Too Many Variants Are Killing Your Google Shopping Performance

The Variant Explosion

A single product with 5 sizes, 8 colours, and 2 materials creates 80 SKUs. Multiply that across 200 base products and you have 16,000 feed entries competing for budget. The problem isn't the product range - it's that Google Shopping treats each variant as a separate listing that needs its own data.

When each variant gets 30 impressions per month, the algorithm can't distinguish between a high-converting hero colour and a dog that never sells. Everything gets averaged, and your feed becomes a liability instead of an asset.

Impact on the Algorithm

Smart Bidding needs data density to work. When data is spread across thousands of low-volume variants:

• • Learning is impossibly slow: Each variant needs clicks and conversions to inform bids. With 12,000 variants, most will never accumulate enough data.
• • Budget allocation breaks: The algorithm can't efficiently allocate budget across 12,000 items. It defaults to showing whichever variant it stumbles into early wins with.
• • Quality Score suffers: Low-CTR variants (niche sizes, unpopular colours) drag down product group quality signals.
• • Impression share collapses: Your own variants compete against each other for the same search query.

Data Dilution Problem

Think of data like water. You need a minimum depth in each pool for the algorithm to swim. With 200 products and £50k/month spend, each product gets roughly £250/month in budget - enough for 100-200 clicks per product. That's workable.

But with 12,000 variants, each gets £4/month - roughly 2 clicks. The algorithm literally cannot learn anything from 2 clicks per month. Your entire feed is in permanent learning mode.

Feed Strategy for Variants

The fix is strategic variant management:

• • Hero variants only: Submit your top 3-5 selling colours/sizes as individual listings. Use item_group_id so Google knows they're related.
• • Suppress dead variants: Any variant with zero sales in 90 days should be excluded from the feed
• • Size consolidation: For apparel, submit the most popular size and let Google show other sizes on the product page
• • Colour champions: Identify your hero colours by conversion rate and give them priority in the feed
• • Seasonal rotation: Rotate which variants are in the feed based on seasonal demand

Consolidation Framework

Apply this decision tree to every variant in your feed:

• • Sales in last 90 days > 5: Keep in feed as individual listing
• • Sales in last 90 days 1-5: Keep but monitor - candidate for consolidation next quarter
• • Zero sales in 90 days: Remove from Shopping feed (still available on site)
• • Out of stock: Remove immediately (wasted impressions and poor user experience)

This typically reduces feed size by 40-60% while concentrating data on products that actually sell. The result: faster learning, better bids, higher ROAS.

Category Differences

Variant management differs by vertical:

• • Fashion: Size is generally not a Shopping differentiator (users don't search by size). Colour matters more. See fashion size/colour complexity.
• • Beauty: Shade variants are critical - users search for specific shades. See shade variant complexity.
• • Electronics: Storage/RAM/colour variants may target different search intents. Keep separate.
• • Home/furniture: Material and finish variants often serve the same search intent. Consolidate.

Next Steps