Ecommerce Product Image Prep Checklist

Answer-first workflow

Start by deciding the destination context first, then run each image through the same sequence: select, crop, rename, clean, export, verify, then hand off with a short decision log.

Source-to-delivery plan

Before touching a pixel, create a short delivery plan. This prevents drift across operators and prevents the common error of renaming or resampling before the rules are fixed.

• Choose source root for untouched captures or exports.
• Create an output root for processing copies only.
• Record platform target, channel, and maximum acceptable file size.
• Define accepted formats, naming pattern, and required shot counts.
• Document who signs off and by when.

Do not process files in place. The original master set must stay unchanged for any later recovery or audit.

Practical workflow

1. Ingestion: pull only candidate images from the source root, and leave everything else separate.
2. Selection: remove blurred, overexposed, or misframed shots, then lock the best sequence for each product variant.
3. Framing rule: apply one framing rule to the full group, for example equal product placement and background margins, so storefront grids look consistent.
4. Crop and resize: perform final crop after confirming all required variants are selected.
5. Background cleanup: remove artifacts only where they affect perceived quality, not brand marks or texture details.
6. Color and contrast pass: fix only visible errors; avoid global, aggressive adjustments that break material matching.
7. Rename with a deterministic pattern before compressing and before moving to final folders.
8. Compression pass: generate web copies using one quality profile that meets your destination limit.
9. Quick audit: run deterministic checks on dimensions, filename format, orientation, and visible artifacts.
10. Handoff package: export a copy list, decision notes, and failed-item log for the next owner.

Decision rules

• Rule 1: If a shot does not match the framing rule, reject it early even if the product is in focus.
• Rule 2: If metadata cleanup changes visual appearance, stop and use source metadata only if there is a catalog-level reason.
• Rule 3: If compression removes print detail or surface finish cues, lower quality settings and re-run.
• Rule 4: If file names do not sort in product order, do not proceed to packaging.
• Rule 5: If a correction takes more than two focused passes, hand off to manual review instead of further automation.

Limits that must be locked

The only way to avoid subjective churn is to set hard limits and treat them as non-negotiable during a batch:

• Dimension policy: one target size for primary listings, one for alternate views.
• Filename policy: one separator, one product id position, one variant token position.
• Compression policy: one image type per destination channel.
• Quality policy: one threshold for texture retention and one for allowed compression artifacts.
• Naming policy for rejects: explicit suffix so excluded files are not mixed with deliverables.

If any limit is unknown, pause and confirm with the catalog owner before editing the batch.

Pre-upload quality checks

Run these checks against the final export folder, not the master folder.

1. Count check: each product has the same number of angles required by the channel.
2. Label check: filenames decode to product id, variant, and view without manual guessing.
3. Consistency check: major edges, angle direction, and background edge distance look aligned across siblings.
4. Visual sharpness check: zoom 100 percent and ensure text, fabric, or fine tooling remains legible.
5. Compression check: open quickly in low bandwidth mode and confirm no blocked halos or banding in critical tones.
6. Context check: test at least one set inside the destination page mockup or storefront tile view.

Common mistakes

• Renaming before final shot selection, creating extra churn when variants are changed later.
• Mixing masters and exports, then accidentally compressing the wrong set.
• Over-processing colors to match one lighting setup and breaking another.
• Applying different crop ratios within one catalog group, causing row misalignment.
• Leaving rejected items in the same folder as accepted items without clear tags.
• Skipping context review and trusting only local zoom views.

Stop condition

Stop the run when all checks pass, all decisions are logged, and the deliverables differ from masters only by intentional, documented operations. Do not continue if any change is based on taste rather than a rule, or if additional edits are requested without updated acceptance criteria.

In short, stop when output matches the plan, metadata is clean, and the next owner can review the result without asking for hidden steps.

Handoff and related tool section

Package the handoff with three outputs in order: approved export folder, reject folder, and a short text note that lists rule exceptions and exact decisions.

• Batch Image Renamer for deterministic renaming once naming tokens are finalized.
• Image Tools section for compression and cleanup actions, then return immediately to this checklist for final validation.
• Keep a shared spreadsheet or notes entry with product id, shot count, and date stamp.

After handoff, the next owner should be able to continue from the decision log instead of restarting from the source files.

Batch Image Renamer is a practical fit after you set the token pattern and before final export folder freeze.

Decision and audit notes

The checklist is only useful if the result is reproducible. Record every decision point, because reproducibility is the real quality bar in ecommerce operations.

• Confirm source type, target platform, and accepted format before any edits begin.
• Use the smallest effective change first and re-check immediately; avoid stacked experiments.
• Track any rule that must be preserved across revisions, such as crop ratio, color baseline, and compression preset.
• Stop as soon as output meets acceptance and the next owner can complete handoff with clear context.
• For public-facing rollout, always run one final context check in the destination page environment, not only local preview.

Related workflow boundaries

This page covers only manual and browser-based preparation tasks. Do not use this checklist to introduce new shooting styles or new AI generation steps. If image quality or model training is required, route to a separate capture or production pipeline before this prep stage.