How many assets do I get a month?

That is the number we agree up front, and it is what sets your rate. Ten statics a month is one arrangement, thirty assets including twelve videos is a different one. We would rather commit to a volume you can afford indefinitely than deliver a large first month and then taper, because the whole value of this is that it does not stop.

Do you shoot original footage?

We can, and it moves the price. The cheaper and usually better route is working from what already exists — your product shots, customer clips, and creators you already have relationships with — which we edit, caption and cut into variations. If you supply raw footage and we only edit, you sit at the lower end of the range.

Who runs the ads, you or us?

This service produces the assets and the brief that explains what each one is testing. Running them, setting budgets and making the kill decisions can be yours or ours — it is a separate arrangement either way. Plenty of advertisers buy creative from us and have someone else in the account, and that works fine as long as the naming convention is respected.

My last agency made beautiful ads and they did not work. Why would this be different?

Probably because they made few of them and made them from the brand's own language. We work from your reviews and support messages, in the customer's words, and we produce enough variations that the winners have a chance to show up. Quality still matters — it just loses to volume plus honest source material more often than people expect.

How do I know which creative is actually working?

Because every asset is named to encode its angle, its hook and its version, so the platform's own reporting groups them without extra work. Without that, a month of testing produces a list of filenames and no conclusions, which is the usual state of affairs.

How often does creative need replacing?

Sooner than a fixed production schedule tends to allow, and the account tells you when. Watch frequency and CPM together — when the same people are seeing an ad three or four times over and the cost to reach them starts climbing while click-through falls, the creative is finished. Nothing is wrong with the account; the audience has simply seen it.

Paid Media & Performance

Ad creative production

A standing supply of ads — angles pulled from your own reviews, one variable changed per version, named so your reporting can tell you which idea worked.

From: $200–$2,500/mo
Typical time: ongoing
Usually bought by: Performance marketers

What this actually solves

On Meta and TikTok, the creative is the targeting. The audience tools do less than they used to and the algorithm does more, which means the thing that decides whether an account scales is how many genuinely different ideas you can put in front of it. An advertiser producing two ads a quarter cannot find a winner, because winners are found by elimination and there is nothing to eliminate.

The second half of the problem is decay. Creative does not fail gradually and gracefully - it works, and then it stops. The audience has seen it. Frequency climbs, the cost to reach a thousand people climbs with it, click-through slides, and the cost per acquisition follows about a week later. Nothing in the account changed. The owner starts adjusting budgets and audiences looking for the fault, and the fault is that the ad got old. Every account that scales has a supply line behind it, and most small brands have no designer at all, let alone one producing on a weekly rhythm.

The raw material for good ads is already in the business. It is sitting in the reviews, the support inbox and the sales calls, in the customer's own phrasing. The best-performing hook is almost never the sentence a marketing team would have written — it is the odd, specific thing a real buyer said about why they finally bought.

What we do, and in what order

Mine the source material. Your reviews, support messages and any recorded sales calls, read properly rather than skimmed, with hooks pulled out verbatim and the review they came from kept attached to each one. Alongside that, a pass through the Meta Ads Library on the competitors and the big spenders in your category, saved into a swipe file in Foreplay. Two days at the start, then about half a day a week.
Build the angle map. The hooks get grouped into angles — distinct arguments for why someone should buy, each aimed at a different kind of buyer. Ten to fifteen angles is a healthy starting map. Each gets a name that will survive into the file naming, so that reporting later reads in ideas rather than filenames. One to two days per cycle.
Produce in variations, not one-offs. Each angle becomes three to five executions - different hook, different opening shot, different thumbnail, same underlying argument. Statics and carousels in Canva, video in CapCut, generated backgrounds and concepts in Ideogram or Midjourney where a shot does not exist, voiceover in ElevenLabs and presenter video in HeyGen where they earn their place. Rolling, five to ten assets a week.
Name everything to a convention, on delivery. Angle, hook, format and version encoded in the filename, so that the platform's own reporting and tools like Motion can group by idea. This costs minutes per asset and is the difference between a quarter of testing that concludes something and one that does not. Same day as each delivery.
Read the results and retire the dead. Monthly, against frequency, CPM trend and click-through together rather than any one of them alone — which angles are still earning, which have fatigued, which never worked and should not be re-cut. That reading feeds the next month's angle map. Half a day a month.

What moves the price

The range is $200–$2,500/mo and three things place you in it.

Volume and format mix. Statics are quick. Video is not, and a month with twelve edits in it is a different production load from a month of ten images, however similar the brief sounds.

Where the footage comes from. If you hand over raw clips and product photography and we cut them, the rate sits low. If we are sourcing creators, directing what they shoot, and handling the back-and-forth on revisions, that is a different job with people-management in it.

How much has to be invented. A brand with a strong body of reviews, clear photography and a defined visual language gives us a running start. A brand with none of that needs the look established first, and the first month is slower than the ones after it.

What usually goes wrong

One great ad gets made, and then everyone waits. This is a numbers business. A winner is found by elimination, which means most of what goes into a month exists in order to be eliminated, and the few that survive are what carry the account. If two ads were produced and neither worked, the problem is almost never craft - there simply were not enough shots taken for a winner to appear. Judge the process by how many distinct ideas got tested, not by how good the best one looked in the deck.

Everything changes between versions. New hook, new footage, new music, new caption, new thumbnail, all in one revision. Something moves and nobody can say what did it, so the next round is guesswork again. Change one variable per version — the hook alone, or the thumbnail alone — and after a month you own a piece of knowledge about your market rather than a folder of results.

The ads get written in the brand's voice instead of the customer's. Copy drifts towards how the company describes itself, and the specific, slightly awkward sentence that a real buyer wrote in a review gets smoothed into marketing language. The smoothing is what kills it. We also hold a hard line here — if a benefit does not appear in the source material, it does not go in the ad, because an invented claim is both worse-performing and, in regulated categories, an actual problem.

See where the site stands before you brief anyone.

If the site is already live, start from what it scores rather than from a guess. The audit reports what it checked, what it found, and — the part most tools leave out — what it could not check, and why.

See the website audit Talk about this work

Questions people ask before they commit.

How many assets do I get a month?: That is the number we agree up front, and it is what sets your rate. Ten statics a month is one arrangement, thirty assets including twelve videos is a different one. We would rather commit to a volume you can afford indefinitely than deliver a large first month and then taper, because the whole value of this is that it does not stop.
Do you shoot original footage?: We can, and it moves the price. The cheaper and usually better route is working from what already exists — your product shots, customer clips, and creators you already have relationships with — which we edit, caption and cut into variations. If you supply raw footage and we only edit, you sit at the lower end of the range.
Who runs the ads, you or us?: This service produces the assets and the brief that explains what each one is testing. Running them, setting budgets and making the kill decisions can be yours or ours — it is a separate arrangement either way. Plenty of advertisers buy creative from us and have someone else in the account, and that works fine as long as the naming convention is respected.
My last agency made beautiful ads and they did not work. Why would this be different?: Probably because they made few of them and made them from the brand's own language. We work from your reviews and support messages, in the customer's words, and we produce enough variations that the winners have a chance to show up. Quality still matters — it just loses to volume plus honest source material more often than people expect.
How do I know which creative is actually working?: Because every asset is named to encode its angle, its hook and its version, so the platform's own reporting groups them without extra work. Without that, a month of testing produces a list of filenames and no conclusions, which is the usual state of affairs.
How often does creative need replacing?: Sooner than a fixed production schedule tends to allow, and the account tells you when. Watch frequency and CPM together — when the same people are seeing an ad three or four times over and the cost to reach them starts climbing while click-through falls, the creative is finished. Nothing is wrong with the account; the audience has simply seen it.

Often taken together.

Start with the brief, not the invoice.

Send what you are trying to achieve. You get a written scope, a price and a date — or an honest “this is not the right service for that”.

Talk about Ad creative production All Paid Media & Performance services