Why traditional investment models no longer apply?

AI investments diverge significantly from conventional investment in how they generate and demonstrate value. While traditional projects typically follow predictable implementation curves with clear operational impacts, AI initiatives offer a more complex value proposition: benefits evolve over time, are often indirect and resist clear isolation or attribution to the AI investment itself. This fundamental difference requires organizations to reconsider how they approach ROI calculation and justification for AI spend.

Let’s first explore why AI returns tend to accelerate over time rather than follow a steady path, and then examine attribution complexities, one of the most overlooked challenges in assessing AI ROI.

AI investments often yield exponential returns by combining powerful technological drivers with subtle indirect benefits. An IDC’s global study found that for every dollar invested in AI, companies realize an average return of $3.70, with some organizations achieving as much as $10.30(3). 

5 key technological factors drive this outsized ROI:

1. New capabilities: AI unlocks opportunities like predictive models that anticipate market shifts or maintenance needs, novel data monetization strategies, and automated decision support. These new capabilities create revenue streams and efficiencies that were previously impossible with manual methods.
2. Scalability: Once deployed, AI solutions can scale at marginal building cost. For example, an AI customer support agent can handle thousands of inquiries simultaneously without added staff, enabling rapid growth without commensurate cost increases.
3. Model Reuse: An AI model or component can be redeployed across multiple use cases or domains with minimal adjustment, leveraging one investment for many returns. For example, a machine learning model trained to detect anomalies in manufacturing can be adapted to monitor financial transactions for fraud, extending its value across different business units.
4. Compounding improvements: AI systems benefit from feedback loops that drive continuous improvement. The more an AI is used, the more data it gathers to refine its algorithms, creating a self-reinforcing cycle where better performance attracts more usage, and more usage further improves performance. Small gains snowball into big advantages.
5. Surge in performance: Breakthroughs in AI hardware and model architectures generate ripple effects across entire systems. When a core algorithm improves, such as through a more efficient transformer model or optimized cloud inference, every application built on that foundation sees an instant performance boost. This amplification effect turns incremental technical progress into substantial, system-wide gains, accelerating value creation across digital products and platforms.

Beyond these direct technological drivers, AI brings indirect benefits that amplify long-term returns. Employee satisfaction, increased innovation, organizational agility, improved customer experience, and reduced error risk are all potential AI benefits that are difficult to quantify. While these gains may not show up directly on a balance sheet, they strengthen the organization’s competitive advantage and resilience, contributing to a sustained non-linear growth in overall returns. 

One of the most overlooked challenges in AI ROI assessment is attribution. Unlike traditional systems, typically deployed to solve a discrete problem with clearly measurable KPIs, AI initiatives are often part of broader transformation efforts. They intertwine with changes in processes, platforms, organizational structures, and skills. This interdependence makes it extremely difficult to draw a clear causal link between the AI investment and observed business outcomes.

Moreover, AI often acts as an enabler, rather than a direct executor of value. It powers insights, automates decisions, and augments human judgment, yet the final outcome still depends on downstream execution. This “distributed impact” makes attribution inherently messy.

Take, for example, a retailer implementing an AI-powered demand forecasting tool. If this occurs in parallel with supply chain optimization and workforce retraining, improvements in stock availability, reduced markdowns, or increased sales are the result of a systemic shift, not a single intervention. The AI may have been a catalyst, but not the sole driver.

As AI becomes embedded in multi-year digital strategies and cross-functional workflows, isolating its financial contribution becomes less meaningful, and potentially misleading, without a systemic evaluation framework. In this context, AI calls for a new attribution logic: one that positions it as a leverage point within a broader transformation fabric, rather than a stand-alone investment line.

AI ROI is far from one-size-fits-all. At the organization level, factors such as data readiness, employee adoption capacity and process maturity create significant variability in AI ROI across organizations, even when implementing similar use cases. Externally, industry-specific forces such as regulations, ecosystem maturity, automation levels and planning culture, impose global constraints and opportunities. 

Let’s explore why these dynamics, together, shape costs and timelines, complicating the application of standardized ROI assessment frameworks.

The return on AI investments varies between organizations based on several contextual factors, such as:

• Data availability and depth: Organizations with rich, accessible clean data resources generally realize higher returns from AI initiatives. 

→ A bank with years of structured customer data can more easily deploy accurate credit scoring models.

• Employee AI adoption capacity: Workforces with higher digital literacy and openness to new technologies show faster uptake and better use of AI tools.

→ A retail chain that invests in frontline AI training often sees quicker productivity gains from AI-assisted scheduling or inventory tools.

• Process standardization: Companies with well‑defined, repeatable processes can more easily integrate AI agents. 

→ A logistics company with standardized routing procedures can more effectively deploy AI to optimize delivery paths.

• Technological & model synergies: Organizations that maintain shared data assets, infrastructure, and AI models across use cases achieve economies of scale. 

→ A healthcare group reusing a medical language model for both diagnostics and patient communication improves ROI by spreading development costs across multiple applications.

The return on AI investments varies between industries based on contextual factors, such as :

• Regulation & ethics: AI initiatives often carry significant regulatory and ethical implications that directly impact ROI. Compliance costs related to regulations like GDPR or certification requirements such as SOC 2 can substantially affect project profitability, adding complexity layers. In healthcare, deploying a clinical decision support system must comply with strict data privacy laws (e.g., GDPR, HIPAA) and validation protocols, significantly increasing time-to-market and compliance costs. By contrast, industries with lighter regulatory oversight, such as e-commerce, can move faster and with lower risk exposure.
• AI market & ecosystem maturity: In industries underpinned by robust AI ecosystems, where mature solution vendors and scalable cloud-service providers abound, implementation costs are lower and time-to-value is compressed. 

→ Insurers can rapidly adopt fraud detection models pre-trained on industry datasets. By contrast, in agriculture or public infrastructure, where vendor ecosystems are less developed, firms face higher integration costs and longer development cycles.

• Process Automation: Industries with highly automated core processes, back-office functions and operations can embed AI agents into existing workflows more seamlessly, accelerating implementation and enabling organizations to capture value, and demonstrate ROI, much more quickly. 

→ In manufacturing, predictive maintenance tools can be embedded into existing asset management systems, cutting downtime. By contrast, in sectors with less structured workflows, like creative industries, AI integration often requires more upfront customization and organizational change.

• Long-term planning culture: Industries where firms embrace structured, multi-year enterprise planning foster tightly aligned business and IT roadmaps, reduce the risk of mid-course reversals. They move beyond “quick-win” mindsets, creating the stable investment horizon needed for consistent, sustainable ROI from AI initiatives.

→ Energy or Aerospace companies can embed AI into long-range grid optimization strategies with clearly defined KPIs. Industries driven by quarterly performance or short-term ROI expectations may underinvest in foundational AI capabilities that take time to mature.

Because AI returns are driven by its unique characteristics and influenced by organizational maturity and industry dynamics, ROI models must be rethought, not merely to evaluate outcomes, but to inform the business case and strategic investment logic from the start.

In our next article, we’ll unveil a structured evaluation framework that enables organizations to quantify AI’s impact across three interconnected tiers: industry context, implementation costs, and multi-horizon benefits.