PulsePulse
Log inGet started free
All articlesAnalytics

What Is a 'Good' ROAS for Facebook Ads in 2026?

A 4x ROAS can still be a losing campaign. A 1.8x ROAS can be wildly profitable. Understanding breakeven ROAS — and why the 'good ROAS' question is the wrong question — is the foundation of scaling correctly.

March 28, 20265 min readPulse Team · Nurdd Solutions

A "good" ROAS is one that is above your breakeven ROAS. There is no universal number. A fashion brand with 65% gross margins has a completely different target than a consumer electronics brand with 18% margins. Benchmarks published online ("aim for 4x!") are meaningless without knowing the margin context.

The Breakeven ROAS Formula

Your breakeven ROAS is the minimum ROAS at which your ad spend returns zero net profit on the attributed revenue:

Breakeven ROAS = 1 ÷ Gross Margin %

Examples

Gross MarginBreakeven ROASWhat 4x ROAS means
60%1.67xProfitable — 2.33x above breakeven
45%2.22xProfitable — 1.78x above breakeven
35%2.86xProfitable — 1.14x above breakeven
25%4.0xAt breakeven — zero profit
18%5.56xStill losing money on ads

A business with 25% gross margins hitting 4x ROAS is at exact breakeven on ad spend — every rupee of "profit" from sales is consumed by the cost of goods. They need 5x+ to actually profit from ads.

Why ROAS Alone Is Insufficient

1. Attribution window distortion

Meta's default attribution is 7-day click + 1-day view. This window often overcounts revenue — it attributes conversions that would have happened anyway (organic, email, direct) to your ads. Your true ad-driven ROAS is typically 15–30% lower than what Ads Manager reports.

2. COD return rates inflate ROAS (India-specific)

For Indian D2C brands using COD, Meta counts the order placement as a conversion — before delivery or return. If your COD return rate is 20%, you're attributing 20% more revenue to your ads than actually lands in your bank. Your effective ROAS = Reported ROAS × (1 − Return Rate).

3. Blended ROAS vs ad-attributed ROAS

Blended ROAS (Total Revenue ÷ Total Ad Spend) is often more reliable than platform-reported ROAS because it captures all revenue regardless of attribution path. If your blended ROAS is significantly lower than your Meta ROAS, you have over-attribution.

Target ROAS by Business Type

Business TypeTypical MarginTarget ROAS (profitable)
D2C Apparel / Fashion50–65%2.0–2.5x+
Beauty / Skincare D2C60–75%1.8–2.2x+
Consumer Electronics12–22%5.0–8.0x+
Food & Grocery D2C25–40%3.0–4.0x+
Supplements / Health65–80%1.5–2.0x+
Lead Gen (SaaS / Services)70–90%Track CPL, not ROAS

The Right Metric Framework

Use ROAS as a directional signal, not the single source of truth. Pair it with:

  • MER (Marketing Efficiency Ratio): Total Revenue ÷ Total Marketing Spend — platform-agnostic efficiency
  • nCAC (new Customer Acquisition Cost): Ad spend to acquire one new customer
  • LTV:CAC ratio: Are you acquiring customers who stick around?

Frequently Asked Questions

What is a good ROAS for Meta Ads in India?

For most Indian D2C brands (apparel, beauty, food) with margins between 40–65%, a ROAS of 2.5x–4x is typically profitable. However, this only holds after accounting for COD return rates. Calculate your specific breakeven: 1 ÷ (gross margin − return rate impact).

My ROAS is 6x but I'm still not profitable. Why?

Common reasons: (1) Very low gross margins — check if your COGS + fulfilment leaves meaningful contribution; (2) High COD return rates attributed as conversions; (3) Attribution over-counting from wide windows; (4) High overheads unaccounted in your "margin" estimate.

Should I use 1-day click attribution or 7-day click?

For most purchases, 7-day click is the industry standard and captures most real intent. For impulse categories (under ₹500), 1-day click is more accurate. The 1-day view attribution is largely noise — it attributes any conversion within 24 hours of an impression, even if the user never engaged.

Stop guessing. Start fixing.

Connect your Meta and Google Ads accounts to Pulse and get a full AI audit in minutes — free.

Try Pulse free →