Catalog SEO Planner

Turn your English product records into a reviewable plan for product titles, meta descriptions and image alt text. Uses supplied text with deterministic rules; no model key, store credentials or additional subscription is required.

Start

1. Keep the sample Product records to try the example, or paste your own JSON array.
2. Supply 1-100 products, each with a unique id and nonblank title.
3. Optionally provide acronyms to preserve and a handle prefix to exclude.
4. Run the Actor. Open Complete plan and skip summary, then the OUTPUT JSON file, for the full report and Proposed changes for the downloadable results table.
5. Review each before/after value before applying changes with your own store tools.

Price

$0.05 per run, charged once when the Actor starts. Up to 100 products and all proposed changes are included; no extra per-result charge or platform-usage surcharge. The run uses 128 MB. Set your maximum cost per run to at least $0.05. The start charge also applies when input is rejected or no changes are needed, so validate your records against the limits below before starting.

Accepted data

The full input is a JSON object with products, optional acronyms (at most 100, 30 characters each), and optional skipHandlePrefix (at most 100 characters). Maximum serialized input is 1 MiB. The supported product fields are:

Field	Meaning
id	Unique nonempty string (200 characters max) or safe integer
title	Required nonblank product title, at most 1,000 characters
handle, vendor, category, productType	Optional strings, at most 1,000 characters
titleTag, metaDescription	Existing SEO text, at most 1,000 characters
seo	Optional object with title and/or description strings
description, descriptionHtml	Optional supplied copy, at most 10,000 characters
images	Up to 20 objects with optional id, alt, altText

Omit optional values that are unknown. null, unsupported fields, duplicate product IDs and out-of-limit inputs are rejected. Image alt strings have a 1,000-character limit; image IDs use the same string/safe-integer format as IDs.

Results

The complete JSON plan includes titles, metas, alts and an overall summary. Each section contains proposed before/after changes and relevant skip counts. Unresolved shared titles remain explicit instead of being silently overwritten. The dataset contains one row per proposed change, labeled with its section. An empty dataset can be a successful run with no changes; read the full plan.

Existing populated image alt text is preserved. The included two-product sample produces one title change, two meta descriptions and two alt-text suggestions. Applying those suggestions and running again produces no additional changes.

Limits and privacy

• English text heuristics only; no translation or search-ranking guarantee.
• Image alt text comes from metadata, not image recognition. Review it against the actual image before publication.
• No websites are fetched and no changes are written to a store.
• Input and output are stored in your Apify run storage under Apify's retention and access controls. Supply only authorized records; remove data when no longer needed.
• The hosted result does not grant redistribution rights to the implementation.

Support

Open an issue from this Actor's page with a minimal non-sensitive reproduction, the run ID, expected behavior and actual behavior. Do not post passwords, API keys, private exports or customer information.