Merge every source into one record One answer per SKU.

Import the catalog and read the quality report. Every product gets a score against the rules you set, and the field view shows completeness per attribute across the whole catalog. That gives you a ranked list of what is missing where, instead of an overall percentage nobody can act on.

That depends on the catalog and on how much of it needs a human decision. Import, mapping and the mechanical corrections of casing, whitespace, units and categories run in a single session per source. Merging duplicates and rewriting thin content are the parts that take real time, because someone has to approve them.

Anything that follows a rule: formatting, casing, whitespace, decimal separators, unit codes, category assignment, dictionary terms and empty fields it can derive from existing data or web research. Deciding that two near-identical products are the same product, or that a price is wrong, stays with you.

Not if the mapping is set up right. An import matches on the identifier you choose and updates the fields that source owns, leaving the rest alone. You control per field whether a source may overwrite it, and Quality Guard quarantines rows that break a rule instead of writing them.

The recurring work of imports, standardisation, validation and exports can run as scheduled automations. What stays human is judgement: approving generated copy, resolving a genuine conflict between two sources, and deciding what a good product looks like in the first place.