How should product variants be structured in a database?

Parent-child model: one parent product record holds shared fields (title, brand, description), each variant child holds SKU-specific fields (color, size, price, stock, GTIN, image). This avoids duplicating shared data while allowing variant-level specificity.

How do you maintain data quality in a product database over time?

Validation at write time (not audit time): every product update is validated against a schema before it saves. Completeness scores visible in the catalog view. Version history that enables rollback when bad data is committed.

How does Seegea function as a product database layer?

Seegea maintains a versioned product database synchronized with your Shopify or PrestaShop store. It adds the validation, enrichment and governance layer that the native CMS does not provide.

Product database: structure and maintain a clean source of truth

A well-structured product database is the foundation on which every downstream operation runs: channel distribution, SEO, Google Shopping compliance, AI copy generation. A poorly structured one multiplies every downstream problem.

The product database schema: what to store and how

Core product fields

title (H1), short_description, long_description, meta_title, meta_description, handle/slug, status (active/draft/archived), created_at, updated_at, version_number, account_id.

Identifier fields

internal_sku (unique per account), gtin (EAN-13/UPC-12), mpn, brand, supplier_reference, hs_code (for cross-border). Each identifier serves a different system — keep all of them.

Attribute fields

color (name + hex), size (label + numeric equivalent), material (% breakdown), weight (value + unit), dimensions (W x D x H + unit), gender, age_group, condition. Stored as a flexible JSONB column to support per-category schemas.

Schema decision	Wrong approach	Right approach
Variant structure	Duplicate parent fields per variant	Parent-child: shared + specific
Attribute storage	Fixed columns per attribute	Flexible JSONB per product type
Version history	Overwrite in place	Append-only with version number
Image storage	URLs in free text	Structured array: url + alt + position
Price	Single price field	price + compare_at + currency + channel

Never store product data in a single "notes" or "description" field as a workaround for missing structured fields. You cannot query it, validate it or distribute it reliably. Add the structured field instead — even if it is just a metafield in Shopify.

Version history: the safety net you did not know you needed

Product databases degrade when bulk edits go wrong — a CSV import overwrites descriptions, an AI generation produces bad copy, a team member edits the wrong product. Without version history, these mistakes are permanent.

Seegea stores a full version of every product after each modification. Ctrl+Z restores the previous version and re-pushes to Shopify or PrestaShop in one operation. No manual recovery, no "can anyone remember what it said before?". Built in France between Annecy and Chantilly, with EU-hosted data and 99.9% SLA on Supabase PostgreSQL.

See the Seegea product database in action

30 min Google Meet · we review your catalog structure together

See the Seegea product database in action

Created in France (Annecy – Chantilly) · Email & Google Meet support

FAQ

Core product fields (title, description, SKU, GTIN, brand, price), attribute values (color, size, material), variant relationships (parent-child SKUs), media references (image URLs, alt text), channel-specific overrides and version history.

See Seegea in action

Book a 30-min live demo on Google Meet. No commitment.

Book a demo