Summarize this article with:
Almost every ecommerce account I audit underperforms for one of the same eight reasons. Not a random mix of issues, though. Always the same eight. Sometimes it’s broken conversion tracking feeding bad data to Smart Bidding. Other times it’s a Performance Max campaign consuming 40% of its budget on branded queries that would have converted organically. Or it’s a product feed built by a developer for Shopify’s export settings rather than Google’s matching algorithm.
The “best practices” articles ranking for this topic read like platform documentation. They list the same things in the same order with the same generic explanations. But what none of them give you is the sequencing: what to fix first, why the order matters, and what each fix typically produces in practice.
So here’s the version from someone who’s still in the accounts.
What google ads best practices actually mean for ecommerce
The term “best practices” gets applied to a lot of settings that don’t actually move performance. True google ads best practices for ecommerce are the account decisions that directly affect whether Smart Bidding gets accurate data, whether your products appear for the right queries, and whether your budget concentrates on profitable SKUs or gets diluted across everything.
Actually, three things determine ecommerce Google Ads performance at a structural level. First, signal quality: the accuracy of your conversion tracking, the quality of your product feed, and the accuracy of your attribution model. Since Smart Bidding optimizes for whatever signals it receives, wrong signals produce wrong decisions at scale. Second, campaign architecture: how products are segmented, whether brand traffic is isolated from non-brand, and whether Performance Max is structured to complement Standard Shopping or compete with it. Third, bidding calibration: whether Smart Bidding has enough monthly conversion data per campaign to optimize rather than restrict, and whether tROAS targets reflect actual account blended ROAS rather than aspirational numbers.
Everything else (ad copy, landing pages, keyword match types, image assets) sits on top of these three foundations. Optimizing the surface layer without the foundation is why so many accounts produce better dashboards than revenue.
Why most ecommerce accounts still underperform
In fact, the accounts I inherit almost always have one thing in common: the person managing them was following platform recommendations rather than account data. Google’s auto-apply recommendations, optimization score pressure, and account-level suggestions are designed to increase spend. They’re not designed to improve your unit economics. A lower optimization score doesn’t mean your account is broken, actually. It means you’ve declined some of what Google wants you to accept.
The most common structural failure is treating Performance Max as a replacement for everything. PMax is an excellent prospecting and reach tool. It’s a poor brand traffic manager, a weak search term reporter, and an unreliable hero product protector. Brands that put their entire catalog into one PMax campaign at a single tROAS target are treating a $400-margin item identically to a $15-margin item, allowing brand queries to overbid against organic traffic that was already converting, and losing the search term visibility that tells you which queries actually convert.
The second common failure: measuring performance on what Google tells you rather than what Shopify confirms. Google’s reported ROAS and your actual blended ROAS (total revenue from Shopify divided by total paid spend) diverge by 20 to 40% in most multi-channel accounts because of multi-day attribution windows and shared conversion credit. Brands making budget allocation decisions based on Google’s self-reported ROAS are working with systematically inflated data.
The 8 google ads best practices, ordered by impact
These are the eight practices I apply to every new ecommerce account, in this sequence. The order matters because each one builds on the previous. Fixing number six without fixing number one is like building a house on a bad foundation.
1. Conversion tracking accuracy before any campaign change. Before touching bids, budgets, or campaign structure, verify that the primary conversion action is a real purchase event with actual revenue value, that enhanced conversions is active (Google Ads Settings > Conversions > Enhanced Conversions), and that there’s no duplicate conversion counting across Google Ads and GA4. I find conversion issues on roughly 65% of inherited accounts. The fix is unglamorous, but it changes every downstream optimization decision the algorithm makes. Indeed, in accounts where enhanced conversions wasn’t running, enabling it alone has improved Smart Bidding efficiency by 15 to 25% within 30 days.
2. Feed quality as the primary campaign performance driver. Since 74 to 97% of Performance Max spend in ecommerce goes to Shopping placements, the product feed is effectively the account’s primary creative asset. Product titles are Google’s keyword matching mechanism for Shopping. A title like “Blue Running Shoe” matches almost nothing useful. A title like “Nike Air Zoom Pegasus 41 Men’s Road Running Shoes Blue White Size 10” matches dozens of high-intent specific queries. Feed title restructuring (Brand + Product Type + Key Attribute + Attribute 2) consistently produces 20 to 40% improvement in impression share on high-intent queries within 30 days. When I rebuilt titles for P-REX Hobby, that was the only change that mattered. Hustle Marketers’ GTIN guide for Google Shopping covers feed validation and attribute requirements that unlock premium placements.
3. Brand traffic isolation from Performance Max. Without a brand exclusion list in PMax (Settings > Brand exclusions > Search terms), PMax bids aggressively on branded queries. These convert at high rates, inflate apparent ROAS, and mask poor non-brand performance. The fix: add brand exclusions to every PMax campaign and create a separate branded Search campaign at tCPA $2 to $5. After making this change, I typically see non-brand ROAS drop 20 to 40% from the inflated baseline. Not because performance got worse, but because you’re finally seeing the real number. That’s the baseline all subsequent optimization improvements are measured against.
4. Campaign segmentation by margin tier. A single PMax campaign covering the full catalog at a uniform tROAS target treats a 60% margin product identically to a 15% margin product. Smart Bidding can’t distinguish between them unless you segment by custom label. The fix: add custom labels to the product feed (custom_label_0 through custom_label_4) for margin tier (high/medium/low) and performance tier (proven/testing/new). Create separate PMax asset groups or separate Standard Shopping campaigns by margin segment. Bid more aggressively on high-margin products and less on clearance or low-margin SKUs. For accounts with 100+ SKUs, this single change typically improves blended ROAS by 8 to 20% within 60 days.
5. Smart Bidding calibration in the right sequence. Launch new campaigns with Maximize Conversion Value and no tROAS target. Wait for 30 to 50 monthly conversions. Then transition to Target ROAS at 10 to 15% below the observed blended ROAS. Tighten in increments of 10 to 15% every 14 days. Never change tROAS by more than 15% at a time, because changes above that threshold trigger learning phase resets. For promotions and seasonal periods, use Seasonality Adjustments (Google Ads Tools > Bid adjustments > Seasonality adjustments) rather than changing the tROAS target mid-flight. In fact, the most common Smart Bidding mistake is setting tROAS too high too early, which starves the campaign of data and produces the self-defeating cycle of low conversion volume leading to more restrictions.
6. First-party data loaded and active. Customer Match uploads CRM email lists as Smart Bidding audience signals for both PMax and Standard Shopping. Enhanced conversions matches hashed purchase data against Google’s signed-in user graph, improving the conversion model without spending anything additional. Together, these two features are the highest-value free tools in Google Ads, and most accounts never enable them. For ThePetsClub UAE, loading an 18,000-person Customer Match list improved non-brand conversion rate by 22% within 30 days, before any campaign structure change.
7. Search term analysis and negative keyword discipline. For Standard Shopping campaigns, review the search terms report every week. Add negatives aggressively: competitor brand queries, off-category terms, low-intent patterns (e.g., “how to,” “free,” “DIY”). For PMax, also review the search themes insight report monthly and add off-target terms to the account-level negative keyword list (Google Ads Settings > Negative keywords). PMax doesn’t support campaign-level negative keywords for search terms, but account-level negatives block them across all campaigns. In accounts where I run weekly negative keyword reviews, wasted spend on off-intent queries typically drops 15 to 25% within 60 days.
8. Blended ROAS measurement rather than platform ROAS. Google’s self-reported ROAS includes multi-day attribution windows that share credit with Meta, email, and organic. The honest metric is total revenue from Shopify or BigCommerce divided by total paid spend across all channels. Track this weekly. If the gap between Google’s reported ROAS and your blended ROAS is above 30%, you’ve been making budget allocation decisions on inflated data. Setting tROAS targets based on actual margin rather than benchmarks is the key step most accounts skip.
Performance Max vs Standard Shopping: the actual decision
The debate between PMax and Standard Shopping gets covered in every google ads best practices article. Most of them say “use both” without explaining the specific role each should play.
First, Standard Shopping belongs on your top 20% of SKUs by revenue and margin: the hero products where you want full search term visibility, product-group-level bid control, and negative keyword management. Standard Shopping shows you exactly which queries triggered which products. That visibility is the primary management tool for high-value SKUs.
Then, Performance Max belongs on the broader catalog: the remaining 80% of SKUs where scale and multi-channel reach matter more than granular control. But PMax needs to be structured properly: separate asset groups by product category, brand exclusions loaded from day one, Customer Match as an audience signal, and tROAS calibrated per asset group rather than across the whole campaign.
The accounts that run everything in one PMax campaign at a single tROAS target aren’t running google ads best practices. They’re running a template.
The auto-apply setting that silently degrades accounts
Google’s auto-apply recommendations automatically accept changes to your account (bid adjustments, keyword additions, match type changes, budget increases) based on Google’s optimization score. Auto-apply is on by default in most accounts and most account managers never touch it.
Turn it off. Every single one. Go to Tools > Account-level recommendations settings > Auto-apply recommendations and disable every checkbox. Google’s optimization score measures how closely your account mirrors what Google wants, not how effectively it serves your business goals. A lower optimization score with better ROAS is the correct trade-off. Review recommendations manually every two weeks. Accept what makes analytical sense. Decline the rest. This is one of the highest-ROI google ads best practices changes available and takes 10 minutes to implement.
What these practices produced for named clients
ArmorGarage, BigCommerce, garage floor coatings. Inherited account: one PMax campaign for the full catalog, no brand exclusions, 60% of SKUs missing GTINs, auto-apply recommendations on. Applied practices 1 through 5 in sequence across 90 days. Started with tracking validation (enhanced conversions wasn’t running). Then GTIN addition, feed title rebuild, brand exclusions, Standard Shopping for hero products, PMax rebuild by product-line asset groups. Result: 1,500%+ ROAS within 90 days. The ArmorGarage case study documents the full sequence.
P-REX Hobby, Shopify, hobby parts for Bin Chen. Tracking was clean. The problem was practice number two: feed quality. Generic titles that didn’t match model-specific search terms. Rebuilding titles to front-load brand, part type, and compatibility model number produced 20 to 40% impression share improvement on high-intent queries within two weeks. ROAS hit 9x over 90 days. The P-REX Hobby case study documents the title restructure process.
ThePetsClub UAE, Shopify Plus, pet food. Tracking was fine. Architecture was reasonable. The missing piece was practice number six: first-party data. An 18,000-person CRM list had never been loaded as Customer Match. PMax was running with broad signals only. Loading Customer Match improved non-brand conversion rate 22% in 30 days. Combined with asset group restructuring and a Meta budget reallocation toward Google Shopping, ROAS reached 14x over 90 days.
The pattern across all three accounts: the highest-impact change in each was a different practice. That’s actually the key insight. Auditing which practice is broken first is the actual skill.
What I’d check in the first 30 minutes of any ecommerce audit
First, pull the conversion actions list. Is the primary conversion action a real purchase event? Enhanced conversions should be active. Is there a duplicate from GA4 also counting as a primary action?
Then check Merchant Center. How many products are disapproved? What’s the GTIN coverage percentage? How old is the most recent feed fetch? Disapproved products have zero Shopping impression eligibility.
Then open PMax search terms insight. What percentage of spend is going to branded queries? Anything above 20% without a separate branded Search campaign is overbidding on traffic that didn’t need a PMax bid.
Then pull 20 product titles from top-revenue SKUs. Do they match actual queries in the Shopping search terms report? If the gap is large, feed work will move ROAS faster than any campaign setting.
Finally, check auto-apply settings. If recommendations are auto-applying, you don’t actually control the account.
What implementing these practices costs
Tracking audit and enhanced conversions setup: 1 to 2 hours, one-time. Feed title rebuild for top 200 SKUs by revenue: 4 to 8 hours, one-time. Brand exclusion list and branded Search campaign setup: 2 to 3 hours, one-time. Custom label setup for margin segmentation: 2 to 4 hours if margin data exists in Shopify. Auto-apply settings review and disabling: 30 minutes.
Total first-pass google ads best practices implementation on an active ecommerce account: $800 to $2,500 in specialist time, one-time. At $20,000 monthly ad spend, a 15% ROAS improvement from these changes adds $3,000 in monthly revenue. The investment pays back within the first month.
Ongoing monthly management (search term reviews, bidding monitoring, Merchant Center diagnostics): 4 to 8 hours per month. Our ecommerce PPC management services cover how this is structured for active accounts.
Why work with Ishant Sharma on google ads best practices
Twelve years. 500+ brands. $780M+ in trackable client revenue. Google Partner and Meta Business Partner. Upwork Top Rated Plus with a 99% Job Success Score and a 5.0/5.0 rating. Clutch Award Winner 2024.
The eight practices in this article are the first things I check on every new account, in this order. ArmorGarage, P-REX Hobby, and ThePetsClub all produced their documented results through the same diagnostic sequence, not through better creative or higher budgets. The highest-performing Google Ads accounts aren’t doing more things. They’re doing the right things in the right order with accurate data underneath all of it.
Hustle Marketers’ ecommerce PPC agency page covers how we scope and run these engagements.
What to take from this
Google ads best practices for ecommerce aren’t a checklist to complete once and revisit quarterly. They’re the diagnostic framework you apply when something underperforms and the maintenance discipline you run monthly to prevent the silent account degradation that comes from ignored Merchant Center disapprovals, auto-applied recommendations, and conversion tracking that drifts.
The accounts that consistently produce strong ROAS aren’t running more sophisticated google ads best practices than everyone else. They’re running the foundational ones correctly, continuously, with accurate data underneath every Smart Bidding decision.
Start with tracking. Fix the feed. Isolate brand. Then build.
About Ishant Sharma
Ishant Sharma is a Google Ads specialist and Founder of Hustle Marketers, a Google Partner and Meta Business Partner agency working with e-commerce and lead-gen brands across the US, UK, UAE, and Australia. 12+ years in performance marketing. Trackable client revenue across his work has crossed $780 million. Upwork Top Rated Plus with a 99% Job Success Score and a 5.0/5.0 rating. Clutch Award Winner 2024. Based in Chandigarh, India.