Skip to content

Content Engine

The merchantCENTRAL Content Engine is the SEO test bench for your listings. It audits the search-engine facing texts of every marketplace listing — title, meta title, meta description, description, URL slug, images and identity data — against rules you set, and writes a finding per breach with a severity and a reason. In phase two, Copilot drafts new texts from the facts of the listing; none of them reaches a channel before a person has applied it.

The SEO fields live channel-neutrally in the hub on the marketplace item. Every connector maps them to its own channel fields on export (Shopware 6: metaTitle, metaDescription, keywords, seoPathInfo). The Content Engine knows no connector — it reads and writes hub fields only.

What does the Content Engine do?

Feature Description
SEO audit Every listing against a rule set of 23 rules: missing, too short, too long, duplicate, invalid — with a severity and one sentence on what search engines make of it
Limits are settings All lengths live in the setup. What search engines display shifts over time; a limit of 0 switches that check off
Duplicate detection Title, meta title, meta description and slug per marketplace and sub-account — Duplicate Slug is an error, because only one of the two listings is reachable under that path
Judged as the channel serves it The description through the hub's description resolver, the images from the channel's and the item's pictures — a listing filled by Icecat does not count as empty
Dashboard Errors, warnings and hints of the last audit, the proposals waiting, and the reason if nothing is being audited right now
Findings list The daily work: errors first, one jump per row to the marketplace item and to the item card
Copilot drafts Title, meta title, meta description, keywords, slug and an HTML description from the facts of the listing — never from the model's world knowledge
Review before the channel Every draft becomes a content proposal with four states. Only Apply writes the hub fields, flags the listing for the next item update and audits it again immediately
Daily run One job queue entry audits the catalogue once a day; the audit on demand is on the dashboard, in the marketplace item list and on every item card

The audit needs no AI and no external service

The audit is pure rule evaluation over hub tables: no credentials, no HTTP call, no tokens. Only the optional Copilot drafts call Azure OpenAI, through the hub's shared AI Setup. You can use the module completely without ever entering an AI resource.

Quick start in 5 steps

  1. Set up the hub — the Content Engine requires a configured merchantCENTRAL hub from version 2.106 on (Initial Setup)
  2. Activate the license — module CONTENTENGINE; a demo can be started once per tenant and module. Without a license nothing is audited, while the setup stays possible
  3. Check the limits — the defaults on the setup card match current search result pages; change only what you know better
  4. First audit — start it from the dashboard and read the findings: errors first
  5. Automate — Create Job Queue Entry schedules the daily run; from then on the dashboard shows where you stand

How this differs from Icecat

Both modules hang on the same description resolver, but they do different things:

merchantCENTRAL for Icecat Content Engine
Role A source Judgement and generation
Where the content comes from From the Icecat database, wherever brand and GTIN match From the facts of your own item — or from you
What it produces Marketing texts, images, data sheets, features Findings and text drafts for the search engine
Where it helps Wherever a manufacturer releases content Wherever no source applies: long tail, own brands, B2B wording — and anywhere the existing text is poor for a search engine

In short: Icecat fills, the Content Engine judges the result and writes where nothing was filled. A listing whose description comes from Icecat counts as present for the audit — what is judged is what the channel serves, not what the marketplace item stores.

Topics in detail