RankingRider

Extracts product metadata (titles, descriptions, tags, collections) from Shopify stores via CSV export, parsing Shopify's native product schema into a tabular format for batch processing. Uses Shopify's REST or GraphQL API to authenticate and retrieve product catalogs, then transforms nested product objects into flat CSV rows with column headers mapping to Shopify's product field taxonomy. Handles pagination for stores with 1000+ products and preserves product IDs for downstream re-import matching.

Generates optimized product titles using a fine-tuned language model that injects high-intent keywords (extracted from product category, tags, or user input) into natural-sounding titles. The model is trained on high-ranking Shopify product titles and follows SEO best practices: keyword placement in first 60 characters, inclusion of brand/category modifiers, and avoidance of keyword stuffing. Outputs multiple title variants (typically 3-5 options) so merchants can choose the best fit for brand voice.

Generates product descriptions using a language model that balances keyword inclusion with readability, targeting a keyword density of 1-2% for primary keywords. The model expands on product features, benefits, and use cases while naturally incorporating keywords in headers, opening sentences, and body paragraphs. Outputs descriptions typically 100-300 words, formatted with HTML line breaks or markdown for Shopify compatibility. Includes fallback logic to preserve existing descriptions if AI generation fails or produces low-quality output.

Accepts a CSV file with updated product metadata (titles, descriptions, tags, collections) and re-imports it back into Shopify using the REST or GraphQL API. Matches rows to existing products via product ID to ensure updates apply to the correct products, handles variant-level updates if applicable, and provides a transaction-like rollback mechanism if errors occur during bulk import. Validates data before import (e.g., title length, description HTML formatting) and reports errors per product so merchants can fix and retry.

Provides a free tier that allows merchants to optimize a limited number of products (typically 10-50) per month before requiring a paid subscription. The quota is tracked per Shopify store and enforced via API-level checks before AI generation or import operations. Free tier users can still export/import CSV and access the UI, but generation requests are rate-limited and queued. Paid tiers unlock higher quotas (100-1000+ products/month) and priority processing.

Uses Shopify's OAuth 2.0 flow to authenticate RankingRider without requiring merchants to manually copy/paste API tokens. Merchants click 'Connect Shopify Store' in RankingRider, are redirected to Shopify's OAuth consent screen, and grant RankingRider permission to read/write product metadata. RankingRider receives an access token scoped to products:read and products:write, stores it securely (encrypted at rest), and uses it for all subsequent API calls. Tokens are refreshed automatically before expiration.

Processes bulk AI generation requests asynchronously, queuing title and description generation for multiple products and returning progress updates via polling or webhooks. Uses a job queue (likely Redis or similar) to manage generation tasks, distributes them across multiple LLM API calls to parallelize processing, and stores results in a database for retrieval. Merchants can check progress in real-time via a dashboard showing 'X of Y products completed' and estimated time remaining. Handles failures gracefully by retrying failed products and reporting errors.

Automatically extracts or infers SEO keywords from product category, tags, and existing title/description using pattern matching and a keyword database. Maps Shopify product categories to common search terms (e.g., 'Women's Shoes' → ['women's shoes', 'ladies shoes', 'female footwear']), combines with merchant-provided tags, and ranks keywords by relevance and search volume. These keywords are then injected into AI-generated titles and descriptions to ensure topical relevance. Merchants can also manually override or add keywords.

Validates CSV data before import to Shopify, checking for common errors: missing product IDs, invalid title/description length, malformed HTML, encoding issues, and duplicate product IDs. Provides detailed error messages per row (e.g., 'Row 5: Title exceeds 255 characters') and allows merchants to download an error report CSV for batch fixing. For export, handles special characters, Unicode, and Shopify-specific field formatting (e.g., variant SKUs, metafield JSON) to ensure re-import compatibility.

Generates AI-optimized titles and descriptions but allows merchants to manually edit, approve, or reject each suggestion before import. Merchants can view side-by-side comparisons of original vs. AI-generated content, select from multiple AI variants, or manually edit the AI output. Only approved changes are imported to Shopify, giving merchants full control over brand voice and accuracy. Tracks which products were manually edited vs. AI-generated for future reference.