ROI
ROI is net profit divided by cost, expressed as a percentage. Unlike ROAS, it nets out COGS and overhead — making it the metric that actually governs profitability decisions.
ROI (Return on Investment)
ROI is net profit divided by the cost of the investment, expressed as a percentage — the profitability of a spend, not just its revenue.
Return on Investment measures how much profit a euro of spend actually returns after costs are subtracted. You take the net gain from an investment (revenue minus all associated costs), divide by what you spent, and express the result as a percentage. A campaign with €10,000 net profit on €5,000 spend has 200% ROI.
Unlike ROAS, which only compares revenue to ad spend, ROI nets out cost of goods sold, fulfilment, payment fees, and operational overhead. That makes it the metric that governs real profitability decisions — whether a channel, a campaign, or a tool earns more than it costs to run.
ROI is the most general profitability metric you have. It works for an ad campaign, a CRO test, a new shipping partner, or a software subscription — anywhere money goes in and money comes back, you can compute it.
The trap is what you count as "cost". Most teams compute ROI on media spend only and call a 300% return excellent. Once you fold in COGS, returns, payment processing, and warehouse pick-and-pack, that same campaign can land closer to 20%. The number isn't wrong — it's just answering a different question.
ROI = ((Revenue - Total Costs) / Total Costs) × 100
Revenue
Revenue generated
Gross sales attributable to the investment, net of refunds and discounts.
Total Costs
Total fully-loaded cost
Media spend plus COGS, fulfilment, payment fees, and any operational overhead allocated to the investment.
A Shopify apparel store runs a €8,000 paid social campaign that drives €40,000 in revenue. COGS on those orders is €16,000, fulfilment and payment fees come to €4,000, and €2,000 of agency time is allocated to the campaign.
Revenue: €40,000
Media spend: €8,000
COGS: €16,000
Fulfilment + fees: €4,000
Allocated overhead: €2,000
Total Costs: €30,000
→ ROI = ((40,000 − 30,000) / 30,000) × 100 = 33.3%
ROAS on the same campaign reads 5.0× (40,000 / 8,000), which sounds great. ROI of 33% tells the truer story — the campaign profits about €10,000, which is healthy but nowhere near 5× the money back.
Read ROI as a margin on the money you put at risk. Anything above 0% means you made more than you spent. Above 100% means you doubled it. Negative ROI means the investment lost money even before you account for the opportunity cost of tying that capital up.
Typical ROI ranges by e-commerce investment type (fully-loaded costs)
| Investment type | Below average | Healthy | Strong |
|---|---|---|---|
| Paid search (branded) | < 80% | 150-300% | > 400% |
| Paid social (prospecting) | < 0% | 20-60% | > 100% |
| Email + SMS (owned) | < 300% | 500-900% | > 1500% |
| CRO / A/B testing program | < 100% | 200-500% | > 800% |
| Site speed / tech investment | < 50% | 100-250% | > 400% |
| Influencer / affiliate | < 0% | 30-80% | > 150% |
Two common ROI mistakes distort decisions. First, omitting COGS — which inflates everything and makes paid acquisition look better than owned channels. Second, attributing the wrong revenue window: a CRO test that lifts conversion permanently shouldn't be evaluated on a single month's gain. Match the cost horizon to the benefit horizon.
ROI: frequently asked questions
ROAS (Return on Ad Spend) is revenue divided by ad spend — a top-line ratio. ROI is net profit divided by total cost — a bottom-line ratio. A 4× ROAS can be break-even ROI or 200% ROI depending on your margins. Use ROAS to manage campaigns day-to-day; use ROI to decide whether the channel is worth running at all.
It depends on the investment type and your gross margin. For paid prospecting, anything above 20-30% on fully-loaded costs is healthy. Owned channels like email routinely clear 500%+ because the marginal cost is tiny. CRO programs typically return 200-500% within the first year because gains compound across all traffic.
Yes — if you want a profitability number that drives real decisions. Excluding COGS gives you a contribution-margin ratio, which is useful for in-flight campaign optimisation but misleading for channel-level budget allocation. The more decisions ride on the number, the more cost categories you should fold in.
Estimate the annualised revenue lift from the winning variant (lift % × baseline revenue × 12), subtract testing program costs (tool, analyst time, design/dev) for the same period, then divide by those costs. A 3% lift on €5M annual revenue is €150K — easily 500%+ ROI on a sub-€30K program.
Yes. Any time total costs exceed revenue, ROI goes negative. A campaign with €10K spend and €8K revenue has ROI of −20% (and that's before COGS). Negative ROI is a signal to pause, not necessarily to kill — diagnose whether it's a creative, audience, or attribution-window issue first.
Match the measurement window to the investment's benefit horizon. A one-off campaign can be evaluated in 30-60 days. A site-speed improvement or CRO win compounds over the next 12 months, so amortise costs accordingly. Measuring a long-lived investment on a short window understates ROI dramatically.
ROI sits at the top of the e-commerce metrics stack as the profitability summary. It depends on inputs like AOV, conversion rate, CAC, gross margin, and return rate. Improving any of those upstream metrics flows directly into ROI — which is why ROI on its own is hard to optimise; you optimise its components.
Not always. Early-stage stores often run negative or break-even ROI on prospecting deliberately, betting on LTV from repeat purchases. In that case, payback period and LTV:CAC ratio are better steering metrics than blended ROI, which can look alarming during a growth push.
They cut both revenue and add reverse-logistics cost, so they hit ROI from two directions. In apparel and beauty, where return rates can top 25%, calculating ROI on gross sales rather than net sales overstates profitability significantly. Always use post-return revenue in the numerator.
Both. Per-channel ROI shows where to shift budget; blended ROI shows whether the overall business is profitable. Be cautious with per-channel ROI when channels assist each other — last-click attribution can make brand search look heroic and paid social look weak, when the real story is in the assist.
Test ideas before you ship them
Run unlimited A/B tests, attach hypotheses to outcomes, and build a searchable archive of what works — and what doesn't.