---
title: "Prepare your product catalog"
description: "Prepare a CSV or Excel product catalog for Interlinked with required fields, recommended attributes, clean descriptions, and business-specific examples."
section: "Products"
url: https://interlinked-ai.com/en/resources/docs/prepare-your-product-catalog
lang: en
lastUpdated: 2026-04-24
reviewedAt: 2026-04-24
---

# Prepare your product catalog

## Direct answer

Prepare your catalog as a CSV or Excel spreadsheet with one product per row. Include product name, price, and stock for every product. Add SKU when available, and include descriptions or attributes that help the AI agent understand what each product is and when to recommend it.

## Required fields

At minimum, include:

| Field | Why it matters |
|-------|----------------|
| Product name | Identifies the item the customer is asking about |
| Price | Lets the AI answer pricing questions and compare options |
| Stock | Helps the AI understand whether an item is available |
| SKU | Recommended for internal tracking; Interlinked can generate one if missing |

If stock cannot be identified for a product, Interlinked may mark the value as `0`. Treat this as a safety fallback. It is better to provide accurate stock values in the file and review them after import.

## Recommended extra columns

Extra columns are allowed and recommended. The more complete your catalog is, the better your AI agent can answer product questions. A short description, product category, size, color, flavor, material, or use-case column can help the agent understand what each product is and when it should recommend it.

Useful extra columns include:

- Description
- Color
- Size
- Flavor
- Material
- Brand
- Category
- Tags
- Use cases
- Ingredients
- Dosage or presentation, where relevant
- Compatibility
- Gender or age group, where relevant
- Care instructions
- Shipping notes
- Return restrictions
- Product benefits
- Product limitations

<div class="docs-callout docs-callout-tip">
  <p><strong>Tip</strong></p>
  <p>A description column is highly recommended because the AI agent uses product information to answer customers more naturally and accurately.</p>
</div>

## Examples by business type

Different businesses benefit from different attributes:

- A clothing store can include size, color, material, fit, and care instructions.
- A supplements store can include flavor, presentation, serving size, ingredients, and usage notes.
- A cosmetics store can include skin type, finish, shade, volume, and application notes.
- A hardware store can include material, measurements, compatibility, and recommended use.

Products must comply with applicable laws and Interlinked's usage policies. Do not upload illegal, restricted, or prohibited products.

## Clean data guidelines

Before uploading, review the file for consistency:

- Use one row per product or variant.
- Keep product names clear and specific.
- Use consistent currency and price formatting.
- Avoid blank names, blank prices, and blank stock values.
- Use consistent category names.
- Keep descriptions factual and customer-safe.
- Remove duplicate products unless they represent distinct variants.
- Keep a backup copy of the catalog file before replacing the live catalog.

## How catalog detail affects AI answers

If a catalog only includes product names, prices, and stock, the AI can answer basic questions. If it also includes product descriptions, categories, sizes, materials, use cases, or restrictions, the AI can respond with more useful context.

For example, if a customer asks for "a black formal shirt in medium," the AI can search across product name, color, category, size, and description when those fields are present. Without those fields, the AI has less context to identify the best match.

## What to do next

After preparing the file, upload it from the Products page or wizard. See [Upload a product catalog](/en/resources/docs/upload-a-product-catalog). For ongoing quality recommendations, see [Product catalog best practices](/en/resources/docs/product-catalog-best-practices).