Storeshift

How Storeshift works

Every step of the pipeline, in plain English. The honest version, with the actual things we run on your data.

The six phases

You watch all six happen live with item-level progress and a running cost-so-far display. Every phase emails you on completion.

  1. 1

    Scrape your marketplace storefront

    We pull every public listing: title, description, price, all images, the source URL. We respect the marketplace's robots.txt and rate limits; if anti-bot defenses fire, we automatically solve them via 2Captcha (we eat the cost) or pause the run cleanly so you don't lose progress.

  2. 2

    Scrape each individual listing

    For each item we pull the full detail page: high-resolution images, the long description, item specifics, condition, brand, MPN, UPC/barcode, weight, dimensions, quantity available. Multi-variant listings (size × color, size × material) get their full variation matrix captured: per-variant SKU, price, and inventory. Everything maps to the right Shopify column on export.

  3. 3

    Download and clean every image

    Every image saved locally for processing. We deduplicate so you don't pay for the same image twice on items that share photos.

  4. 4

    Background-remove with rembg

    Each product image runs through rembg, an open-source background removal model. Output is transparent WebP: ~30% smaller than PNG, faster Shopify page loads, and cleaner for TikTok Shop catalog review.

  5. 5

    Upload to a CDN, get permanent URLs

    Cleaned images uploaded to Cloudflare R2 (yours for 30 days; the CSV references them by URL). Shopify imports them by reference, so you don't have to re-upload anything when you import.

  6. 6

    Generate the AI copy and Shopify CSV

    For every product, the AI writes: a Shopify Title in your chosen voice (40-70 chars), the full Shopify Body HTML (paragraphs, structure, brand-aware copy), an SEO meta description (140-155 chars, unique sentence structure per product), a category assignment. Two review passes follow before the CSV is built: a copy QA pass that flags generic AI phrasing, repeated openings, and punctuation artifacts, then a TikTok Shop policy review against the Commerce policy categories. Both surface as per-listing badges on the Review page so you decide what to fix, keep, or ignore. Multi-variant listings emit one row per variant under a single Shopify Handle, with proper variant pickers on import. Output gets packaged into a Shopify-ready CSV plus a database mapping every SKU back to its source URL.

Quality gate

We check our own AI before you see it.

Storeshift is more than a rewrite button. Each run has deterministic checks around source facts, images, barcodes, and CSV structure so the result looks like a catalog, not a scraped marketplace dump.

Verify source facts

Generated titles, bodies, and SEO are checked against canonical facts pulled from the source listing. If copy mentions material, condition, compatibility, or included parts that the source did not support, the Review page flags it before import.

See the product

Image-product matching catches stock-photo placeholders, wrong-variant photos, and first images that may not depict the item being sold. We surface the warning so you can delete the item or ask support to retry with a better image.

Validate the boring details

GTIN check digits, CSV headers, required variant values, title length, and malformed HTML are checked before the file ships. The small rules are what keep Shopify and Merchant Center imports clean.

+1

Keep it fresh (optional)

After your first conversion, you can subscribe to a monthly or weekly re-pull. We scrape your eBay store again, dedupe against your existing Shopify catalog, AI-process only the new listings, and email a delta CSV. From $19/mo.

Up to 500 listings processed per cycle. Bigger catalogs sync over multiple cycles.

What's actually in the AI copy

Prompts are revised weekly as we close the gap on edge cases. Current production rules:

Shopify product titles

AI-rewritten in your chosen tone of voice. 40-70 characters, with a configurable suffix (e.g. " - YourBrand"). Distinctive design names and seller language are preserved for review; risky phrases are flagged after generation so you can keep, edit, or fix them with context instead of having the model quietly erase branding before you see it.

Source title (eBay):
  NEW Nike Air Max 90 Mens Size 10 White Black Sneakers Shoes Running NIB

Storeshift output (tone: Direct & benefit-focused):
  Air Max 90 - Classic Streetwear in White/Black, Size 10 - YourBrand

Full Shopify Body HTML

This is the full product page body, not the SEO meta. Each item gets a multi-paragraph description rewritten in your voice: opening hook, key benefits, materials/condition/specs where the source listing surfaced them, and a closing line. The eBay description is fed in as factual grounding so we don't hallucinate features the seller didn't claim. Output is HTML so Shopify renders it straight into the product page.

SEO meta descriptions

140-155 characters. We deliberately vary sentence structure across products in the same run so Google doesn't flag your catalog as duplicate-content boilerplate (the silent penalty that kills templated Shopify stores).

Category detection

Keyword match first against your store's category list. Items that don't cleanly match get sent to the AI for a structured call constrained to the valid category enum, so we can never invent a category your storefront filter doesn't know about.

Copy QA pass

A second pass over the AI-rewritten titles, bodies, and SEO descriptions before the CSV is built. We catch generic AI phrasing (the cliché filler that screams machine-written), repeated opening clauses across listings (the silent duplicate-content tell that templated stores ship by accident), and punctuation artifacts (em dashes, stray smart quotes) that read as machine output.

Findings don't auto-rewrite anything. They surface as per-listing badges on the Review page so you decide whether to refine or ship as-is. The QA layer is why your $19 catalog reads in your voice and not as a templated dump.

TikTok Shop policy review + review-and-fix UI

Each finished listing reviewed against TikTok Commerce policy categories: copyrighted IP, restricted substances, safety certifications, hate symbols, misleading claims (medical, "best," "guaranteed"), prohibited categories. Block-severity findings come back with a specific suggested fix, not a generic "rejected." Soft warnings come through as recommendations.

Flagged items show up on a per-product review screen where you can click Fix (re-prompts the AI with the violation context and the specific suggested change), edit the title or body inline, delete the item from the CSV with an undo button, or click Ignore for cases where you have legitimate authorization (an authorized brand reseller with trademark rights, for example).

We don't promise TikTok will approve every listing. Their review is opaque. We promise the obvious-rejection-reasons are handled before you even submit.

What you ship with

  • →shopify_import.csv: drop straight into Shopify Admin → Products → Import
  • →product_database.db: SQLite mapping every Shopify SKU to its source listing URL. You always know where each product came from.
  • →tiktok_review.json: full policy review report with violations, severity, suggested fixes
  • →Background-removed images on a 30-day CDN URL (you import by reference; Shopify mirrors them on first import)

What you can configure

  • →Run kind: Convert (default; every scraped listing lands in the CSV) or Sync (you upload your existing Shopify products CSV, we dedupe by title and only AI-process the new items). Sync is for sellers who are already on Shopify and want to bring their eBay catalog over without duplicates.
  • →Voice: pick a preset (Direct, Premium, Punchy, Story-driven, etc.) or describe your own
  • →Audience: free-form description. "US-based resellers aged 25-55 who care about clean catalog hygiene."
  • →Listing cap: scope to a specific count (don't pay for more than you want)
  • →Exclude keywords: skip listings whose titles contain these words ("used," "broken," "parts only"). Filtered BEFORE any AI step so excluded items cost nothing.
  • →Markup % + rounding: apply a markup over the source price with charm-pricing ($X.99 / $X.95) or whole-dollar rounding
  • →Google Shopping fields: toggle on (default) to populate Variant Barcode, Brand, and MPN columns for multichannel syndication (Google Shopping, TikTok Shop catalog, Meta Commerce). Off if you only sell on Shopify proper and want a leaner CSV.

Try it on a 5-product preview, free.

See the actual output before you pay. Tweak the voice and regenerate up to five times before paying.

Generate a free preview →
How it works | Storeshift · Storeshift