Do These Supplier Images Need to Be Reshot Before I Advertise the Product?

01

What should you record before reviewing an image?

Create one asset record per supplier file so that its source, rights status, product variant, evidence, and decision do not become separated.

Record the supplier, product or SKU, variant, source URL or original filename, date received, and any statement about how the image was created or edited. Preserve the original file rather than working only from a compressed screenshot.

Add fields for written reuse permission, Content Credentials or other provenance metadata, physical-sample availability, reviewer, evidence location, open questions, and final decision. Use one of four decisions: research only, verify, reshoot, or reject.

Permission must be reviewed independently from product accuracy. The U.S. Copyright Office explains that photographs published online remain protected and that using another person’s photograph generally requires permission from the copyright owner. A supplier’s possession of a file does not, by itself, document your right to advertise with it. Use the supplier asset intake guide to collect files, source information, and permissions consistently. (U.S. Copyright Office)

1. Supplier, product or SKU, and exact variant
2. Source URL, original file, and date received
3. Supplier’s stated creation or editing method
4. Written reuse permission: yes, no, or unclear
5. Content Credentials or metadata: present, absent, or not checked
6. Physical sample: available or unavailable
7. Decision, reviewer, evidence location, and open questions

02

What counts as a visible product promise?

A visible promise is any product fact or experience that a reasonable shopper could infer from an image, even when the ad copy does not say it explicitly.

List the image’s apparent product shape, dimensions, proportions, color, finish, transparency, texture, and material. Record the number of units shown, included accessories, packaging, labels, controls, ports, seams, openings, fasteners, and attachment points.

Then note contextual promises: fit on a person, compatibility with another object, scale relative to a hand or room, intended environment, movement, transformation, capacity, strength, waterproofing, cleaning ability, heat resistance, durability, or a before-and-after result. Shadows and reflections matter when they change perceived depth, finish, transparency, or scale.

Google Merchant Center’s image guidance requires the image to represent the correct product and variant, including details such as color, pattern, and material. It also warns against showing accessories that are not sold with the main product. This is a useful review principle even when you are preparing assets for another channel. (Google Merchant Center)

03

How do you assign an evidence level to each promise?

Grade each visible promise separately. One image may contain a verified product color, an unsupported accessory, and an unrepeatable performance demonstration.

Use five evidence levels. None means the supplier image is the only support; keep that promise research-only. Documentary means a listing specification, dimension sheet, package list, or written supplier statement supports it; verify important advertising details against the sample. Sample-observed means you inspected the same SKU and variant. Reproduced means you recreated the depicted setup or result under normal conditions. Contradicted means the sample, packaging, documentation, or repeated test disagrees with the image; remove the claim and reshoot or reject the asset.

Supplier confirmation is useful evidence, but it should not automatically settle performance, health, or safety claims. The FTC says advertisers need a reasonable basis for express and implied objective claims before dissemination; what constitutes reasonable support depends on factors including the claim, product, consequences of error, and evidence experts would consider appropriate. (FTC advertising substantiation policy)

1. Write down each visible promise.
2. Assign none, documentary, sample-observed, reproduced, or contradicted.
3. Attach the relevant file, photograph, test record, or supplier statement.
4. Choose research only, verify, reshoot, or reject for that promise.
5. Set the image’s overall status according to its least-supported material promise.

04

When should a supplier image remain research only?

Keep an image internal when it is useful for ideas or investigation but its product depiction, rights, or relationship to your SKU is unresolved.

Research-only images can support moodboards, composition studies, positioning analysis, sample-inspection checklists, claim discovery, and shot planning. They can also reveal what suppliers or competitors want shoppers to believe without treating those presentations as established facts.

Use research-only status when product details vary between images, scale references may be artificial, accessories lack a confirmed package list, packaging text is malformed or inconsistent, the scene cannot be reproduced, the variant cannot be tied to the SKU you will ship, or reuse rights are missing.

Label the status clearly and prevent the file from quietly migrating into an ad, product page, marketplace listing, email, or organic social post. Customer-facing organic use does not resolve an accuracy or permission problem.

05

What should you verify against the real sample?

Verify every plausible but unconfirmed detail against the exact SKU and variant customers will receive, not merely a similar catalog item.

Compare the image with the current supplier listing, purchase order, written dimensions and materials, package-content list, unboxing record, physical sample, variant selector, and landing-page copy. If manufacturing consistency is important, inspect more than one sample rather than assuming one unit represents every shipment.

Check geometry from all sides; dimensions beside a ruler; color under neutral and ordinary room lighting; material through close inspection; and every included item through a continuous unboxing. Confirm that ports, controls, attachments, packaging, instructions, compatibility, and depicted use match the shipped variant.

Use the sample custody and condition log to connect observations to a received unit. Google requires ads and landing pages to refer to the same product or variant, reinforcing the need to compare the creative, selected offer, and destination together. (Google Merchant Center landing-page requirements)

1. Confirm the SKU and variant on the order, package, and product.
2. Photograph the unopened package and a complete unboxing.
3. Capture front, back, sides, top, and bottom.
4. Measure dimensions and inspect materials, finish, controls, and accessories.
5. Compare each observation with the supplier image and landing page.
6. Record discrepancies and move unresolved promises into the shot backlog.

06

When is a real-sample reshoot mandatory?

Reshoot when an image is the main support for a material customer expectation that you cannot independently verify and reproduce.

Automatic triggers include a visible mismatch with the sample; changing proportions; disappearing controls, seams, or attachments; an unreproducible material finish; distorted scale; unconfirmed accessories or quantities; inaccurate packaging or labels; and a pictured SKU that cannot be identified.

Also reshoot demonstrations or scenes implying waterproofing, strength, capacity, speed, cleaning, heat resistance, durability, transformation, or before-and-after results unless you can substantiate and recreate the presentation. Reject unsafe or instruction-inconsistent setups rather than reproducing them.

The FTC states that a visual presented as an actual demonstration must truthfully represent what the product can do. TikTok separately prohibits exaggerated effects, misleading comparisons, inconsistent ad and landing-page information, and deceptive before-and-after presentations. A label such as “illustration” or “AI-generated” does not correct a misleading central impression. (FTC; TikTok Advertising Policies)

07

How should you build the reshoot backlog?

Turn each unresolved promise into a specific shot or test with a pass condition, owner, priority, and evidence location.

Use P0 for details that determine what the buyer receives or what the product can do. Examples include a complete unboxing flat lay for included accessories, a ruler-based scale shot, or a continuous demonstration under normal use. Use P1 for material, finish, controls, and all-angle coverage. Use P2 for realistic lifestyle context that does not introduce unsupported performance implications.

For every row, record: priority; unresolved promise; missing evidence; required shot or test; pass condition; owner; sample identifier; status; and approved output. A useful pass condition is observable—for example, “every delivered item is visible” or “dimensions and proportions match the received sample”—rather than subjective wording such as “looks authentic.”

Archive rejected shots and test conditions as well as approvals. This preserves why a scene failed and helps prevent the same unsupported concept from returning in another brief.

1. P0 — Included accessories: shoot a continuous unboxing and complete flat lay.
2. P0 — Scale: show verified dimensions with a ruler and undistorted reference.
3. P0 — Demonstrated result: record the complete process without hidden edits.
4. P1 — Material and finish: capture neutral-light close-ups.
5. P1 — Controls and ports: photograph every side of the exact variant.
6. P2 — Lifestyle context: stage a realistic, instruction-consistent use scene.

08

Can Content Credentials prove an image is trustworthy?

No. Content Credentials can provide tamper-evident provenance information, but they do not prove that the depicted product, configuration, or result is factually accurate.

The C2PA specification defines a framework for recording assertions about an asset’s provenance and edits. That information can help you understand where a file came from and how participating tools handled it. Preserve credentials and original files when they are available. (C2PA specification)

Absence of credentials does not prove that an image is false. Credentials can also be incomplete or removed, and provenance does not establish the truth of the depicted subject. Treat it as one evidence source alongside supplier documentation and physical inspection, not as an approval badge. (C2PA explainer)

Meta describes creative tools that can generate backgrounds, expand images, animate static images, and generate variations. Any generated derivative therefore needs a fresh review for changed geometry, scale, accessories, or implied performance. Follow the AI creative provenance workflow to preserve originals, metadata, and derivative history. (Meta Advantage+ creative)

09

How can ATIYO support this workflow?

ATIYO can keep the asset decision, visible claims, evidence, reshoot brief, iterations, and reusable learning connected in one creative-strategy workflow.

Use ATIYO to record why an asset was classified as research-only, verification required, reshoot, or rejected. Attach the relevant brand context, sample observations, open questions, shot backlog, and approved iteration so later collaborators can understand the decision without reconstructing it from messages.

ATIYO does not authenticate an image, inspect a physical product, connect to ad accounts, buy media, calculate ROAS, or independently know performance. Media performance remains in the ad platform. ATIYO preserves creative context and learnings. Any performance result must be recorded by the user if it should inform later briefs or iterations.