Business ROI Guide 2026: Calculate Return on Investment

The ROI Formula

Example: $20K marketing spend generates $60K profit: (60-20)/20 = 200% ROI.

Annualized: 200% over 2 years: (3)^(1/2) - 1 = 73% per year - the honest comparison number.

ROI ignores timing — annualization fixes it. $10K growing to $14K in 6 months is 40% ROI but roughly 96% annualized; the same $10K to $14K over 3 years is 40% ROI but only 11.9% annualized. The original example's 200%-over-2-years = 73%/year is exactly this correction in action. For anything under 12 months, always annualize before comparing: a 25% ROI in 3 months (~144% annualized) crushes a 60% ROI over 2 years (26.5% annualized), and the raw percentages completely hide that ranking.

Negative ROI needs the same rigor. A $50K campaign returning $35K in year-one cash is −30% ROI — but if it acquired 200 customers worth $150 in lifetime value each ($30K of future value), true ROI is ($35K + $30K − $50K) ÷ $50K = +30%. The classic mistake is stopping at first-year cash: include the asset created (customers, equipment, brand equity) in "gain." Conversely, a project showing +40% ROI that consumed 6 months of your best engineer's time at $180K loaded cost is really deeply negative once labor is included.

Marketing ROI Deep Dive

$5K ad spend, $25K revenue, $10K COGS:

• ROAS: 25/5 = 5:1
• True ROI: (25 - 10 - 5)/5 = 200%

ROAS ignores COGS - a 4:1 ROAS can still lose money on thin margins. Always compute true ROI with all costs.

Attribution windows change the number without changing reality. A campaign showing 5:1 ROAS on 7-day click attribution often drops to ~3.2:1 on 28-day view — same spend, different truth. Since iOS 14.5+, modeled and underreported conversions mean ad platforms undercount results by 15-30%; calibrate with a holdout test (pause spend in two matched geos for 4 weeks and measure the sales delta). True incrementality beats reported ROAS every time: a "4:1 ROAS" campaign where 60% of buyers would have purchased anyway is really 1.6:1 incremental — barely above breakeven.

Use MER (Marketing Efficiency Ratio = Total Revenue ÷ Total Marketing Spend) as the guardrail metric: $2M revenue on $500K spend = 4.0 MER. It can't be gamed by attribution settings and captures halo effects across channels. Working rule: if MER holds above 4.0 while you scale spend 20%/month, keep scaling; if MER decays below 3.0, you're buying diminishing returns — cut the worst channel and retest. Pair MER with new-customer CAC ($500K ÷ 4,000 new customers = $125) to keep growth efficient, not just revenue-flattering.

Equipment Payback Example

$80K machine, saves $30K/yr labor:

• Simple ROI (year 1): (30-80)/80 = -62.5% (meaningless short-term)
• Payback period: 80/30 = 2.67 years
• 5-year ROI: (150-80)/80 = 87.5%
• Annualized: (1.875)^(1/5) - 1 = 13.4%/yr

For multi-year investments, payback period + annualized ROI beat single-year ROI.

Payback ignores the time value of money — NPV fixes it. The $80K machine saving $30K/year for 5 years at an 8% discount rate: $30K × 3.993 (annuity factor) = $119,790 present value − $80K = +$39,790 NPV. Positive means buy. But add $4K/year maintenance and the real savings are $26K/year: $26K × 3.993 = $103,818 − $80K = +$23,818 — still positive but 40% thinner. Any equipment ROI that omits maintenance, energy, and downtime is fiction; model all three before signing the purchase order.

Tax treatment can flip the decision outright. Section 179 lets many businesses expense the full $80K in year one — at a 25% combined tax rate that's a $20K year-one tax shield, cutting effective cost to $60K and payback to 2.0 years ($60K ÷ $30K). Without 179, 5-year MACRS depreciation dribbles the shield out and payback stays near 2.7 years. Run both versions: the after-tax ROI is the only one that matters, and it can run 30-50% better than the pre-tax headline suggests.

ROI Decision Framework

1. Hurdle rate: Set a minimum (e.g., 15%) - reject below it.
2. Risk-adjust: A risky 30% ROI may lose to a safe 15%.
3. Include ALL costs: Time, training, disruption, opportunity cost.
4. Time-bound it: ROI without a timeframe is meaningless.
5. Compare alternatives: Best ROI among options wins, not just positive ROI.

Opportunity cost is the comparison ROI hides. Project A returns 40% on $100K ($40K profit) while Project B returns 25% on $400K ($100K profit) — B's lower percentage makes $60K more absolute profit. With limited capital, rank by profit dollars per constrained resource; with limited management bandwidth, rank by profit per executive-hour. A "great 60% ROI" project that consumes your only senior engineer for a year usually loses to a "boring 25%" project that runs itself — the engineer's time is the real constraint, not the cash.

Set hurdle rates by risk tier, not one flat number: safe cost-savings projects 10-12%, core business expansion 15-20%, new-market bets 25-30%+, speculative R&D 40%+. Reject anything below its tier's hurdle regardless of positive ROI — a 12% return on a speculative bet is a failure, not a win. And treat sunk costs as zero: $30K already spent on a struggling project doesn't raise its hurdle; the decision is forward ROI only. Write the hurdle into the proposal template so every pitch arrives pre-scored.

Frequently Asked Questions