SMBs must tie AI costs to outcomes to achieve real ROI

IBM says the AI industry is facing a "cost reckoning," as 79% of executives expect AI to drive major revenue by 2030 but only 24% know where that revenue will come from.

Categorized in: AI News Management
Published on: Sep 05, 2026
SMBs must tie AI costs to outcomes to achieve real ROI

IBM's Neil Dhar said the AI industry is entering a "cost reckoning," as most businesses still lack the financial models to accurately measure what AI costs them. For small and midsize businesses, the path to real returns is not cutting AI adoption but tying every project to specific, measurable business outcomes from the start.

The disconnect between AI spend and business results

Companies track AI license costs, but few have a reliable framework for attributing token consumption, compute, or spend to actual outcomes. The gap is stark: 79% of executives expect AI to drive significant revenue by 2030, yet only 24% know where that revenue will come from, according to "The Enterprise in 2030" report from the IBM Institute for Business Value.

When AI is layered on top of existing workflows as a general productivity tool, token spend has no anchor. Embedded in specific processes tied to defined outcomes, the value becomes traceable. SMB leaders should measure AI's return like any other capital allocation: through time saved, better customer and employee experiences, or new revenue. If those markers don't move, the investment may not be worth it.

Fine-tuning AI to the company's purpose

Any business can adopt a plug-and-play agent or off-the-shelf AI. That doesn't mean they should. Fine-tuning models, agents, and data to meet the specific needs of a company's core mission helps SMBs stand apart. Agents and models should work in lockstep with the company's purpose and augment people to advance product innovations, improve efficiency, and enhance value to customers. Placing measurable targets on those outcomes shows where AI has moved the needle or where teams can gain ground faster.

Managing token costs is an essential piece of the ROI puzzle. Achieving savings in any area - through economies of scale or by lowering specific budget lines - improves overall ROI and time to value. For professionals looking to build these skills, AI for Management training covers the frameworks needed to connect technology investments to business outcomes.

How metered pricing changes the equation

In June, Microsoft announced metered Copilot Cowork pricing, which offers SMBs a way to optimize AI costs while unlocking value. Copilot Cowork provided a lower-cost alternative from its inception, with a runtime that finds the right information and tools, model choices that match the right task, and billing that charges businesses only for what they use.

Under the new structure, Copilot Cowork 1 is expected to handle tasks at substantially lower costs than competitors, with variable pricing models for different use cases. It requires a Microsoft 365 Copilot User Subscription License and bills customers through Copilot Credits on a usage-based model. The price of each task is calculated from four inputs: model use, context retrieval, tool calls, and runtime. Organizations deploying these tools can benefit from Microsoft AI Courses to ensure teams understand the cost levers and model choices available.

Building a budget model around task types

Microsoft has defined Copilot Cowork task types to help businesses improve AI budgeting. Light, medium, and heavy tasks have been assigned estimated Copilot credits - 100-300, 300-700, or more than 700, respectively. Combined with four typical user personas - corporate knowledge workers, management and senior leaders, customer-facing knowledge workers, or technical workers - applying estimated prices per prompt helps SMBs build a flexible cost model. They can then estimate and refine expected AI costs over time.

Why this matters for management

As SMBs accelerate AI adoption, the work of defining precisely where and how AI will add value determines whether an investment delivers returns. The businesses that connect token spend to measurable outcomes - revenue growth, time saved, customer retention - will be the ones that avoid the cost reckoning Dhar described. For managers, the immediate step is auditing current AI usage against defined business targets and shifting any untethered spending toward specific, measured processes.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)