AI investments need workflow changes to improve returns

AI spending forecasts raise the stakes for corporate returns

AI can generate financial gains in individual applications while falling short of a company’s investment expectations. Forrester identifies part of the problem as companies investing in AI technology without changing their organizations and work processes alongside it. The gains then remain within individual uses of the technology, limiting their contribution to the financial results the company expected.

Expected spending makes that gap consequential. A new Forrester study forecasts that AI will account for 20% of corporate technology spending in 2027, a fifth of the technology budget. Alongside that spending forecast, The Register reports that more than half of IT executives are expected to miss their return-on-investment targets.

Return targets and operating profit measure different things

The return deadlines can be demanding: in the survey, 45% of AI decision-makers say an investment must produce a positive return within one year to qualify as successful. Return on investment, or ROI, measures the financial return relative to the amount invested. For these decision-makers, generating a benefit is therefore only part of the test; the investment must produce a positive return within the specified year.

Company-wide results present a different measure of financial performance. Only 22% of respondents say AI had a positive effect on their company’s operating profit over the past year, meaning its earnings from operations. This finding concerns the effect on the business’s results, while the one-year requirement concerns the return expected from an investment.

The measures also describe different periods and statuses. A one-year return requirement sets a deadline for success, while the operating-profit finding reports respondents’ experience over the preceding year. The Register’s report that executives are expected to miss ROI targets is a forecast, so it describes a different result from the reported effect on operating profit.

Organizational changes help local gains reach company results

Forrester’s explanation connects disappointing financial results to how companies introduce AI. Companies invest in the technology while leaving their organizations and work processes unchanged, so the surrounding work does not change alongside the deployment. In this pattern, financial gains arise within individual use cases, the particular tasks or applications where companies put AI to work.

Those individual gains explain how an application can produce value while the broader investment falls short of expectations. Forrester identifies the lack of accompanying organizational and process changes as a partial cause of the problem. Its explanation places the way a company organizes and performs work within the same financial question as the technology it buys or develops.

Cheaper models could improve the cost side of AI investment

Alongside changes to how companies work, the choice of AI model affects running costs. Forrester expects companies to increase their use of open AI models, which others can use and adapt, because they are cheaper to operate. The expected shift is a response to the need to contain costs as companies spend more on AI.

That shift would give different model categories different roles. Companies are expected to reserve more expensive, advanced AI models for more critical tasks, while increasing their use of open models elsewhere. The forecast therefore describes a selective allocation of work, with the importance of the task influencing when companies use the more expensive technology.

Main highlights

Plan for rising AI spending: Forrester forecasts AI will account for 20% of corporate technology spending in 2027. Budget owners need to assess that growing commitment against return targets that more than half of IT executives are expected to miss.
Separate investment returns from operating profit: While 45% of AI decision-makers require a positive return within one year, only 22% report a positive effect on company operating profit over the past year. Finance teams need distinct measures for investment performance and company-wide financial impact.
Pair AI deployments with workflow changes: Forrester links disappointing returns partly to unchanged organizations and work processes. Business owners need to connect individual AI applications with process changes that help local gains reach company results.
Match model costs to task requirements: Forrester expects greater use of cheaper open models, with more expensive, advanced models reserved for critical tasks. Technology teams can review model choices against task requirements to contain operating costs.