PPC for E-commerce Stores: Shopping, Performance Max and Beyond

Running paid search for an online shop is a different job from running it for a clinic or a renovation contractor. You are not advertising one service with a handful of keywords; you are advertising hundreds or thousands of products, each with its own price, margin, stock level and seasonality. The ads themselves are increasingly generated from your product data rather than written by hand.

That is why PPC for e-commerce starts with the product feed, not the keyword list. If your feed is thin, inaccurate or disapproved, no bidding strategy will save the account. If the feed is strong and conversion tracking reports real order values, Shopping, Performance Max, search and remarketing can work together instead of competing for the same budget.

This guide walks Singapore e-commerce owners through the main campaign types, how Google Merchant Center feeds work, how to think about ROAS targets using your own margins, and a practical plan for building or cleaning up an account. It also covers what AI tools can and cannot do for a small online store.

How PPC for e-commerce differs from lead-generation advertising

A lead-generation advertiser optimises for form fills or WhatsApp enquiries, which are worth roughly the same amount each. An online store records a different value for every order, and the profit on each order depends on the product mix, discount and delivery cost. Three consequences follow:

• Revenue is not profit. A campaign that drives SGD 10,000 in sales on low-margin electronics may be worth less than one driving SGD 4,000 on high-margin accessories.
• Product data drives targeting. For Shopping and Performance Max, Google matches searches to your products largely using titles, descriptions, categories and attributes in your feed.
• Measurement must capture order value. Purchase conversions need to pass the actual transaction value and currency, or value-based bidding will be guessing.

If you are still choosing a shop platform, check how well it supports Merchant Center feeds and purchase tracking; our overview of e-commerce platforms in Singapore covers the wider decision.

Google Merchant Center and the product feed

Google Merchant Center is where you upload and manage product data for Google. The feed can come from a platform integration (Shopify, WooCommerce and others have connectors), a scheduled file, a spreadsheet or the Content API. Once approved, products can appear in Shopping ads, Performance Max and, where available, unpaid free listings.

Required and important attributes

Google’s product data specification sets out which attributes are required. The core set includes an ID, title, description, link, image link, price and availability. Brand, GTIN or MPN, and condition are required in many cases, and apparel needs extra attributes such as colour, size, gender and age group, with an item group ID for variants.

Feed quality habits that pay off

• Write titles the way customers search: brand, product type, key attribute (size, colour, material) and model, in a sensible order for your category.
• Keep price and stock in sync automatically; manual uploads go stale quickly during sales such as 11.11 or year-end promotions.
• Use custom labels to tag products by margin band, price band, seasonality or clearance so you can set different targets later.
• Check the Merchant Center diagnostics regularly and fix disapprovals before they pile up.
• Make sure your on-page Product markup agrees with the feed; our schema markup guide explains how product structured data works.

Shopping, Performance Max, search and remarketing compared

Most online stores use a combination of campaign types. The right mix depends on catalogue size, budget and how much control you need.

Standard Shopping

Standard Shopping gives you the search-terms report and product-group level control. Many smaller stores still start here, or use it alongside Performance Max for specific product groups they want to manage closely. Add negative keywords for irrelevant queries such as “free”, “second hand” or competitor names you do not want to target.

Performance Max

Performance Max uses your feed plus text, image and video assets to serve across Google’s inventory, and relies on automated bidding toward your conversion goals. It works best with accurate purchase values, enough conversion volume and asset groups that reflect real product categories. Google has adjusted how Performance Max and Standard Shopping interact when both can show the same product, so check the current Google Ads Help guidance on Performance Max before designing your structure. Our guide to optimising Performance Max campaigns covers asset groups, signals and reporting in depth.

Search campaigns

Text ads still matter for brand searches, category searches that do not map neatly to one product, and service-like queries such as gift delivery or corporate orders. Keep them focused and avoid bidding on the exact product queries already covered well by Shopping.

Remarketing

Shoppers often compare across several sites before buying, and many browse on mobile and buy later. Remarketing brings back cart abandoners and past buyers, but it needs consent-aware tracking under the PDPA and sensible frequency caps so you are not following people around for weeks.

Setting ROAS targets from your margins

Return on ad spend is revenue divided by ad cost. It is useful, but a “good” ROAS depends entirely on your margins. A quick way to find your break-even ROAS is to divide 1 by your gross margin after variable costs.

These are illustrative calculations, not benchmarks. Two refinements make the numbers more useful. First, factor in repeat purchase: if customers typically order again, you may accept a lower first-order ROAS for new customers. Second, segment by margin band using custom labels, so high-margin products are not held back by a single account-wide target.

Google’s Target ROAS and Maximise conversion value strategies need reliable conversion values to work. If your purchase tracking misses orders or double-counts them, fix that first using the steps in our guide to conversion tracking and UTM parameters. Read more about value-based bidding before tightening targets.

A practical e-commerce PPC plan

The timings below are rough planning ranges for a small store with an in-house owner or marketer and, optionally, a freelancer.

1. Measurement first (week 1, owner and developer). Confirm purchase conversions fire once per order with the correct value and SGD currency, and that consent handling is in place.
2. Feed audit (weeks 1 to 2, marketer). Fix disapprovals, rewrite the top 50 to 100 product titles by revenue, add GTINs and brand, and set custom labels for margin and seasonality.
3. Initial structure (week 2). Start with one Performance Max or Standard Shopping campaign split by margin band or main category, plus a small brand search campaign.
4. Learning period (weeks 3 to 6). Use Maximise conversion value without a target, or a modest target, while data builds. Avoid daily changes.
5. Review search terms and placements (fortnightly). Add negatives, check which products get spend without sales, and fix feed issues behind them.
6. Set targets (from week 6). Introduce Target ROAS per margin group, based on actual results and your break-even maths.
7. Add remarketing and seasonal plans (month 2 onwards). Plan budgets around 9.9, 11.11, 12.12, Black Friday, Christmas and Chinese New Year if relevant to your category.

For budget sizing, see our guide on how to determine a PPC budget.

Using AI in e-commerce PPC, with human oversight

Google’s own automation already handles much of the bidding and placement in Performance Max. Beyond that, AI tools are useful for rewriting feed titles in bulk, drafting ad text and asset variations, summarising search-term reports, and flagging products with spend but no sales.

The limits are important. AI-written titles can introduce claims you cannot back up (for example “best” or “authentic” on a reseller item), and automated suggestions do not know your margins, stock commitments or supplier issues unless you tell them. Always review changes before they reach the live feed or account.

Marketing2u’s AI Ads Agent monitors Google Ads performance, search terms and wasted spend through a read-only sync, and its page states it never changes your ad accounts; recommendations are for you to apply. For broader store workflows, see AI agents for e-commerce.

Common e-commerce PPC mistakes

• Launching Performance Max before purchase values are tracked correctly.
• Treating revenue ROAS as profit and scaling low-margin products that lose money.
• Ignoring feed disapprovals and letting best-sellers drop out of ads.
• Lumping every product into one campaign with one target.
• Running promotions on site without updating feed prices, causing mismatches and disapprovals.
• Making large budget or target changes every few days, so automated bidding never stabilises.
• Forgetting delivery costs and free-shipping thresholds when calculating break-even ROAS.

Bringing it together

Strong PPC for e-commerce is built on three things: a complete, accurate product feed; purchase tracking that records real order values; and targets based on your margins rather than a generic ROAS figure. Get those right, then choose the mix of Shopping, Performance Max, search and remarketing that matches your catalogue and budget. Review regularly, change one thing at a time, and let the data, not the dashboard’s headline numbers, guide your next step.

Frequently asked questions

Should a small online store use Performance Max or Standard Shopping?

Many small stores start with one of the two and add the other once they have data. Performance Max offers wider reach and automation, while Standard Shopping gives more control and search-term visibility. The right choice depends on catalogue size, conversion volume and how closely you want to manage product groups.

What is a good ROAS for e-commerce?

There is no universal figure. Work out your break-even ROAS by dividing 1 by your contribution margin after product, payment and delivery costs, then set targets above that. Repeat purchase value can justify a lower target for new customers.

Do I need Google Merchant Center to run Shopping ads?

Yes. Shopping ads and the product side of Performance Max use product data from a Merchant Center account linked to Google Ads. Most e-commerce platforms offer an integration that keeps the feed updated.

Why are my products disapproved in Merchant Center?

Common reasons include price or availability mismatches with the landing page, missing GTINs or brand, images with promotional overlays and policy issues. The diagnostics section in Merchant Center lists each issue and the affected products. See the product data specification for requirements.

Sources and further reading

• Google Merchant Center Help: Product data specification
• Google Merchant Center Help
• Google Ads Help: About Performance Max campaigns
• Google Ads Help
• PDPC: Personal Data Protection Commission Singapore

Related guides

• 10 Ways To Optimise Performance Max Campaigns In Google Ads For Real Results
• What Is Return On Ad Spend (ROAS): Guide To Smarter Ad Spend
• Remarketing and Retargeting: How to Win Back Website Visitors
• How Does Value-Based Bidding Work? A Simple Guide
• E-commerce SEO Checklist: 26 Crucial Steps To Boost Google Rankings

Browse every article in our Paid Advertising guides.