How to Structure Google Shopping Campaigns for Maximum ROAS

Most D2C brands treat Google Shopping as a set-and-forget channel. Upload the product feed, turn on Performance Max, let Google do its thing. The problem: Google will optimize for conversion volume — and conversion volume often means spending your budget on low-margin products, branded searches that would have converted organically, and queries that produce high click volume but poor purchase intent. Structure is how you take back control.

Standard Shopping vs. Performance Max: when to use each

Google has been aggressively pushing Performance Max (PMax) as the future of Shopping campaigns. And for many brands, PMax does perform well — it pools signals across Search, Shopping, Display, YouTube, Gmail, and Maps. But it comes with a fundamental trade-off: you give up control and transparency in exchange for automation.

Use Performance Max when:

• You have a large, diverse product catalog (500+ SKUs) where manual campaign management would be impractical
• You have strong conversion history (50+ purchases/month) for the algorithm to optimize against
• You want Google to handle cross-channel placement automatically
• Your goal is volume and you can accept some inefficiency in the mix

Use Standard Shopping when:

• You have a small, high-margin catalog where every SKU matters
• You need granular control over which products appear for which queries
• You want full visibility into search term reports and negative keyword management
• You're building a new account and want to understand your traffic before automating

For most D2C brands, the right answer is a hybrid: a PMax campaign for the broad catalog, plus Standard Shopping campaigns for your top-margin products with manually controlled bidding. The Standard Shopping campaigns act as a safety net — they compete against PMax for your priority products and give you a window into actual search terms.

Campaign segmentation: organize by margin and product type, not by feel

The biggest structural mistake in Google Shopping is putting all products in one campaign and letting Google allocate budget however it wants. Google will find the path of least resistance — highest conversion rate, not highest margin. Your $15 product with a 90% conversion rate will eat budget from your $150 product with a 55% margin.

The framework we use segments campaigns along two dimensions: margin tier and product category. This allows you to set different target ROAS (tROAS) goals by segment, because your profitable math is completely different across segments.

Segment by margin tier

Create separate campaigns for high-margin products (40%+ gross margin), mid-margin (25–40%), and low-margin (<25%). Set tROAS targets accordingly — high-margin products can afford a lower ROAS target because each sale is more profitable.

Segment by product category

If you sell across multiple categories (e.g., apparel + accessories + home goods), segment by category. Bidding behavior and conversion rates differ significantly across categories, and a single campaign averages these out in ways that hurt performance.

Separate bestsellers

Your top 20% of products drive 80% of revenue. Give them their own campaign with more budget and more aggressive bidding. These products have the most conversion data — use it to your advantage.

Isolate new arrivals

New products need traffic to generate conversion data. Put them in a separate campaign with a Traffic or Maximize Clicks strategy initially, then migrate to tROAS once they have enough history.

Bidding strategy selection: the order of operations

Bidding strategy determines how Google spends your budget. Choose wrong and you'll either underspend (if your tROAS target is too high) or overspend inefficiently (if it's too low). Here's the progression we use for new campaigns:

| Phase | Strategy | Note | |-------|----------|------| | Phase 1 (0–30 days) | Maximize Clicks | Build impression share and conversion data. Set a max CPC cap to avoid runaway spend. | | Phase 2 (30–60 days, 30+ conversions) | Maximize Conversion Value | Transition when you have enough data. Let Google optimize for value without a ROAS constraint. | | Phase 3 (60+ days, 50+ conversions/month) | Target ROAS (tROAS) | Set your tROAS target 10-20% below your actual recent ROAS to give the algorithm room to operate. Tighten gradually. |

A common mistake: setting tROAS immediately on a new campaign. If you tell Google you want a 5x ROAS with no conversion history, it will either not spend your budget (too constrained) or quickly overspend on high-volume, low-value products to hit the number. Let the algorithm learn before you constrain it.

Negative keywords and query sculpting

In Standard Shopping, you can't add keywords — but you can add negative keywords to exclude irrelevant traffic. This is called query sculpting: shaping which queries trigger your ads by systematically excluding the bad ones.

Pull your search terms report weekly. Look for patterns in non-converting queries:

• Queries with informational intent ("how to clean X", "what is X") — exclude these
• Competitor brand names if you're not running a conquesting strategy — exclude or separate into a low-bid campaign
• Irrelevant product categories triggered by broad product titles in your feed
• Queries with high click / zero purchase history over 30+ days — exclude after confirming the pattern

For Performance Max, query sculpting is harder — Google doesn't show full search term data. Use campaign-level negative keywords (available at the account level) and URL exclusions to at least exclude categories Google shouldn't touch.

Feed optimization: the foundation everything else is built on

Your product feed is your targeting. In Shopping, you don't choose keywords — Google matches your products to search queries based on your feed attributes. A weak feed means poor query matching, higher CPCs, and lower impression share. A strong feed is a structural advantage.

Product titles This is the most important feed attribute. Include: primary keyword + brand + key differentiator + size/color where relevant. Wrong: "Blue Sneaker". Right: "Nike Air Max 270 Men's Running Shoe — Navy Blue, Size 10". Lead with the search term, not your brand name.

Product descriptions Stuff with relevant search terms and feature keywords — not marketing copy. Google indexes descriptions for query matching. Use this field to capture long-tail queries your title can't cover.

Google product category Be as specific as possible. Wrong: "Apparel & Accessories". Right: "Apparel & Accessories > Clothing > Activewear > Yoga Pants". More specific categories mean better query matching and eligibility for category-specific promotions.

Custom labels Use custom labels to segment your feed by margin tier, bestseller status, or seasonality. This is how you implement your campaign segmentation strategy — custom labels in the feed map directly to campaign product filters.

Putting it all together: the campaign structure blueprint

Here's the exact structure we implement for a mid-scale D2C brand ($50k–$200k/month in Shopping spend):

| Campaign | Name | Detail | |----------|------|--------| | Campaign 1 | Top Sellers — Standard Shopping — tROAS 400% | Top 20 products by revenue. Manual CPC or tROAS. Full negative keyword management. | | Campaign 2 | High Margin — Standard Shopping — tROAS 300% | All products above 40% gross margin, excluding top sellers. Lower tROAS target = more volume. | | Campaign 3 | Catalog — Performance Max — tROAS 250% | Full catalog minus products in Campaigns 1 & 2. Let PMax handle the long tail. | | Campaign 4 | New Arrivals — Standard Shopping — Maximize Clicks | All products added in last 60 days. No ROAS constraint. Rotate out when they have history. |

This structure gives you control where it matters most (top sellers and high-margin products) while using automation efficiently for the rest of the catalog. Revisit the structure quarterly — as products move up or down in performance, migrate them between campaigns accordingly.