When your scaling campaign and your testing pool share a budget, the signal you call creative fatigue may be something the architecture itself produced.
A decision framework — 1 section · ≈150 words · 3 sourced, linked quotations
Edition pricing to be announced.
14-day unconditional refund
The Gap
It is designed to shift how readers evaluate creative efficiency: away from volume as a proxy and toward the structural and strategic conditions the evidence identifies as causal.
The Evidence
The lever the evidence licenses is a structural separation between a scaling campaign and a testing pool. “The fix is splitting into two campaigns: 1/ Scaling CBO (~$120/day) - only proven winners live here.” [Q1] That architecture matters to the refresh-or-scale decision because it changes what the dashboard signal actually means.
The cost of choosing the testing side of the lever is that it demands more from the practitioner upstream. “You need twice the creative volume and earlier signals to know when to rotate.” [Q2]
“It is part of getting serious.” [Q3] The separation between scaling and testing is not a sophistication-level optimization; it is the prerequisite condition under which a practitioner's judgment about when to refresh, kill, or scale a creative becomes trustworthy at all.
Confidence 80%
Scored against the cited record — claims the evidence didn't support are refused, never softened into a hedge.
The diagnosis and recovery brief. The evidence-backed read on what recovery requires — and what it refuses to promise.
The offer
Without structural separation, the data you act on when you decide to refresh, kill, or scale a creative is generated by a system mixing two different questions into a single number. You cannot isolate fatigue from competition, or rotation from noise, because the structure never asked you to.
The decisions you make on that number compound. Scaling a creative that was only winning because it shared spend with weaker tests. Pulling something that looked tired but was outbid. Each call feels considered. None of them are defensible on the record.
When scaling and testing live in separate campaigns — with only proven survivors on the scaling side — each side produces a signal that is actually about what you think it is. Your judgment about when to refresh, kill, or scale becomes something you can cite a reason for, not just a read you acted on.
What the full record includes
You can stop second-guessing whether a creative is fatiguing or simply being outcompeted inside a mixed pool — the separation answers that structurally, before you spend against the wrong read.
You can see what your scaling spend is actually buying, because the architecture keeps unproven creative out of that pool entirely and the dashboard reflects that discipline.
You can decide when to rotate on the testing side with earlier signals — the structure is built to surface those signals before the spend question becomes expensive, though it asks more creative volume from you upstream to work at all.
You stop treating this as optional sophistication — what the product facts establish is that this separation is the prerequisite condition under which your creative judgment is trustworthy, not an optimization layered on top of one.
What you receive
1 section · ≈150 words · 3 sourced, linked quotations — the full record, nothing summarized away.
Read on the web + a machine-readable markdown edition.
Access by email link — yours to keep. Revoked only if refunded.
Edition pricing to be announced.
14-day unconditional refund
The honesty apparatus
Every claim in this record carries a confidence score — the mean here is 80% — and claims the evidence didn't license were refused, not softened.
Method 1 claim scored against the cited record.
What this refused to claim
Most sales pages claim everything. This record refused 22 claims the evidence didn't support — they're in the full record, struck through.
A practitioner staring at a declining ROAS on a dashboard faces two very different causes that look identical from the outside: the creative itself is fatiguing, or the creative is still sound but has been stranded in the wrong structural environment, competing against untested work in a pool that cannot give it a fair read. — The cited evidence did not support this claim as stated.
Conflating those two causes produces the wrong move almost every time. — The cited evidence did not support this claim as stated.
Rival readings
of the market's story this record examines — retained because the evidence doesn't exclude them
The narrative front-loads creative volume and angle-generation as the primary lever, but the evidence reveals that structural campaign architecture—separating testing from scaling budgets—is the practitioner-validated prerequisite that the market-facing message omits entirely.
Retained as a competing explanation not excluded by the cited evidence.
The omission is incidental: the narrative assumes practitioners already understand campaign structure and intentionally scopes its message to the upstream creative-production problem, treating architecture as a given rather than a gap.
Retained as a competing explanation not excluded by the cited evidence.
The two legs address sequentially distinct problems—narrative targets creative-supply bottlenecks while the evidence targets budget-allocation discipline—and the tension is a deliberate editorial split rather than a true omission.
Retained as a competing explanation not excluded by the cited evidence.
Questions
Is this only relevant once I am spending at a certain scale?
The product facts are direct on this point: the separation is not a sophistication-level optimization reserved for larger budgets. It is described as the prerequisite condition under which a practitioner's judgment becomes trustworthy at all. The question is not whether you are spending enough to warrant it — it is whether you want a defensible signal or not.
What does moving to this structure actually cost me?
More upstream than you may be doing now. The testing side of the architecture demands greater creative volume and earlier rotation signals before the spend question gets expensive. The structure does not reduce what is required of you — it changes where the work lives and makes the output of that work readable.
What qualifies a creative to move to the scaling side?
The product facts establish the structural rule — the scaling side holds only what has already survived the testing side's filter — but they do not set a universal threshold for what 'survived' means. That definition belongs to the practitioner. The architecture makes the answer to that question consequential in a way that a mixed-pool structure does not.
This is not a pitch for a more complicated system. The architecture either separates the two questions or it does not. If it does not, the signal is not clean. That is the whole thing.