Profit Margins: Gross, Operating & Net Margin Explained

Industry

Net margin

Internet / software

20–40%

Manufacturing

5–15%

Restaurant

3–10%

Retail

2–5%

Construction

1–4%

Key Takeaways

Waterfall: revenue → −COGS = gross profit → −opex = operating → −interest/tax = net

Three margins: gross (product power), operating (management efficiency), net (final take)

Compare within industry: net margins — software 15-25%, retail 1-4%, manufacturing 5-10%

Diagnostic rule: margin gaps between levels locate the problem (COGS vs overhead vs financing)

Core answer: "Profit rate" has three distinct meanings: cost-based profit rate (profit ÷ cost), sales-based margin (profit ÷ revenue), and capital-based return (profit ÷ invested capital). ¥60 cost selling ¥100: cost profit rate 66.7%, sales margin 40%. Chinese business speech (利润率) usually means profit ÷ revenue for margins and profit ÷ cost in trade contexts — always clarify the denominator before comparing numbers.

The three denominators

Metric	Formula	Use
Cost profit rate 成本利润率	profit ÷ cost	trade, manufacturing quotes
Sales margin 销售利润率	profit ÷ revenue	financial statements
Return on capital 资本金利润率	profit ÷ capital	investment efficiency

China's industrial statistics use 利润率 = total profit ÷ operating revenue (sales basis).

The profit ladder (income statement)

Gross margin = (revenue − COGS) ÷ revenue — production efficiency
Operating margin = operating profit ÷ revenue — core business quality
Net margin = net profit ÷ revenue — the bottom line after everything

A company can show 40% gross and 3% net — the gap (37 points) is opex, interest, and tax. Read all three before judging.

Worked examples

Example 1 — The quote confusion. Supplier says "我们只有10个点利润" on a ¥100 quote — almost certainly cost-based: their cost ≈ ¥90.9 (100 ÷ 1.10). As sales-basis it would be cost ¥90. Small difference here, huge at 40 "points": cost-based 40% = sales-based 28.6%.

Example 2 — Restaurant reality. Revenue ¥800k/month, ingredient cost ¥260k (67.5% gross margin), labor ¥180k, rent ¥90k, utilities/misc ¥60k → operating profit ¥210k = 26%… then owner salary, depreciation, tax → net ~15–18%. The "restaurants make 70%" myth confuses gross with net.

Example 3 — Capital efficiency. Two shops both earn ¥500k/year: A invested ¥2M (25% return), B invested ¥5M (10%). Same profit, very different quality — A's model scales, B's ties up capital.

Example 4 — Listed-company reading. 贵州茅台 net margin ~50%, supermarkets ~2%, both excellent for their industries. Cross-industry margin comparison is meaningless; compare against industry quartiles and the company's own history.

Common mistakes and myths

Denominator ambiguity — "20% profit" without a basis is noise; cost-basis and sales-basis diverge more as numbers grow.
Revenue growth ≠ profit growth — scaling at negative unit economics accelerates losses; check margin trend alongside revenue trend.
One-off gains in "profit" — asset sales and subsidies inflate net margin for a year; operating margin strips most of the noise.
Ignoring cash conversion — profit on paper with receivables piling up is not money; operating cash flow ÷ net profit near or above 1 signals quality.
Forgetting owner's salary — small-business "profit" often includes the owner's unpaid labor; deduct a market salary before calling it profit.

Frequently Asked Questions

What are the three profit margins?

Gross margin = (revenue − COGS) ÷ revenue — product economics. Operating margin = operating profit ÷ revenue — after SG&A and R&D: management efficiency. Net margin = net income ÷ revenue — after interest, taxes and one-offs: what shareholders keep. ¥1M revenue, ¥600k COGS, ¥280k opex, ¥40k interest+tax: gross 40%, operating 12%, net 8%. Each level strips another layer of cost.

How does the profit waterfall work?

Revenue 100 → COGS 60 → gross profit 40 → selling 15 + admin 10 + R&D 5 → operating profit 10 → interest 2 + tax 2 → net 6. Reading it: where the 100 leaks tells the story — heavy COGS = pricing/cost problem; heavy opex = scale or discipline problem; heavy interest = capital structure problem. One waterfall beats three ratios for spotting where profit dies.

What margins are normal by industry?

Net margin guideposts: software/SaaS 15-25% (mature), pharma 15-20%, banks 25-35% (different model), consumer brands 5-15%, manufacturing 5-10%, retail 1-4%, supermarkets 1-3%, airlines single digits and cyclical. Gross margins: SaaS 70%+, retail 20-35%, manufacturing 20-40%. Never compare across industries — 5% net is elite for a grocer, alarming for software. Trend over 3-5 years beats any single-year snapshot.

How do margin gaps diagnose problems?

Compare the three levels against peers: gross margin below peers → pricing power or production cost issue. Gross fine but operating weak → overhead bloat or undersized scale. Operating fine but net weak → debt load (interest) or tax inefficiency. Example: gross 45% (peer 42%) but operating 8% (peer 15%) → ¥7 of every ¥100 is leaking in SG&A — audit headcount and marketing ROI before touching the product.

How do I improve each margin level?

Gross: raise prices (strongest lever), cut COGS, shift mix to high-margin products. Operating: scale fixed costs over more revenue, cut zombie projects, automate processes — but starve R&D/marketing and tomorrow's gross margin dies. Net: refinance expensive debt, use tax incentives legally, clean up one-off losses. Watch the trade-offs: cutting the sales team lifts this year's operating margin and next year's revenue problem.

What mistakes happen with profit margins?

(1) Comparing net margins across industries; (2) one-year snapshots (one-off gains/losses distort net — check the notes); (3) chasing margin % while profit ¥ shrinks (50% of nothing is nothing); (4) ignoring cash flow (profitable companies die from cash gaps — margin ≠ liquidity); (5) cutting into muscle (R&D, quality) to dress up margins; (6) forgetting owner compensation in small-business margins (a "30% margin" that includes no market-rate salary for the founder is fiction); (7) blended margins hiding a loss-making segment that should be fixed or killed.

Industry margin benchmarks