (2016). The modern advertising agency selection contest: A case for stipends to new participants. Journal of Marketing Research, 53(5), 773 to 789.
Kohavi, R., Longbotham, R., Sommerfield, D., and Henne, R. M. (2009). Controlled experiments on the web: Survey and practical guide. Data Mining and Knowledge Discovery, 18(1), 140 to 181.
Koslow, S., Sasser, S. L., and Riordan, E. A. (2006). Do marketers get the advertising they need or the advertising they deserve? Agency views of how clients influence creativity. Journal of Advertising, 35(3), 81 to 101.
Lewis, R. A., and Rao, J. M. (2015). The unfavorable economics of measuring the returns to advertising. The Quarterly Journal of Economics, 130(4), 1941 to 1973.
Rosengren, S., Eisend, M., Koslow, S., and Dahlen, M. (2020). A meta-analysis of when and how advertising creativity works. Journal of Marketing, 84(6), 39 to 56.
Sasser, S. L., Koslow, S., and Kilgour, M. (2013). Matching creative agencies with results-driven marketers. Journal of Advertising Research, 53(3), 297 to 312.
Schmidt, S., and Eisend, M. (2015). Advertising repetition: A meta-analysis on effective frequency in advertising. Journal of Advertising, 44(4), 415 to 428.
Smith, R. E., Chen, J., and Yang, X. (2008). The impact of advertising creativity on the hierarchy of effects. Journal of Advertising, 37(4), 47 to 62.
Spake, D. F., D'Souza, G., Crutchfield, T. N., and Morgan, R. M. (1999). Advertising agency compensation: An agency theory explanation. Journal of Advertising, 28(3), 53 to 72.
Industry and vendor sources (not peer reviewed)
Superside. Pricing. Read 18 November 2026. Vendor source.
Design Pickle. Pricing. Read 18 November 2026. Vendor source.
Billo. Pricing. Read 18 November 2026. Vendor source.
Webtonic. 21 Ad Creative Agencies Worth Paying For in 2026. Read 18 November 2026. Industry source.
Free workbook
The Kill Sheet
Know when to kill an ad and when to let it run, with thresholds for the five signals that matter.
Five signals, three verdicts, ten minutes every Monday. Free.







