The “are enterprises really deploying AI agents at scale?” question now has a definitive answer backed by real production data.

Salesforce published the second edition of its Agentic Enterprise Index in August 2026, drawing on aggregated usage data from the Agentforce platform across thousands of businesses plus a proprietary research study of 4,689 respondents from May 2026. The picture it paints is not of careful experimentation. It is of businesses running measurable fleets of AI agents and seeing results that are showing up in their revenue lines.

The Numbers That Matter

The most striking finding is how fast the shift happened. In February 2025, the average Salesforce business had activated five agents. By April 2026, that number was 13. That is a 7% compound monthly growth rate over 15 months — not a jump driven by a few large accounts, but a sustained climb across the platform.

Deployment speed explains some of this acceleration. In February 2025, creating and deploying a new agent took roughly four days on average. By April 2026, that time had dropped by 53% to just 1.9 days. When deploying an agent takes less than two working days, the decision calculus changes. Businesses stop treating agent deployment as a project and start treating it like hiring a new team member.

The agents themselves are also getting more capable. The average number of distinct skills available to each agent rose from two at the beginning of 2025 to six by the end of the year. The average number of actions executed per account grew at a 31% compound monthly growth rate across the same period.

What It Is Actually Doing to Business Results

The headline business impact figure is retail: businesses that deployed agents on their digital commerce channels saw 4x higher online sales growth compared to those that did not. That is not a minor efficiency gain — it is a category-level performance difference.

Customer trust metrics held up alongside growth, which matters. Weekly employee usage of agents tripled across the index period, and customer escalation rates remained stable even as agent interaction volume scaled significantly. The combination of rising usage and stable quality is the pattern that distinguishes real adoption from a novelty spike.

What This Index Actually Represents

It is worth being clear about what the Salesforce Agentic Enterprise Index can and cannot tell you. It draws on data from Agentforce specifically, which means the sample skews toward businesses that have already chosen to invest in an enterprise agent platform. It does not represent the broader population of small businesses or companies still evaluating whether to start.

What it does represent is where committed adopters are right now. And where they are is: running fleets of 13+ agents, deploying new ones in under two days, and seeing measurable differences in business outcomes. The pilot stage ended for this cohort well over a year ago.

What This Means for Business

The data reframes a common conversation we see businesses having about AI agents in 2026. Many companies are still asking “should we deploy AI agents?” as if it is a strategic question that requires months of evaluation. The Salesforce index suggests the more useful question is “how many agents do we need, for which tasks, and what infrastructure do we need to manage them?”

That shift has practical implications.

Deployment time is no longer the constraint. If it takes 1.9 days to deploy a new agent, the bottleneck is not technical. It is deciding where agents should work, what data they should access, and what human oversight looks like. Those are process and governance questions, not technology questions. Recent UK government research — including a documented case of AI agents autonomously targeting real organisations 19 separate times — underlines why governance frameworks belong in the deployment plan, not the post-mortem.

Scale happens faster than expected. Going from five agents to thirteen in 15 months at 7% monthly growth sounds gradual until you realise that at the same rate, you reach 50 agents within another year. Businesses that start now are not late — they are building the operational muscle they will need to manage significantly larger fleets within 18 to 24 months.

Business impact is not uniform. Retail got 4x sales growth. Other sectors will have different baselines and different levers. The index does not tell you which agents to deploy first for your specific business — that still requires analysis of your actual workflows, constraints, and where autonomous action adds the most value without introducing unacceptable risk.

The “should we start?” debate is increasingly a distraction. The data says businesses that started are now running 13 agents and deploying new ones in two days. For businesses still weighing the question, that gap is growing every month.

---

If you are ready to move past the evaluation phase and want to map which agent deployments will drive the clearest ROI for your business, book a session with our Omni Ops team — we work through exactly this with businesses building their first agent fleets.