AI Measurement Gap: 90% Spend, 12% Prove Revenue – Investor Filter

A new survey from Comviva exposes a structural disconnect in the AI adoption cycle: 90% of organisations increased their AI marketing investments over the past two years, yet only 12% can quantify the revenue those investments generate. For anyone constructing a watchlist of AI-exposed equities, that 78-point spread is not a curiosity. It is a practical risk filter.

The report, titled “The AI Efficiency Divide: Measuring AI’s Real Value Beyond the Hype”, surveyed more than 200 senior IT and business executives across telecommunications, retail, and e-commerce. The core finding is that money flows into AI faster than the tools to measure its return. Investors who treat every company’s “AI tailwind” claim as equal are ignoring a critical distinction: the firms that can prove AI ROI are a small minority. The rest operate on estimates and hope.

The 12% Edge: Why Revenue Attribution Defines AI Stock Risk

The headline number is the gap between spenders and measurers. 90% raised AI marketing investment. 12% can track the revenue it brought. That leaves 78% of organisations funding a technology whose business contribution they cannot isolate.

Rajesh Chandiramani, Chief Executive Officer at Comviva, framed the shift directly.

“AI is rapidly moving from experimentation to enterprise-wide adoption, and the industry is entering a phase where accountability and outcomes will define success.”

For investors, the practical question is whether the companies in a portfolio can answer the board-level challenge the survey describes. 86% of marketing leaders have been asked by their board or senior management to justify AI spending in the past 12 months. Only 16% were confident they could defend their budgets with quantified business value.

A stock whose management cannot articulate AI ROI in concrete terms carries more execution risk than one that can. The 12% minority that can quantify revenue is the group that can sustain – or increase – AI budgets without facing a board backlash. The rest may face budget scrutiny that slows product rollouts and depresses margins.

Risk to watch:

The measurement gap is not temporary. It reflects a deeper problem in how companies budget for AI. When board pressure intensifies, spending cuts could follow across the 78% that cannot prove value.

Cost Fragmentation and the Three Barriers to AI ROI

The survey identified why measuring AI ROI is so hard. 67% of organisations cannot accurately determine the total cost of AI initiatives once infrastructure, talent, and data-related expenses are factored in. 79% rely on estimates rather than precise measurement.

AI spending is scattered across multiple line items: software subscriptions, cloud infrastructure, hardware, talent, and integration costs. That fragmentation makes it difficult for finance teams to build a complete picture of total investment and return.

Three barriers stood out in the survey:

• Cost fragmentation – cited by 62% of respondents as a major challenge
• Attribution – 58% said they struggle to link revenue directly to AI
• Experience-to-revenue disconnect – 55% reported that customer experience improvements do not translate into measurable revenue gains

For holders of cloud infrastructure stocks and AI platform providers, this creates a two-sided dynamic. Demand for compute and software remains high as companies spend to stay competitive. If the measurement problem persists, the next budget cycle could bring a pullback. A company that cannot prove AI ROI is a company that may cut cloud spend or delay new subscriptions.

Practical rule:

When a cloud provider reports slowing growth from enterprise customers, check whether the customer base overlaps with sectors that score low on AI measurement capability. The survey suggests telecom, retail, and e-commerce are still early in the measurement journey.

Which AI Use Cases Deliver Measurable Outcomes?