Measuring AI ROI: Companies Struggle to Quantify Gains, New Study Offers Three Paths
When a firm pours money into artificial intelligence, the promise is clear: higher profits, sharper efficiencies, or a competitive edge. Yet most executives are still guessing what the numbers look like. A fresh study—built on interviews with more than 30 CEOs and senior leaders—says the answer isn’t obvious and offers three concrete ways to turn AI projects into real dollars.
The data paints a confusing picture. Industry surveys and benchmarks often tell companies to "invest, experiment, build capabilities," but they rarely explain how to turn those inputs into measurable outputs. The study confirms that a standard method for measuring AI return on investment (ROI) is still missing. In fact, two firms that invested almost identically in AI defined success in entirely different ways, underscoring how subjective the metric can be. When generic AI tools are deployed without a clear measurement framework, lasting returns rarely materialize.
ROI also hinges on the type of AI technology. Analytical AI—predominantly machine‑learning models that predict and optimize—tends to deliver directly attributable financial gains, but it’s usually applied to narrowly defined use cases. Generative AI, on the other hand, is a broad‑stroke tool that can accelerate knowledge work, improve quality, and boost volume. Turning those gains into dollars, however, requires deliberate effort and a disciplined measurement plan.
Industry context shapes expectations. In the consumer‑goods sector, analytical AI is often used to streamline supply chains and sharpen demand responsiveness. A B2B marketing agency might instead lean on generative AI to boost creative throughput, raise proposal win rates, or improve lead conversion metrics—each a distinct definition of "return." The lesson? The same technology can mean different things in different rooms.
From these insights, the study lays out three practical pathways that match a company’s maturity level and strategic intent:
1. Input‑heavy experimentation – For firms still in the discovery phase, this approach focuses on building capabilities and testing a wide range of pilots. 2. Outcome‑focused pilots – Middle‑stage companies can hone in on specific, high‑impact use cases and track tangible results. 3. Integrated business‑process alignment – Mature organizations can embed AI into core processes, turning technology into a strategic asset that drives consistent, measurable value.
Financial discipline emerges as a recurring theme. Executives noted that, unlike traditional capital projects, AI initiatives often lack a clear cost‑benefit model. The research recommends treating AI as a strategic asset: assign budgets, set performance targets, and link outcomes to executive compensation where appropriate. By doing so, firms create accountability and ensure that AI investments move beyond experiments into revenue‑driving initiatives.
The takeaway is simple but powerful: without a clear, explicit framework, many organizations miss credible, lasting gains from AI. Adopting one of the three recommended pathways lets companies align investments with measurable outcomes, track progress over time, and justify future spending. For leaders who want to turn AI projects into proven business value, this roadmap offers a clearer path to success.
