Being first to deploy artificial intelligence agents doesn't guarantee being first to see returns. According to a global survey of 2,025 AI decision-makers, the organizations achieving meaningful return on investment (ROI) within eight months share a common trait: they prioritized preparation over pace.

Why Do Some Companies Hit ROI Faster Than Others?

Salesforce's State of Agentic AI in the Enterprise study reveals a counterintuitive finding. Professional and Business Services, along with Supply Chain and Logistics, were among the slowest sectors to adopt AI agents, yet they achieved meaningful ROI in just 6.5 months. Meanwhile, the High Tech industry, one of the largest deployers of AI agents, took 10.1 months to reach the same milestone. This disconnect between deployment speed and profitability suggests that early adoption alone doesn't drive business value.

The research identifies three critical success factors that separate winners from laggards:

• Data Quality: Clean, accessible data available at the moment an agent needs to act emerged as the top predictor of success, regardless of whether data was fully unified across the entire organization.
• Narrow Scope: Tightly bounded use cases, where AI agents focus on specific, well-defined tasks rather than broad applications, delivered faster and more measurable results.
• Human Oversight: Established escalation paths and governance frameworks built before deployment prevented costly errors and ensured accountability.

Notably, only 31 percent of companies unified their data before launching AI agents. The other 69 percent proceeded with fragmented or partially integrated data sources. However, those who did invest in data preparation beforehand achieved ROI in 7.3 months, compared to 8.8 months for those who addressed data gaps after deployment.

"Every boardroom is asking whether it's moving fast enough. Two years into the agentic shift, the answer from the data is that the advantage was never in starting first; it's in starting deliberately," said Shibani Ahuja, Senior Vice President of Data and AI Strategy at Salesforce.

Shibani Ahuja, Senior Vice President of Data and AI Strategy, Salesforce

What's the Trade-Off Between Speed and Safety?

The research uncovered a genuine tension between launching quickly and building robust safeguards. Companies with lighter governance structures reached positive ROI in 7.2 months, while those with heavier oversight took 9.3 months. However, this speed came with a cost: organizations with below-average governance were nearly twice as likely to discover an agent operating outside its intended parameters only after a consequential error occurred, at 32 percent versus 18 percent for those with above-average governance.

More than a third of respondents whose AI initiatives slowed, stalled, or failed named stronger governance frameworks and escalation protocols as changes they would make in hindsight. The lesson is clear: launching with minimal oversight accelerates time to ROI but increases the risk of discovering problems too late.

How to Build a Foundation for AI Agent Success

• Define Use Cases First: Start with a narrow, well-defined problem that AI can solve independently, rather than attempting broad transformation across multiple departments simultaneously.
• Prepare Data for Agents: Focus on making relevant data accurate, mechanized, and semantically described so agents understand what information means and how to use it, rather than waiting for perfect enterprise-wide data integration.
• Embed AI Into Existing Workflows: Ninety-four percent of deployers reported that embedding AI into core workflows delivered more value than running it as a standalone tool, so integrate agents into the systems employees already use daily.
• Establish Governance Before Launch: Define guardrails, escalation paths, and monitoring frameworks in advance, balancing the need for oversight with the flexibility to learn and iterate.
• Measure ROI Continuously: Track adoption rates, customer satisfaction, and business outcomes from day one, rather than assuming value will be obvious after deployment.

The research shows that among the 30 percent of organizations already running agents in production, deployments achieve an average employee adoption rate of 53 percent and a 29 percent lift in customer satisfaction. Retailers deploying AI agents grew online sales at four times the rate of those that did not, demonstrating that when executed thoughtfully, AI agents deliver tangible business results.

Why Measurement Remains the Biggest Challenge Across Industries

While AI adoption is accelerating across sectors, measurement lags significantly behind. In corporate indirect tax, 65 percent of tax professionals now use generative AI tools like ChatGPT for work, yet only 26 percent measure the resulting ROI. This gap between adoption and measurement represents a critical blind spot for organizations trying to justify AI investments and identify which use cases deliver the most value.

Thomson Reuters' three-year tracking of AI adoption in professional services shows a clear evolution. In 2024, the question was whether to adopt AI at all. By 2025, organizations asked whether they had a strategy. In 2026, the differentiating question became whether companies were measuring what they were doing and going deep enough with their AI initiatives.

Organizations with a formal AI strategy now drive three times the ROI of those adopting tools informally. Yet across the entire profession, 82 percent of organizations either aren't collecting ROI metrics on AI or don't know if they are. This measurement gap is particularly risky in indirect tax, where high transaction volumes, multiple jurisdictions, and tight filing deadlines mean that shallow AI adoption without proper oversight can cascade into significant errors.

The path forward for enterprises is becoming clearer: preparation beats speed, measurement unlocks value, and deliberate strategy separates organizations seeing real returns from those simply checking the AI adoption box. As AI agents move from pilot projects to core business operations, the companies that invested upfront in data readiness, clear governance, and continuous measurement will pull further ahead of those racing to deploy without a plan.