Losing Money on Facebook Ads? Here's What to Do About It

You’re looking at your Meta Ads Manager and the numbers seem OK. This campaign has a 2.8x ROAS. That one has 3.1x. You’re thinking: “We’re crushing it.”

Then you pull your Shopify reports and cross-reference against the dates the campaigns ran.

The math doesn’t add up.

Meta claims $12,000 in attributed revenue. Your Shopify store recorded $8,100 in actual orders from people who clicked those ads. The gap is $3,900 — and that’s after accounting for customers who ordered on day 2 or 3 after clicking the ad.

After subtracting the cost of goods (35% of revenue), your margin on those $8,100 orders was $5,265. You spent $4,300 on ads. You made $965 profit.

But Meta’s ROAS would tell you this was a huge win: $12,000 / $4,300 = 2.79x. Looks great. You’d keep scaling the campaign. In reality, you’re barely breaking even.

This scenario plays out in a thousand Shopify stores every day. Merchants are losing money without realizing it because they’re trusting the wrong metrics. Let’s fix that.

The good news: once you know your real ROAS, you can set up automated kill rules to pause losing campaigns before they drain your budget.

The Attribution Problem — Meta Claims Credit for Sales It Didn’t Make

Meta’s attribution system is generous. Very generous.

Here’s how it works: if someone sees your ad on Wednesday and clicks “Learn More,” that click is recorded. But Meta also uses “view-through attribution.” If someone sees your ad but doesn’t click, and then comes back to your site on Friday (from a search, a direct link, whatever), Meta can still take credit for that sale — even though the original ad didn’t cause the purchase.

The window is 7 days for clicks and 1 day for views. This means:

• Customer sees your ad Monday evening
• Comes back to your site Thursday from a Google search
• Buys something Thursday
• Meta attributes that Thursday sale to your ad, even though the customer came from Google

This isn’t malicious. It’s just how Meta’s system is designed. But for you as a merchant, it inflates your ROAS and makes underperforming campaigns look like winners.

Here’s what Meta’s attribution engine doesn’t see:

1. Customers who would have bought anyway — Your brand name searches are probably going to convert regardless of ads. If someone searches “your-store.com” and comes from Google, Meta can’t take credit. But if they saw your ad yesterday, Meta might.

2. Multi-touch attribution — Only the last interaction before purchase counts in Meta’s system. Your email campaign converted the sale, but if the customer saw your ad 5 days earlier, Meta gets 100% credit.

3. Honest cross-device gaps — Someone sees your ad on mobile, but completes the purchase on desktop hours later. Meta might miss this.

4. Returns and chargebacks — Meta attributes the full order value at purchase. But if 10% of orders are returned within 30 days, Meta’s ROAS is inflated by 10%.

The result: Across hundreds of stores, we’ve found that Meta’s attributed revenue is 20–40% higher than the actual revenue tied to those campaigns based on Shopify data.

If Meta says a campaign generated $10,000 revenue, the real number is probably $7,000–8,000.

The COGS Problem — High ROAS Doesn’t Mean High Profit

Even when attribution is accurate, ROAS is a revenue metric, not a profit metric.

Let’s say your campaign has a 3.5x ROAS. That sounds fantastic. But here’s the real math:

• Campaign spend: $2,000
• Attributed revenue (Meta): $7,000
• ROAS: 3.5x ✓ Looks great

But dig deeper:

• Actual orders (Shopify): 60 units (let’s assume $116 average order value)
• Cost of goods sold (COGS): 40% of revenue = $2,784
• Shopify fees (2.9% + $0.30): $200 × 2.9% + ($200 × 0.30) = $320
• Payment processing (Stripe): $200 × 2.2% = $44
• Shipping costs: $600 (assuming $10 per order)
• Refunds/chargebacks: 5% = $348

Total costs: $4,096

• Gross revenue: $6,960
• Total costs: $4,096
• Profit: $2,864

But wait, we already spent $2,000 on ads, which is in the “costs” above. So:

• Real profit after ads: $2,864 - $2,000 = $864 on $7,000 attributed revenue = 12% net margin

That’s actually healthy. But now let’s look at what happens when COGS is higher (which it often is):

• Campaign spend: $2,000
• Attributed revenue (Meta): $7,000
• ROAS: 3.5x (unchanged)

But COGS is 50% instead of 40%:

• COGS: $3,480
• Payment processing, shipping, refunds: $1,240
• Shopify fees: $320
• Total costs: $5,040
• Gross revenue: $6,960
• Profit before we even count the $2,000 ad spend: $1,920
• Profit after ad spend: -$80 (losing money)

Same 3.5x ROAS, but you’re losing money because your product costs more.

This is why COGS matters so much. A 3x ROAS at 30% COGS is wildly profitable. A 3x ROAS at 55% COGS is barely breakeven.

Most merchants don’t track COGS closely in their Meta reports. They see ROAS and assume profit. This is the core mistake that leaves thousands of dollars on the table.

The Slow Bleed — Campaigns That Start Strong and Gradually Die

Some campaigns don’t crash immediately. They start well and gradually decline.

Here’s what it looks like:

• Day 1-2: Campaign spends $50, makes 2 sales, ROAS is 4.2x. Excited.
• Day 3-4: Campaign spends $95 total, makes 4 sales total, ROAS is 3.1x. Still good.
• Day 5-6: Campaign spends $180 total, makes 7 sales total, ROAS is 2.8x. Declining but OK.
• Day 7-8: Campaign spends $310 total, makes 11 sales total, ROAS is 2.4x. Hmm.
• Day 9-10: Campaign spends $450 total, makes 15 sales total, ROAS is 2.1x. Uh oh.
• Day 11-12: Campaign spends $650 total, makes 19 sales total, ROAS is 1.8x. Pause?
• Day 13-14: Campaign spends $900 total, makes 22 sales total, ROAS is 1.6x. Definitely pause.

You waited 14 days to pause a campaign that’s been declining the entire time. Total wasted spend: maybe $150–200 from days 9–14.

This is the slow bleed. Manual monitoring misses it because you check once a day or once a week. By the time you notice the trend, you’ve already lost money.

An automated system that checks every 15 minutes would have paused this campaign on day 9 when ROAS first dropped below 2.0x. You’d have saved $200+.

How to Calculate Your Real Meta Ad Profitability

Here’s the formula every Shopify merchant should be using:

Real profit = (Shopify revenue from campaign) - (COGS) - (Platform fees) - (Ad spend)

Specifically:

1. Pull Shopify revenue for the campaign period — Use a custom report filtered by UTM source = “facebook” or “instagram” (make sure your Meta pixel is properly set up to pass UTM parameters). Take the sum of all orders in that window.

2. Calculate COGS — If you track COGS per product in Shopify, pull that directly. If not, estimate it: take your annual COGS / annual revenue × campaign revenue. Or track it manually in a spreadsheet.

3. Subtract platform fees — Shopify (2.9% + $0.30 per order) + payment processing (typically 2.2% for Stripe). This is usually 3.5–4% of revenue.

4. Subtract ad spend — The total you spent in Meta Ads Manager.

5. Compare to ROAS — Campaign ROAS = Shopify revenue / ad spend. If your ROAS is 3x but your real profit is only 10% margin, you’re underpricing or overspending.

Example with real numbers:

• Campaign ran from April 1–7
• Meta reports $8,500 attributed revenue
• Shopify reports $5,800 actual revenue from linked orders (April 1-8, accounting for 2-day lag)
• Average COGS on those orders: 38% = $2,204
• Platform fees: $5,800 × 3.8% = $220
• Ad spend: $1,900
• Real profit: $5,800 - $2,204 - $220 - $1,900 = $1,476
• Real profit margin: $1,476 / $5,800 = 25.4%
• Meta’s reported ROAS: $8,500 / $1,900 = 4.47x (misleading — 32% higher than real revenue)
• Your actual ROAS: $5,800 / $1,900 = 3.05x

So you’d scale up a campaign that Meta says is 4.47x, when the real number is 3.05x. This leads to over-spending because your benchmarks are wrong.

How to Stop the Bleeding Automatically

The first step is understanding when you should actually pause a campaign. Once you know your pause thresholds, automation takes over.

Manual spreadsheets and daily checks aren’t sustainable. You need three things:

1. Real-time data — Know the real ROAS (based on Shopify data, not Meta’s inflated number) every 15 minutes
2. Automated kill rules — Pause campaigns as soon as they cross your profitability thresholds
3. COGS integration — Account for the cost of goods in your pause rules

Here’s what this looks like in practice:

You set a kill rule: “Pause if ROAS < 2.0x after $50 spend”

The system:

• Checks Meta’s API every 15 minutes for spend data
• Pulls Shopify orders for customers who clicked the campaign
• Calculates real ROAS based on actual orders (not Meta’s attribution)
• If ROAS drops below 2.0x, pauses the campaign automatically
• Logs the pause with context (spend, ROAS, why it was paused)

You can set multiple rules:

• Aggressive: Pause if (ROAS < 1.5x AND spend > $100)
• Conservative: Pause if (ROAS < 3.0x after $200 spend)
• Sustainable: Pause if (contribution margin < 10%) — accounting for COGS

You can also set rules that scale up instead of pausing:

• Scale up: Resume a paused campaign if (ROAS > 2.5x for 48 hours)

This way, you’re not just killing underperformers — you’re doubling down on winners.

Tools like Calatrix automate this entirely. You connect Meta and Shopify once, set your rules, and the system runs 24/7. You get a log of every pause and can review decisions later.

FAQ

Q: Should I trust Meta’s ROAS or my own calculations?

A: Your own calculations based on Shopify data. Meta’s ROAS is typically 20–40% inflated due to view-through attribution and multi-touch modeling. Always cross-check against Shopify.

Q: What if my product has high COGS (60%+) — does that mean I can’t run profitable ads?

A: It’s harder, but possible. You’d need a 4–5x ROAS minimum to be profitable after all costs. This is only achievable for high-ticket items (>$200 AOV) or repeat customers. For low-AOV products with high COGS, Meta ads are often not profitable without either raising prices or improving customer lifetime value.

Q: How often should I check campaign performance?

A: Daily minimum, but hourly is better. Automated rules are best — they check every 15 minutes. The sooner you pause a loser, the less money you waste.

Q: Can I use Meta’s conversion pixels for ROAS instead of Shopify order data?

A: You can, but it’s less accurate. Meta’s pixel fires on successful purchase, but it doesn’t capture returns, chargebacks, or product COGS. Use Shopify as your source of truth and compare to Meta’s pixel data to understand the gap.

Q: What’s a realistic profitability target for Facebook ads?

A: Most profitable stores target 20–35% contribution margin (profit after COGS, platform fees, and ad spend, before overhead). Anything below 15% is risky.

Stop Guessing — Start Measuring

The only way to stop losing money on Meta ads is to measure real profit, not vanity metrics. Connect your Shopify store to an automated kill rule system, account for COGS, and let it run.

Related articles

• When to Pause a Facebook Ad Campaign
• How to Automatically Pause Underperforming Meta Campaigns

Start your free trial today. Within 5 minutes, you’ll have a system that catches losing campaigns before they cost you hundreds more.