Most ecommerce Google Ads accounts are bleeding money in the same five places. The product feed isn’t optimized. Performance Max is running with no audience signals and cannibalizing branded search. Target ROAS is set before the account has enough conversion data to use it. Negative keywords haven’t been touched in months. And nobody is reading the search term report.
If any of those sounds familiar, this guide is for you.
Google Ads in 2026 looks different from 2020. Performance Max has replaced Standard Shopping as Google’s default recommendation. Smart Bidding controls most budget allocation. The advertiser’s job has shifted from micromanaging bids and match types to feeding the machine better signals. But the fundamentals of profitable ecommerce advertising haven’t changed: you need the right products in front of the right people at a cost that leaves margin after the sale.
Here’s how to do it.
The Google Ads Landscape for Ecommerce in 2026
The major shift of the past three years is Performance Max. Google has aggressively pushed advertisers toward this campaign type, which runs across Search, Shopping, Display, YouTube, Gmail, and Maps simultaneously using a single asset group and Smart Bidding.
Most advertisers running Performance Max are running it wrong. Not because it’s a bad campaign type, but because Google’s setup wizard omits several configuration steps that are critical for ecommerce, and because the default behavior is to spend budget wherever Google thinks conversions are likely, which often means branded search terms and high-funnel display placements that provide the illusion of performance without the substance.
The opportunity for ecommerce advertisers who understand how Performance Max actually works is significant. Accounts that configure it deliberately outperform accounts running it on defaults by a wide margin.
But before getting into Performance Max mechanics, campaign type selection matters.
Campaign Types and When to Use Each
Performance Max. Best for ecommerce stores with product catalogs connected to Google Merchant Center. It’s the most powerful when configured correctly and when the account has sufficient conversion data (minimum 30-50 conversions per month, ideally 100+). It’s the worst when running without audience signals, without negative keywords, and without proper brand exclusion.
Standard Shopping. Still useful alongside Performance Max, particularly for high-value SKUs where you want explicit bidding control and transparent search term data. Standard Shopping gives you visibility into which search queries trigger your ads - data Performance Max obscures. Running both simultaneously lets you use Standard Shopping as a controlled test environment and Performance Max as the volume driver.
Search campaigns. Essential for branded keywords (protecting your brand from competitors bidding on your name), competitor campaigns (bidding on competitor brand terms with comparison messaging), and category campaigns targeting high-intent transactional terms where you want precise keyword control and ad copy flexibility.
Display and YouTube. Rarely the first priority for ecommerce. Use for remarketing (showing ads to people who visited your site but didn’t purchase) and for building brand awareness in competitive markets where top-of-funnel investment makes sense given LTV. Don’t run prospecting Display before Search and Shopping are performing.
The hierarchy for most ecommerce stores: get Search (branded) and Shopping (Performance Max + Standard Shopping) working first. Add remarketing Display once you have sufficient audience size. Add YouTube when you have video creative and a budget that can sustain the higher CPMs.
Product Feed Optimization: The Most Important Lever in Shopping
Before campaign structure, before bidding, before audience signals: fix the product feed. Your Google Merchant Center product data is what determines which queries your Shopping ads appear for. A poorly optimized feed means Google matches your products to irrelevant queries and misses the queries that actually convert.
The product feed fields that matter most for SEO-like matching:
Title. This is the single most important field. Google matches product titles against search queries the same way it matches organic titles against keywords. Include the primary keyword first. Use the format: [Brand] [Product Name] [Key Differentiator] [Category]. For a women’s running shoe: “Nike Women’s React Infinity Run Flyknit 3 Running Shoes.” Not “Women’s Shoe Nike React” - that’s how most default feeds are formatted.
Description. Secondary to title but still relevant for query matching and for providing context to Smart Bidding. Include secondary keywords naturally. Aim for 500-1,000 characters for complex products.
Product type. Use the full taxonomy path: Apparel > Shoes > Running Shoes > Women's Running Shoes. Don’t truncate. This data helps Google categorize your product correctly.
Custom labels. These are five fields (custom_label_0 through custom_label_4) that you define for your own segmentation purposes. Use them to flag high-margin products, seasonal items, clearance inventory, new arrivals, and bestsellers. Custom labels let you create separate campaign strategies by product profitability rather than treating your entire catalog as one uniform asset.
GTIN (Global Trade Item Number). Include GTINs for branded products that have them. Google uses GTINs to match your products to search queries more precisely and to enable seller ratings and Shopping annotations. Missing GTINs on branded products leaves performance on the table.
The most common feed quality issue: automatically generated titles pulled from the product name field without modification. Product names as written in a CMS rarely match how people search. Rewrite titles to match search queries, not product names.
Performance Max Campaign Structure for Ecommerce
The default Performance Max setup recommended by Google’s interface: one campaign, one asset group, all products. Don’t do this.
A well-structured Performance Max campaign for a mid-size ecommerce store:
Separate campaigns by budget priority, not product type. Run a “Core Catalog” Performance Max campaign with 60-70% of Shopping budget, covering your full product range. Run a “Best Sellers” Performance Max campaign with 20-30% of budget, targeting only your top-converting products (use a custom label to identify them). Run a “New Arrivals” Performance Max campaign with the remainder, covering products launched in the last 60 days.
This structure lets you set different ROAS targets by business priority. Your bestsellers can tolerate a tighter ROAS target. New arrivals need more impressions and looser ROAS to build data before optimizing.
Asset groups by product category and creative angle. Within each campaign, create separate asset groups for distinct product categories. A furniture store might have asset groups for “Sofas,” “Dining Tables,” and “Bedroom Furniture.” Each asset group gets relevant images, headlines, and descriptions for that category. This ensures the AI isn’t mixing sofa creative with bedroom furniture creative.
Audience signals. This is where most accounts are leaving money. Audience signals guide the AI toward the types of users most likely to convert, without hard-restricting targeting. Provide: your customer email list (Matched Audiences), website visitors who completed a purchase, high-intent Google audiences relevant to your category, and visitors from your own remarketing lists. Better signals produce faster and more accurate Smart Bidding optimization.
Search themes. Add 3-7 search theme keywords per asset group. These tell Performance Max which search queries are relevant for each asset group, improving coverage on specific terms that matter to your business. Don’t use them as a substitute for a separate Search campaign - use them as a signal layer.
Bidding Strategy: When to Use What
Smart Bidding in Google Ads optimizes bids automatically based on conversion signals. The question isn’t whether to use Smart Bidding - you should - it’s which strategy to use given your account’s maturity.
Maximize Conversion Value (no target). Use this when launching a new campaign or when you have fewer than 30 conversions per month. The algorithm explores broadly to find conversions without a ROAS constraint. Accept lower efficiency initially in exchange for building the conversion data the algorithm needs.
Target ROAS. Use this once you have 30-50 conversions per month per campaign, ideally 100+. Set your initial target ROAS at 20-30% below your current actual ROAS to give the algorithm room to find volume. Gradually tighten the target as the algorithm demonstrates consistent performance. Jumping straight to a high ROAS target with insufficient data is the most common Smart Bidding mistake.
Target CPA. Less common in ecommerce (where order values vary), but useful for stores with relatively uniform AOV where you want to cap cost per acquisition directly.
Enhanced CPC. Legacy strategy. Not recommended for new campaigns. If you’re running it on existing campaigns, test a transition to Target ROAS.
One important note on ROAS targets: they are calculated on revenue, not profit. A 5x ROAS on a product with 20% gross margin is break-even at best once you account for ad costs. Know your margins by product category and set ROAS targets accordingly.
Negative Keywords: The Most Neglected Lever
In a Performance Max world where Search campaigns handle less budget, negative keywords are still critical for two reasons: they apply across Standard Shopping campaigns, and they can be added as account-level negative keyword lists that apply across campaign types.
The negative keyword categories every ecommerce account needs:
Irrelevant job and career terms. Searches like “[product category] jobs,” “[brand] careers,” “[product] salary” consume budget on zero-intent queries. Block all job-related terms at the account level.
Informational modifiers with no commercial intent. Terms like “what is,” “how does,” “history of,” “definition of” preceding your product categories bring researchers, not buyers. Add as broad match negatives.
Competitor brand terms (if not running competitor campaigns). If you’re not intentionally bidding on competitor brand terms with dedicated campaigns and relevant landing pages, add competitors as negatives to prevent accidental spend.
Product types you don’t sell. A store selling women’s clothing should negative-match men’s and children’s product categories at the account level to avoid Shopping mismatches.
Free, DIY, and budget variants (if you’re premium). A luxury watch brand should negative-match “cheap,” “affordable,” “replica,” “knockoff” terms.
The search term report in Standard Shopping campaigns is your primary source of negative keyword intelligence. Pull it weekly for new accounts, monthly for mature accounts. Sort by cost, identify non-converting queries with spend above your target CPO, and add them as negatives.
Search Campaign Structure for Ecommerce
Most ecommerce operations need at least two Search campaigns:
Branded campaign. Targets searches for your brand name and variations. Bidding on your own brand might seem counterintuitive, but it serves two purposes: it protects your top search position from competitors who bid on your brand (which they can legally do), and it provides data on your branded search volume that’s separate from organic traffic. Keep branded campaign budget separate from category campaigns so branded performance doesn’t inflate category ROAS calculations.
Category campaigns. Targets high-intent transactional terms for your core product categories. Use phrase match as the primary match type. Build ad groups around tight keyword themes, not broad categories. “Women’s leather ankle boots” and “ankle boots for women” can share an ad group; “boots” as a standalone term is too broad to control cost-effectively without extensive negative keyword coverage.
Competitor campaigns (bidding on competitor brand terms) are optional but effective when you have a genuine comparison advantage. If you’re launching one, your landing page must directly address the comparison - don’t send competitor traffic to your homepage.
The SEO and Google Ads Feedback Loop
Running both SEO and Google Ads creates data advantages that neither channel provides alone.
Google Ads reveals which queries convert, not just which queries generate impressions. When a search term consistently converts in paid search, it’s a candidate for organic content investment. For an ecommerce client in the luxury vertical, paid search data identified a cluster of terms around a specific product attribute that had strong conversion rates but minimal organic presence. Building collection pages targeting those terms generated $606K in organic revenue over 12 months without the ongoing ad spend.
In reverse, organic ranking data informs which paid keywords have high impression share but your site doesn’t rank organically for. These are high-priority paid terms because you’re entirely dependent on paid search for visibility there. If you turn off those campaigns, the traffic disappears.
Share keyword data bidirectionally between your SEO and paid teams. The ecommerce SEO service and Google Ads management work best when both channels are informed by shared keyword intelligence.
Attribution in 2026: Reading It Without Being Misled
GA4’s default attribution model is data-driven, which distributes conversion credit across multiple touchpoints based on Google’s proprietary model. This is more accurate than last-click but creates reporting challenges.
Three things to understand about attribution in 2026:
Google Ads and GA4 will never agree. Google Ads uses its own attribution model for reporting within the platform. GA4 uses a different model. Discrepancies of 20-40% are normal and don’t indicate a tracking error. Decide which data source is your source of truth for optimization decisions and stick to it.
View-through conversions inflate Performance Max numbers. Performance Max attributes conversions that happen after an ad impression even if the user didn’t click. This can make Performance Max appear significantly more efficient than it is. In your Performance Max reports, filter to “click-through conversions only” for a more realistic performance picture.
Import offline conversions where possible. For stores with both online and offline sales, importing offline conversions (point of sale, phone orders) into Google Ads gives Smart Bidding a more complete picture of the value each click generates. This is especially relevant for high-ticket items where a significant portion of the sales cycle happens offline after initial search research.
How to Audit an Underperforming Google Ads Account
When I take over a new ecommerce Google Ads account, these are the first five things I check:
-
Conversion tracking. Is it firing correctly? What’s being counted as a conversion (purchases only, or also add-to-carts and form fills)? If the account is optimizing for add-to-cart instead of purchase, Smart Bidding is optimizing for the wrong event.
-
Search term reports. What percentage of spend over the last 90 days went to irrelevant queries? High-waste accounts often have 30-40% of spend on non-converting terms with no negative keyword coverage.
-
Performance Max configuration. Are brand exclusions in place? Are audience signals populated? Are asset groups organized by product category or is everything in one asset group?
-
Bidding strategy vs. conversion data. Is the account using Target ROAS with fewer than 30 conversions per month? If yes, Smart Bidding is operating on insufficient data and either underdelivering or overspending.
-
Product feed quality. Pull a sample of 20 products from Merchant Center. Read the titles as a consumer searching for those products. If the titles don’t match natural search language, feed optimization is the highest-priority fix.
For a detailed account diagnostic framework, see the Google Ads account audit guide. If you’d prefer a professional audit and management service, the Google Ads management service covers account setup, feed optimization, and ongoing management.
For businesses considering whether to prioritize Google Ads or Meta Ads, the comparison and budget allocation framework is covered in the Google Ads vs Meta Ads guide.
Useful References
FAQ
What budget do I need to start Google Ads for an ecommerce store? The minimum viable budget depends on your product price point and target ROAS. For products with an average order value of $50-100, a budget of $1,500-3,000 per month is the minimum to generate enough conversion data for Smart Bidding to function. For high-ticket products ($500+), you can start with $1,000/month because each conversion carries more value. Below these thresholds, the algorithm doesn’t have enough data to optimize, and you’re essentially paying for learning without generating enough signal to act on.
Should I run Performance Max or Standard Shopping for my Shopify store? Both. Start with Standard Shopping to build conversion data and gain visibility into which search terms convert. Once you have 30+ conversions per month, introduce Performance Max with proper configuration (audience signals, brand exclusions, organized asset groups). Run Standard Shopping alongside Performance Max to maintain search term visibility and as a control for high-value product segments. Don’t turn off Standard Shopping entirely in favor of Performance Max.
How do I stop Performance Max from cannibalizing my branded search? Add your brand name and all brand variations as brand exclusions in the Performance Max campaign settings. This prevents Performance Max from bidding on branded queries, which should instead be handled by your dedicated branded Search campaign. Without brand exclusion, Performance Max will often absorb branded conversions (which are easy to get) and report inflated ROAS numbers that don’t reflect incremental paid performance.
What ROAS should I target with Google Ads for ecommerce? It depends entirely on your gross margin. If your products have 40% gross margin, a 2.5x ROAS (250%) is break-even on ad spend before accounting for COGS. Anything above that is profitable. Most ecommerce businesses target 4x-8x ROAS, but this is only meaningful in the context of your margin. Calculate your break-even ROAS (1 divided by gross margin percentage) and set your target above that threshold.
Why is my Google Shopping cost-per-click so high? High CPCs in Shopping usually indicate one of three things: heavy competition for the category (check auction insights), a low Quality Score caused by poor landing page experience or low CTR on your product listings, or a Target ROAS set so high that Google bids aggressively on every query to hit the target. Audit your product feed titles (better titles improve CTR and lower effective CPC), review your landing pages for relevance to the search query, and test loosening your ROAS target to see if CPC normalizes.
How long does it take for Google Ads to start working for ecommerce? Realistic timeline: 4-8 weeks to exit the Smart Bidding learning phase and begin seeing stable performance. The first 2-4 weeks are learning phase, where the algorithm explores and performance may be erratic. Weeks 5-8 typically see performance stabilize. Meaningful optimization decisions (raising ROAS targets, expanding budget, adding campaign types) should wait until week 8-12 when you have sufficient data. Accounts that make major changes every week during the learning phase never exit it properly.
About the Author Luciano Bonanno is an independent SEO and Growth Consultant with 18 years of experience. Founder of SameAPI and DeLeak.co. Book a strategy call →