Product data automation services

What a product data standard covers

Agreeing the standard

Work through a sample of real products with the people who buy, list and sell them. Decide title order, which attributes are structured fields, which units to use and which tags are allowed. Write it down in a page or two, with examples. Then measure the current catalogue against it: how many products need fixing, and which fields are most often wrong. That shows the clean-up needed before automation can rely on the data.

Keep the standard short enough that people will follow it.

How the service runs

1. 01

   Standard

   Data rules agreed and written down.

   Output: Product data standard

2. 02

   Measure

   Catalogue checked against the standard.

   Output: Gap report

3. 03

   Map suppliers

   Each supplier file mapped with import checks.

   Output: Supplier mappings

4. 04

   Map channels

   Feeds mapped to each channel's specification.

   Output: Working feeds

5. 05

   Routine

   Daily reports and feed error handling set up.

   Output: Clean data kept clean

Mapping each supplier

For each supplier, record which columns map to which fields, how their units and categories translate, and what checks apply. Ask whether the supplier can provide data in a more consistent format, such as a structured file instead of a PDF price list. Document the mapping so it can be updated when the supplier changes their file, as suppliers sometimes do.

A routine for feed errors

Channels report rejected or limited listings with reasons. Review these on a set day each week, group them by cause and fix the product record or the mapping rule, then confirm the listing recovers. A rising error count after a channel changes its requirements is a prompt to update the mapping.

Example: a catalogue from many suppliers

What you should receive

Warning signs in a proposal

• Imports without a standard

  Suppliers' inconsistencies flow straight into the store.

• Fixes made in the feed only

  The store record stays wrong and errors return.

• AI descriptions published unreviewed

  Claims nobody has checked reach customers.

• No documentation of mappings

  Nobody can update them when a supplier changes.

Cleaning the existing catalogue

Before automation takes over, the existing catalogue may need bringing up to the standard. Prioritise best sellers and products listed on channels with strict requirements. Bulk tools can fix consistent problems, such as units in the wrong format; others, such as missing attributes, need checking against supplier specifications. Track progress against the gap report so the clean-up has a clear end.

Agree who maintains the standard as new product types arrive.

Training the people who list products

Show everyone who adds or edits products the standard, the completeness report and how to fix held import rows, so the data stays clean after the project ends.

When channels change requirements

Channels update their attribute requirements from time to time. Assign someone to watch for announcements from each channel you sell on and update the mapping before listings are affected.