Which Ad Platform Works for Ecommerce (And How to Track Results)
You're running an ecommerce business. You've got four major advertising platforms demanding your attention and budget: Facebook, Google, TikTok, and Amazon.
Each one promises incredible results. Each one uses different metrics that don't match. And you need to know—before the budget disappears—which platform actually delivers customers who buy.
Here's what matters: not clicks, not impressions, not even traffic. Revenue. Specifically, revenue that costs less to generate than it brings in.
Let's break down which platforms work for ecommerce and how to track what actually matters.
The Platform Landscape (And What Each One Actually Does)
Facebook and Instagram dominate ecommerce advertising—they capture roughly 60% of social commerce ad spend. There's a reason. Their targeting lets you reach people based on behavior, interests, and purchase history. You're not guessing who might buy; you're targeting people who've bought similar products.
Google Shopping works differently. These are high-intent searches. Someone typing "waterproof hiking boots size 10" isn't browsing—they're buying. Conversion rates typically run 2-3x higher than social platforms, but the audience is smaller because you're limited to people actively searching.
TikTok is the wild card. If your product is visual, demonstrates well, and appeals to buyers under 40, you'll see cost per acquisition 30-40% lower than Facebook (in most cases). The catch? Creative burns out fast—you need fresh video content every 7-10 days.
Pinterest drives purchases for home decor, fashion, and lifestyle products. The platform captures people in planning mode—they're thinking about their kitchen renovation or fall wardrobe three months before they buy. Long consideration cycles, but strong conversion rates once they're ready.
The Metrics That Actually Matter
Most ecommerce brands track the wrong numbers.
They obsess over click-through rates and cost per click. Those metrics tell you nothing about profitability. What matters is the full picture—from first click to completed purchase to repeat order.
Cost Per Acquisition (CPA) is your starting point. This is what you pay to acquire one customer. Calculate it simply: total ad spend divided by number of customers acquired. If you spent $3,000 and got 50 customers, your CPA is $60.
But here's where it gets crucial: you need to know your target CPA. Take your average order value, subtract your product costs and fulfillment, then subtract your desired profit margin. What's left is what you can afford to pay for a customer.
Return on Ad Spend (ROAS) shows revenue generated per dollar spent. A 3:1 ROAS means every dollar in ads returns three dollars in revenue. Sounds good—but it depends entirely on your margins. A 3:1 ROAS with 15% margins loses money. A 3:1 ROAS with 60% margins prints cash.
Conversion Rate varies wildly by platform. Google Shopping typically converts at 2-3%. Facebook and Instagram run 1-2%. TikTok can hit 3-4% for the right products. These benchmarks help you diagnose problems—if your Facebook ads convert at 0.3%, something's broken (either your targeting, your landing page, or your offer).
The metric almost nobody tracks properly? Customer Lifetime Value (LTV) to CAC ratio. This compares what you pay to acquire a customer against what they spend over their entire relationship with your brand. You want this ratio at 3:1 minimum—meaning each customer generates three times what you spent to acquire them.
Platform-Specific Tracking (The Technical Stuff That Makes or Breaks Campaigns)
Facebook's Pixel captures everything—page views, add to carts, purchases. Install it on every page of your site. But here's what changes after iOS 14.5: you're now working with modeled data, not perfect tracking. Facebook estimates 20-30% of conversions go unreported in their dashboard.
That means you need server-side tracking through the Conversions API. This sends purchase data directly from your server to Facebook, bypassing browser limitations. Shopify merchants can set this up in about 15 minutes using the native integration.
Google's tracking runs through Google Analytics 4 (GA4) and Google Tag Manager. You'll want enhanced ecommerce tracking enabled—this captures product impressions, clicks, add to cart events, and purchases with full revenue data. Link your Google Ads account to GA4 so you can see the complete customer journey from ad click to purchase.
For TikTok, install their Pixel and enable Enhanced Match. This passes hashed customer data (email, phone number) to TikTok, improving attribution accuracy by 15-25%. You'll also want to set up TikTok Events API—their version of server-side tracking.
Pinterest requires the Pinterest Tag and event tracking for add to cart and checkout. Their attribution window is crucial: they track conversions up to 30 days after the click because Pinterest users plan purchases weeks in advance.
How to Actually Choose Your Platform
Start with one platform. Not three, not five—one.
Here's why: each platform requires different creative, different messaging, different optimization strategies. Spreading $5,000 across three platforms gives you $1,667 per platform—not enough to gather meaningful data or achieve platform-specific learning.
Choose based on your product and customer.
If you sell products people search for specifically (replacement parts, specific tools, known brands), start with Google Shopping. The intent is highest, conversion rates are strongest, and you're capturing existing demand.
If your product requires education or solves a problem people don't know they have, Facebook and Instagram work better. You can target based on interests and behaviors, then educate through your ad creative. The longer your sales cycle, the more Facebook's retargeting becomes essential.
Visual products that demonstrate well—especially if they're under $100 and appeal to younger buyers—should test TikTok first. The platform rewards entertaining content that doesn't feel like ads. Your cost per thousand impressions (CPM) will run $4-8 compared to $15-25 on Facebook.
Home decor, fashion, food, and DIY products belong on Pinterest. The audience is 60% female, skews 30-50 years old, and has household income 20% above average. They're planners who save ideas for months before purchasing.
The Testing Framework That Prevents Budget Waste
Don't "test" platforms with $500 and three days. You'll learn nothing except that small budgets don't work.
Proper testing requires $2,000-3,000 minimum per platform and 14 days of data. Here's the framework:
Days 1-7: Learning phase. The platform's algorithm is figuring out who converts. Your numbers will be terrible—CPA might be 2-3x your target. Don't panic and don't turn off campaigns. You're paying for the platform to learn.
Days 8-14: Optimization phase. The algorithm has enough data to improve targeting. You should see CPA dropping and conversion rates improving. This is where you evaluate whether the platform can hit your targets.
Days 15-30: Scaling phase (if results are good). Increase budget by 20% every 3-4 days. Faster increases reset the learning phase and tank performance.
Track these specific benchmarks by day 14:
- CPA within 40% of your target (if target is $50, you should be at $70 or better)
- ROAS above your breakeven point (even if below your goal)
- Conversion rate within platform norms for your industry
If you're not hitting these by day 14, the problem is usually your offer, your landing page, or your product-market fit—not the platform.
When to Scale, When to Cut, When to Pivot
You've been running ads for 30 days. Now what?
Scale when your CPA is at or below target, ROAS exceeds goals by 20%+, and you're seeing consistent daily performance for 7+ days. Increase budget by 20% every 3-4 days and watch for performance drops.
Cut when CPA remains 50%+ above target after 21 days, ROAS stays below breakeven after testing multiple creative variations, or you've spent 5x your target CPA without a single conversion.
Pivot when one element works but others don't. Good CTR but no conversions means fix your landing page. Good conversion rate but terrible CTR means improve your creative. High costs across the board means test different targeting.
The platform that works best isn't always the one you expect. Test methodically, track religiously, and let the data decide.
Related Guides: Platform Comparison Guide, Platform Benchmarks, Tracking Setup Guide.