OperationsKPI / metricModelAccessible

Return on innovation investment (ROI2)

How should return on innovation investment (roi2) be measured and interpreted?

AccessibleStrategicTeam3 min read

Contents

Helps managers answer: To what extent are our investments in innovation generating a return?

Historical cross-industry reporting placed average research, development and innovation spending at 3.5% of revenue, rising to roughly 7% in manufacturing and around 15% in some technology or pharmaceutical businesses. Accounting treatment and economic intent differ: some R&D is expensed, but managers fund it to create future products, capabilities and cash flows. Return on innovation investment (ROI2) asks whether that commitment creates value.

When to use it

Answer the key performance question: “To what extent are our investments in innovation generating a return?”

Include the KPI in the operational processes and supply-chain perspective.

Compare innovation initiatives at decision gates and after launch.

Examine portfolio value alongside learning, option value and strategic capability.

Origins

ROI2 adapts the long-established return-on-investment concept to innovation portfolios. Consulting and innovation-management practice popularised the label as companies sought a comparable financial view of R&D and new-product investment. It has no single standard-setting origin, and definitions vary materially.

What it is

Perspective: Operational processes and supply chain perspective.

Key performance question: To what extent are our investments in innovation generating a return?

ROI2 compares attributable benefits from new products, services or processes with the costs required to develop and commercialise them. A Boston Consulting Group report from 2010 stated that 50% of surveyed managers were dissatisfied with their innovation return. The finding is historical, but the management problem remains: large activity pipelines do not necessarily produce valuable outcomes.

Alexander Kandybin of Booz & Company argued that the method can compare innovation with other investments and contrast small initiatives with large ones. That comparability is useful only when boundaries, timing, risk and attribution are consistent. Early research, platform capabilities and failed experiments may create learning or options not captured in a project’s immediate profit.

How to use it

Measurement

Define the innovation unit, baseline, time horizon and counterfactual. Include research, development, testing, launch, capital, change and support costs as appropriate. Estimate incremental—not total—profit, and state how shared platforms, cannibalisation and benefits after the measurement window are treated.

Data collection method

Combine project accounting with product, service and customer data. Reconcile forecast versions with actual expenditure and contribution. Keep direct evidence separate from modelled attribution.

Formula

One common form is:

ROI2 = [(Net profit from new products and services) – (Innovation costs for these products and services)]/(Innovation costs for these products and services)

If “net profit” already deducts the innovation costs, subtracting them again would double count. Define the numerator as incremental benefit before those specified investment costs, or use an alternative formula whose terms do not overlap.

ROI2 can be retrospective, using realised costs and benefits, or prospective, using probability-weighted scenarios. Forecast results should be reported as a range with assumptions rather than as an achieved return.

Frequency

Measure at decision gates and after enough market evidence exists. Under a simple, constant and non-discounted convention, a 33.3% annual return implies approximately three years of payback![](./assets/premium/return-on-innovation-investment-roi2-01.png); 50% implies two years and 200% six months. Real cash flows rarely arrive evenly, so calculate the actual payback schedule and discounted value when timing matters.

Source of the data

Use innovation-project records, accounting data, product economics, sales data and validated benefit models.

Cost/effort in collecting the data

Effort is moderate when costs and incremental benefits are traceable. It rises when benefits are shared, delayed or forecast, and when the counterfactual is uncertain.

Target setting/benchmarks

Generic benchmarks are weak because innovation projects differ in uncertainty, horizon, strategic purpose and option value. Set stage-appropriate thresholds and evaluate the portfolio, not only successful launches.

Example

If an innovation costs $50,000 and produces $25,000 in annual incremental profit under the chosen definition, the simple calculation is:

The decision still depends on duration, risk, cash-flow timing, ongoing costs and what would have happened without the innovation.

Top practical tip

Publish a metric dictionary with cost boundaries, benefit attribution, time horizon and treatment of failed experiments. For comparability, distinguish annual return from annualised return. Under a simple non-compounded convention, a period from 1 January through 31 December is annual; 1% for one month scales to 12%, and 10% over two years scales to 5%. For investment returns, use a compound annual rate when reinvestment and timing matter.

Top pitfall

Do not optimise the pipeline for retrospective ROI2 alone. It can encourage teams to underinvest in uncertain discovery, attribute all launch profit to the latest project, hide failures and ignore learning, cannibalisation or long-dated platform value.