Merchant Center catalog operations workshop / 5 MIN READ
Count the packs the retailer combines rather than every piece inside them
Create a pack-level representation record that separates retail grouping, piece count and offer identity.
AI-assisted research and analysis for the SNS Creative Fieldbook
Sources checked 2026-11-07
A pack contains several possible counts: manufacturer units, retailer-combined units and individual pieces. The feed needs the count with the meaning defined for each attribute.
What you will leave with: Create a pack-level representation record that separates retail grouping, piece count and offer identity.
Use this when
- A retailer combines identical products into a new offer.
- A supplier changes between individually packaged items and factory multipacks.
Not warehouse bundle assembly guidance or a rule that every multi-item package needs a multipack value.
Identify who created the sold grouping
Google defines multipack for identical manufacturer-defined products combined by the retailer. Its documentation distinguishes that count from pieces already contained inside a manufacturer-created pack. A single factory pack is not treated as a retailer-created multipack merely because it contains several pieces.
Start with the supplier's saleable package and the retailer's actual offer. Record whether the merchant resells one factory unit unchanged or combines several such units. Photograph or reference the packaging evidence where permitted, and identify the product master record that describes it. The word pack in a title is not enough to determine the attribute. It can refer to either layer, so the export rule must use structured evidence.
Keep three counts separate in the data model
Maintain fields for the number of manufacturer units combined, pieces within each manufacturer unit and total pieces delivered. Calculate the total only after the two inputs are verified. Do not overwrite the base package definition with the consumer-facing total, because other mappings may still need the original unit.
Add checks for mixed assortments and individually selectable items. A retailer-combined set of different products is not the same classification problem as repeated identical units. Likewise, a basket quantity selected by a customer is not automatically a fixed packaged offer in the catalog. If the business changes the offer from one factory pack to several, review its identity, title, price and product page together instead of merely changing a count field.
Reconcile the complete offer representation
For each affected item, compare the feed price, title, image, landing page and checkout quantity. A correct multipack value does not help when the page initially shows one unit and only reaches the advertised configuration after extra selections. Confirm that the customer can identify exactly what the submitted price buys.
Review identifier requirements for the particular merchant-created grouping using current documentation; do not invent a new GTIN to describe the arithmetic total. Preserve the source package identifiers and the approved item mapping in the evidence record. Test exports for numeric formatting and missing values, but remember that a valid integer can still describe the wrong packaging layer. Semantic review comes before format acceptance.
Use packaging changes as a catalog review trigger
When a supplier changes a factory pack from four pieces to six, the merchant's number of combined packs might stay the same while total content changes. Have purchasing notify catalog operations of such changes. A reused title or unchanged supplier SKU should not conceal a material difference in what buyers receive.
Build regression cases for a single item, one factory multipack and a retailer combination of factory multipacks. Compare the expected values with the processed result and the customer page after release. Include one mixed-product assortment that should not inherit the same rule. The durable outcome is a representation that can explain each count, not simply an export with no empty quantity fields.
Make the reasoning concrete.
A fictional cleaning retailer sells two factory cartons together. Each carton contains six identical cloths. Its spreadsheet has one quantity column containing 12, with no record of the factory packaging level.
Export the total piece count
Opening: Set multipack from the only quantity field.
Hypothesis: The number matches the pieces the customer receives.
Risk: It describes the wrong grouping layer for the attribute.
Record both packaging levels
Opening: Keep two combined factory units and six pieces per unit, with 12 total pieces.
Hypothesis: Each attribute can draw from the count it actually means.
Risk: Supplier packaging changes need an explicit review trigger.
The owner verifies the factory package, maps the retailer-combined count and checks the entire two-carton offer. The total piece count remains available for truthful customer information.
Take the questions into your next brief.
Use these prompts in your own notes. Nothing is submitted from this page.
Creator
Did the manufacturer or retailer create this sold grouping?
Base unit
What exactly is one manufacturer-defined product here?
Counts
How many base units and total pieces does the customer receive?
Offer
Do price, page and image describe the same grouping?
Change
Who reports a supplier packaging change before the next export?
Watch for the wrong conclusion.
- Using total inner pieces as the retailer multipack count.
- Applying the same rule to identical units and mixed assortments.
- Inventing a GTIN for a spreadsheet-created package total.
Your next action
Audit one nested pack and write its manufacturer-unit count, inner-piece count and total separately.
Sources & scope
Workflows, worksheets and fictional examples are original editorial analysis, not Google instructions or measured business results. Sources support only the specifically attributed platform facts. Platform controls and eligibility can change; verify current account guidance before implementation.
- Google Merchant Center: Multipack →Primary platform documentation · Supports: Supports the retailer-defined multipack meaning, its distinction from manufacturer packs and the correct packaging level to count. Data-model and regression controls are original editorial analysis.