Observation
What changed in the available data? Which assets received enough delivery to examine? Where do concept, hook, format, audience, placement, or time differ?
Paid-media analysis for creative decisions
Read the account for questions, not easy answers.
ShelfHatch reviews agreed paid-media data to identify creative patterns, coverage gaps, fatigue questions, and possible next tests. We keep observation, interpretation, and recommendation separate.
Inventory / Next brief
ConceptGroup the idea
HookName the opening
OfferRecord the context
FormatSeparate the cut
WindowCompare fairly
QuestionBrief what follows
Analysis inputs
Connect performance records to the creative that produced them.
Metrics and field definitions vary by platform and account. ShelfHatch confirms the comparison window and source before drawing observations.
- Creative IDs, names, thumbnails, and launch dates
- Campaign, audience, placement, and spend context
- Available view, hold, click, conversion, CAC, and ROAS fields
- Offer, product, inventory, and landing-page changes
- Past briefs, concept labels, and production notes
- Known measurement or attribution limitations
Analysis layers
Three layers keep the readout honest.
Interpretation
What are plausible explanations? Which product, offer, media, landing-page, or measurement factors could also account for the pattern?
Recommendation
What should the team keep, adapt, pause, or test next—and what would a better test need to control?
Typical outputs
A readout designed to brief production.
Outputs connect a documented observation to a bounded next-test question. See the creative testing system.
- Creative inventory grouped by concept family
- Naming and metadata gaps
- Coverage map across angles, proof, formats, and funnel roles
- Performance observations with comparison context
- Fatigue or delivery questions needing validation
- Prioritized next-test hypotheses
- Brief-ready scripts, shots, hooks, edits, or adaptations
Boundaries
What paid-media analysis does not prove on its own.
Creative caused the change.
Offer, audience, spend, placement, landing experience, seasonality, and tracking may also explain a pattern.
One metric predicts profit.
Platform reports do not automatically establish incrementality, profit impact, or long-term value.
The answer transfers.
Results from one audience, placement, period, or offer may not carry into another context.
- A low-delivery asset is not automatically a settled loser.
- Reported attribution may not capture every customer interaction.
- A past winner should not be copied indefinitely.
Access and handling
Use the least access needed for the agreed review.
ShelfHatch can work from customer-prepared exports or agreed view access. The engagement should define the source, fields, date range, account owner, permitted use, retention, and deletion or return process before data is shared.
Do not send credentials or customer-level personal data by email.
Least-privilege checklist
01Defined source
02Agreed date range
03Necessary fields only
04Named account owner
05Return or deletion plan
Media-analysis questions
Context before conclusions.
Which platforms can you review?
Meta, TikTok, YouTube, and other paid-media sources can be discussed when the necessary creative-level fields and context are available. Reviews use customer-prepared exports or agreed view access rather than a native data connection.
Do you need administrator access?
Not by default. A scoped export may be sufficient. If view access is useful, the client should grant the least privilege needed through the platform's normal access controls. Credentials should never be shared.
Will you change campaigns or budgets?
Not under the core analysis service. ShelfHatch provides creative observations and recommendations. Campaign changes remain with the client's media owner unless a separate written scope states otherwise.
Can you calculate our true incrementality or profit impact?
Not from platform reports alone. Incrementality and profit analysis require appropriate experimental design and reliable business data beyond a standard creative review.
Start with context
Bring one account question and the context around it.
Tell us which product, offer, channel, date range, and creative decision matter. We will identify the data and labeling needed for a useful review.
Prepare the analysis question