E-Commerce Advertising: Google, Meta and Amazon (2026)
E-Commerce Advertising: Google, Meta and Amazon (2026)
How e-commerce ads work: which platform for which product, reading ROAS/ACOS/POAS, budget testing and scaling. A practical framework from a team running ads for its own brands.

Kerem Başbuğ
Founder
Advertising is where e-commerce burns money fastest — and where the fastest growth comes from. The difference is platform choice and reading the numbers right. This guide is the distilled framework of a team running ads daily for its own 3 Amazon brands and Shopify stores: which platform for which product, what the core metrics say, how to test and scale budget.
First, the right question: which intent are you buying?
Three platforms target three different buying moments:
- Google Ads → products with existing demand. The user is already searching; you meet the search. "Buy leather wallet" is a ready buyer. High intent, higher conversion.
- Meta (Instagram/Facebook) → products whose demand you create. The user isn't searching; you create interest with visuals/video. The channel for impulse, visual, "saw it, want it" products.
- Amazon Ads → products at the moment of purchase. The user is already on Amazon to buy. Competition is per keyword, and ads feed organic rank too. See the Amazon guide.
Promoting the wrong product on the wrong platform is the quietest budget killer.
Learn to read ROAS, ACOS, POAS
- ROAS: ad revenue / ad spend. 4 ROAS = $4 revenue per $1 spent. But revenue isn't profit.
- ACOS (Amazon): ad spend / ad revenue — the inverse of ROAS. 25% ACOS = 4 ROAS.
- POAS (Profit on Ad Spend): the one that matters. With a 30% margin, 4 ROAS profits; with a 15% margin, the same ROAS loses money. Lock your target to profit, not revenue.
Rule of thumb: your break-even ROAS = 1 / gross margin. At 25% margin, break-even is 4; you target above it for profit.
1. Set up measurement (don't advertise without it)
Ads without conversion tracking is throwing money blind. Meta Pixel/Conversions API, Google Ads conversion tags and GA4 come first. A misconfigured pixel is the #1 reason for the "ads don't work" illusion — they do; you just can't see it.
2. Test on a small budget
The first week's goal is data, not profit: which product, creative and audience convert. Start with a few clear variations (not 20 ads at once) and give each enough budget to learn. Early shutdown is the most common mistake — killing a campaign in the learning phase resets the data.
3. Find the winner, scale it
Once the data is clear, shift budget to the winning ad/audience/product. Scaling discipline: 5×-ing budget overnight breaks learning; grow in steps (20–50% increments). When a winner tires (frequency up, CTR down), rotate in new creative — ad fatigue on Meta is real.
4. Build the funnel: cold, warm, hot
- Cold: people who don't know you; create interest (video, strong hook).
- Warm (retargeting): browsed or added to cart but didn't buy; the highest-ROI layer. Abandoned-cart + retargeting is most stores' most profitable spend.
- Hot: existing customers; repeat sales via email/SMS (cheaper than ads). See the Shopify guide.
5. What we do with our own software
To be transparent: on our own brands we run ads with automation we built — budget shifting, rule-based bidding, performance alerts — freeing human effort for strategy and creative. We use the same infrastructure on client accounts; see ad management and Amazon PPC management.
Five common mistakes
- Advertising without conversion tracking (blind flight)
- Watching ROAS and forgetting POAS (profit)
- Killing campaigns in the learning phase
- Skipping the retargeting/email layer and always paying for cold traffic
- Never refreshing creative and surrendering to ad fatigue
FAQ
How much monthly budget? Product- and platform-dependent; a practical floor for meaningful testing is a daily budget 5–10× your product price — the algorithm needs enough conversion data to learn.
Agency or DIY? Start DIY on a small budget and one channel; outsource when multi-channel, scale and time constraints hit. What matters isn't who runs it, but having a discipline that watches profit (POAS).
Google or Meta first? Google if your product is searched, Meta if it's visual/impulse. If budget won't cover both, start with the one matching your product's intent profile.
The takeaway
E-commerce advertising is the strongest growth lever with the right platform, the right metric and patient test-scale discipline — and the fastest money-burner when set up wrong. Lock onto profit (POAS), build measurement, scale winners in steps, complete the funnel.
To have us run your ad accounts against a profit target, there's ad management (Google & Meta) and Amazon PPC management. The first discovery call is free.

About the author
Kerem Başbuğ · Founder
I've run my own e-commerce brands for many years: 3 registered trademarks, 14+ active stores and automation tools I built myself. I founded Revoba to put that same operating discipline to work for a small number of carefully chosen clients.