Supplements
$60k/mo
- Before
- Control ad 40 days live, frequency past 4
- Diagnosis
- Hook rate falling while CPA still looked fine
- Hypothesis
- Pattern interrupt with a delayed product reveal
- Creative
- Skeleton-style rebuild
Blended CPA down 31% over 30 days
The creative supply system for scaling Meta accounts
Your media buyer isn’t the bottleneck. Your winning ad is dying.
Meta burns through creative faster than most teams can replace it. That is a supply problem, not a talent problem, and it is why accounts stall at the exact moment they should be scaling.
When your control starts dying, the next control is already on the runway.
- Which ads are already fatiguing
- The first sprint we would run, no charge
AR
Talking object / Concept
Sponsored · Advantage+
Hover to play
3s Hook38.4%
ROAS3.42x
- Hold rate 15s
- 21.7%
- CPA vs control
- -31%
- Days live
- 9
This is how it goes, every time, on every account.
- Day 1 to 4
Frequency low, CPA holds
The winner scales. Budget goes up. Meta pushes the same asset into a wider slice of your audience.
- Day 5 to 9
Hook rate slides first
Thumb-stop drops before CPA moves. Most teams miss it, because they are watching the cost column.
- Day 10 to 14
CPA spikes, budget gets pulled
Frequency crosses 3. The control ad dies. You are back to launching whatever is left in the folder.
By the time the cost column tells you, the ad has been dying for a week and the replacement has not been briefed.
You cannot replace winning creative as fast as Meta burns through it.
This is a supply problem. Spend goes up, the algorithm exhausts the audience faster, and the production calendar stays the same length it always was. Most accounts past $10k a month end up with one control ad carrying the revenue and nothing tested behind it.
You know how this goes
- 1Your media buyer says the account needs more creative.
- 2You ask the team for ten new ads.
- 3Two weeks later you have seven variations of the same concept.
- 4None of them becomes the new control.
- 5Your existing winner is now at 4.2 frequency.
- 6CPA starts climbing and the buyer pulls budget to protect the blended number.
- 7Everyone is in a call asking what to test next.
Nobody in that chain did anything wrong. There was just no system feeding the account.
Creative fatigue is not a production problem. It is economic leakage.
A fatiguing control ad does not announce itself as a cost. It shows up as acquisition efficiency quietly deteriorating while spend stays flat. Put a number on that deterioration and the conversation changes.
- At $30k/mo on Meta
- $4,500/mo
- lost to a 15% efficiency drop
- At $100k/mo on Meta
- $20,000/mo
- lost to a 20% efficiency drop
- At $300k/mo on Meta
- $60,000/mo
- lost to a 20% efficiency drop
Illustrative, not a forecast. Run it against your own numbers on the audit call.
The mechanism
The 10-Day Creative Runway™
A closed-loop system for replacing fatigued Meta creative before it becomes a CPA problem. We read the account, name what is decaying, form the next hypothesis, build it, ship it, then resprint whatever the data says is pulling. Every ten days, whether or not anything broke.
It is a loop rather than a project, which is the reason it keeps working after the first batch of ads stops.
Where you are
Winning ads fatigue faster than you can replace them.
Where this puts you
A predictable supply of new, tested creative before the current control dies.
- Get your current ads reviewed
- Free hook swipe file (47 openers)
- First sprint scoped on the call
Fill out the formBook your slot
Creative Fatigue Audit
Let’s find out what is killing your control ad.
October 2026
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Please fill out the form before choosing your time slot.
Five stages. Ten days. Then it starts again.
The same loop runs whether you are on the regional tier or the enterprise line. The last stage feeds the first, which is why the account never runs dry.
- 01
Days 1 to 2
Diagnose
We read account-level creative decay. Asset-level 3-second hook rates, hold rates and CPA decay, so we can name which ads are dying and how long they have left.
Fatigue map of your library
- 02
Days 2 to 3
Hypothesize
We pick the next angle, hook and format against what the audit found. Every concept goes into production with a stated reason it should beat the control ad.
Sprint brief your buyer signs off
- 03
Days 3 to 7
Produce
Scripts, storyboards, creator briefs, claymation and 3D built in parallel from assets you already own. No shoot required to start.
Net-new cuts in every placement
- 04
Days 8 to 10
Deploy
Ads land ready to upload in 9:16, 4:5 and 1:1. Named, tagged and mapped to the hypothesis so results can be read back against it.
Delivered on a date you planned for
- 05
Day 15, then repeat
Resprint
We cut the data again mid-month. Winners get rebuilt across new hooks before they fatigue, losers get killed, and what we learn sets up the next diagnosis.
Live data decides the next cycle
We don't pick formats because they look cool.
We pick them from the problem inside your account. Claymation is not a style choice, it is what you reach for when an audience has stopped seeing your ads. Three diagnoses, three different jobs for the creative to do.
Hover to play
What we see
Thumb-stop rate is falling while frequency climbs. The audience has seen this shape of ad from you, and from everyone else in the category, too many times.
What we reach for
- Claymation
- Crochet
- Talking objects
- Skeleton
The hypothesis
Novelty can restore thumb-stop rate.
Formats the feed has no reference point for. The scroll stops before the brain finishes classifying it as an ad, which buys the first three seconds back.
The problem is not who makes your ads. It is when the work starts.
Most creative gets requested. Something goes wrong, somebody asks for assets, and the clock starts then. By the time the replacement lands, the ad it was meant to replace has been dying for three weeks.
The request model
Starts after the damage
- 1CPA climbs far enough that someone notices
- 2Media buyer asks for more creative
- 3Someone writes a brief
- 4Freelancer or editor picks it up
- 5Revisions
- 6Delivery
- 7Launch
- 8Wait for performance data
- 9Repeat, once the next ad is already dying
Three to five weeks, triggered by a problem. The ad is dying the entire time the replacement is being made.
The runway model
Starts before it
- 1Account data read on a fixed clock
- 2Hypothesis formed from what is decaying
- 3Production starts the same week
- 4Delivery on a date your buyer planned for
- 5Launch into a live account
- 6Data cut again at day 15
- 7Resprint decides what gets rebuilt
Ten days, triggered by the calendar. Nothing has to go wrong for the next concept to already be in production.
The difference is not production quality. It is whether the work starts on a problem or on a clock.
You are not buying videos. You are buying the end of thinking about creative production.
Three things stay with you. Everything between the account data and the delivered ad is ours.
What you do
Three things
- Walk us through the ad account on the call
- Approve the sprint brief
- Ship product to creators, only for creator-led cuts
What we do
Everything else
- Read the account and map what is fatiguing
- Decide which hypothesis gets tested next
- Write the scripts and storyboards
- Brief and manage the creators
- Build the claymation, 3D and animation
- Edit, grade and caption every cut
- Localise into your markets
- Deliver in every placement ratio, named and tagged
- Cut the data mid-month and resprint the winners
See the formats in motion.
Claymation, skeleton builds, podcast slices and talking objects. Rest your pointer on any clip and it plays, or tap one on a phone.
- Skeleton style
- Claymation
- UGC podcast sliceCPC$0.33
- Talking object
- Crochet style
- Talking head + B-roll
- Cinematic short filmCVR+212%
- Skeleton style
- Claymation
- UGC podcast sliceCPC$0.33
- Talking object
- Crochet style
- Talking head + B-roll
- Cinematic short filmCVR+212%
Not just what moved. Why it moved.
A result on its own is a claim. The same result with the diagnosis and the hypothesis behind it is a method you can judge. Here is the full chain on three live accounts.
- 38%
- Best-in-account 3s hook rate
- -31%
- CPA vs. previous control ad
- 19
- Net-new assets per flagship cycle
- 9 days
- Average brief to delivery
Skincare
$120k/mo
- Before
- Static-heavy library, nothing holding past 10 seconds
- Diagnosis
- Product needed explaining before it could be wanted
- Hypothesis
- Show the transformation instead of describing it
- Creative
- Claymation ingredient build
15s hold rate 9.4% to 24.1%
Home & tech
$22k/mo
- Before
- One winner carrying four markets, no succession
- Diagnosis
- Same creative burning out at different rates per market
- Hypothesis
- One concept, localised, rather than four separate builds
- Creative
- Talking-object cuts in four languages
ROAS 2.1x to 5.8x on prospecting
What the buyers say once the loop is running.
Feedback from the people who live inside the ad account: growth leads, media buyers and the founders signing off on spend.
Our CPA went from $45 to $9
Our account had sat between $40 and $50 per acquisition for as long as I can remember, and we had stopped expecting that to move. Then they sent a claymation ad. I did not think it would work for a business like ours. It came in at $9. Same offer, same landing page, same budget behind it.
Sander
CEO, Lead Printer
ROAS from 2.2 to 4.6 by month two
Before The Ad Runway, our Meta account was running on 4 winning creatives we'd recycled for months. CTR had slid to 0.8%, and every scale attempt past $500/day broke the account. In the first 2 weeks, they shipped 14 new concepts. Not variations of one ad. Genuinely different angles, formats and hooks. By month 2:
- CTR went from 0.8% to 2.9%
- We scaled spend from $500 to €1200/day without performance falling apart
- ROAS moved from 2.2 to 4.6
The biggest difference is we're no longer waiting for one winning ad to die. There's always a new batch in testing, so fatigue stopped being a crisis and became a routine.
Cev
CMO, BellaLab
Our ROAS climbed from 2.4 to 3.9
Honestly, it's the creatives. We used to run a couple of safe ads aimed at the same customer we'd always had. The Ad Runway came in with a completely different range of concepts, and some of them landed with people we'd written off as not our market. That's the part I didn't expect. It's not just better ads. It opened up a segment of customers we genuinely thought we'd never reach, and they're buying.
Nick
CEO, Swishtime
The person who diagnoses the problem is the person responsible for solving it.
No account managers, because the role does not exist here. Mart reads your account and decides what gets built. Rico is accountable for that decision becoming a finished asset inside ten days. Neither can hand the consequences to anyone else.
Mart Hendrik
Reads the account, decides what gets built
Runs the diagnosis. Which ads are decaying, which hypothesis is worth the next cycle, and what gets cut when the data says it lost. If a concept is dropped, he is the one who tells you why.
Rico Antoniak
Turns the hypothesis into an asset in ten days
Owns the clock. Animation builds, creator briefs, 3D and the delivery date your buyer plans budget against. If a sprint is late, there is nobody else to ask.
Behind them
Editors, 3D artists and a performance analyst
Tallinn, Estonia
Animation and edit studio
Creator characters
Designed and built in house
Retainers sized to your monthly Meta spend.
Priced against the spend it protects, not the number of files it produces. The question is not whether this is expensive for creative. It is whether the loop prevents more waste than it costs.
For Accounts Spending $10k–$30k/mo
Two new concepts every cycle, so the account always has something untested to run.
From $1,989
per month
At $20k a month, a 15% efficiency drop is $3,000 gone. This costs less than the leak.
- 2
- Core concepts
- 7
- Total ads
- 4
- Markets
- 2 core concepts per cycle
- 7 total ads across hooks and angles
- Creative built to recover declining hook rates
- Localised audio and captions in ET, LV, LT and EN
Flagship
For Accounts Spending $30k–$100k/mo
Five new concepts every cycle, chosen from what your account data says is decaying, plus a rebuild of the winners at day 15.
From $4,950
per month
At $50k a month, a 15% efficiency drop is $7,500 gone. This costs less than the leak.
- 5
- Core concepts
- 19
- Net-new variations
- 2
- Resprints
- 5 core concepts per cycle
- 19 net-new variations
- Formats chosen from the diagnosis, not from a menu
- New concepts every 10 days, based on your account data
For Accounts Spending $100k–$300k+/mo
Eight new concepts every cycle across your markets, with licensing and creator usage scoped to your portfolio.
From $9,900
per month, scoped to the account
At $200k a month, a 20% efficiency drop is $40,000 gone. This costs a quarter of it.
- 8
- Core concepts
- 32
- Deliverables
- Full
- Custom builds
- 8 core concepts per cycle
- 32 total deliverables
- Custom asset and 3D animation production sprint
- Bespoke claymation and cinematic short films
All tiers: 10-day sprint cadence · raw files and source projects included · no long lock-in
Next cycle: 4 of 9 sprint slots remaining.
We cap the roster so every account gets a real audit and a named producer. When the cycle fills, the next intake opens the following month.
The Creative Fatigue Audit
Thirty minutes. We tear down your top three ads live, map where your library is decaying and show you the first sprint we would run. You keep the audit either way.
- 01Which of your ads are already showing fatigue
- 02Where the decay begins, and how long the control ad has left
- 03Why your current control is vulnerable
- 04Which creative hypotheses we would test next, and in what order
- 05What your first 10-day sprint would look like
Fill out the formBook your slot
Creative Fatigue Audit
Let’s find out what is killing your control ad.
October 2026
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Please fill out the form before choosing your time slot.
Notes from the production line.
What we learn running creative against live spend, written up for the people who have to make the next call.
About
What The Ad Runway is, and who it is not for
A plain explanation of what we do, what a sprint contains, what it costs, and the kinds of accounts we are the wrong choice for.
18 Sep 20266 min read
Channels
ChatGPT ads: what DTC brands scaling Meta should do with them
How the ads work, what they cost, how targeting differs from Meta, and how to test the channel without starving what already works.
19 Sep 202612 min read
Creative strategy
Ad fatigue: how to spot it before your CPA moves
Hook rate slides days before cost does. Here is the exact sequence of metrics that tells you an ad is dying while it still looks profitable.
28 Aug 20268 min read