Salesforce has warned commerce leaders chasing AI opportunities that customers are moving faster than many organizations can keep up with.
Its latest State of Commerce report found that 86% of commerce professionals believe AI is raising the bar for customer expectations. Yet 61% say meeting those expectations is harder than ever.
These two stats get to the crux of the situation. The tools promising more personalized, faster, and more autonomous journeys are also making yesterday’s service standards look increasingly outdated.
Caila Schwartz, Head of Agentic Commerce Shopper Insights at Agentforce Commerce, summed up the challenge, stating:
“The rules of commerce are being rewritten in real time.”
For Salesforce, AI is no longer being framed as a future-facing experiment; instead, it is becoming a necessary response to the pressure facing commerce, service, sales, and marketing teams.
AI Agents Move Closer to the Service Front Line
Salesforce surveyed 3,450 commerce professionals across 20 countries and 13 industries between April 10 and June 4, 2026.
The vendor also analyzed B2C and B2B buyer behavior from more than 1.5 billion customers globally.
The findings point to a market already well into the AI transition, with 78% of organizations currently using AI in some form, and nearly half planning to deploy agentic AI within the next six months.
In B2C commerce, autonomous customer service resolution is among the most common AI-agent use cases, alongside AI shopping concierges that guide customers through product discovery and purchase.
This is particularly interesting because the distinction between a commerce journey and a service journey is starting to blur.
A customer looking for a product, checking stock, tracking an order, querying a promotion, or seeking a return is not thinking about the internal handoff between commerce and service; they simply expect an answer.
According to Schwartz, “the rules of commerce are being rewritten in real time.
“For years, commerce leaders have navigated a world of compounding complexity: more channels to manage, more vendors to integrate, more customer expectations to meet. And now, AI has entered that equation on both sides.”
The potential upside is obvious. AI agents could reduce customer effort by resolving routine issues quickly and providing product guidance without making buyers wait for an agent.
But the report also makes clear that AI-led service will only be as dependable as the data behind it.
The Data Problem Is Still Very Much Alive
Only 27% of commerce organizations say their customer data is fully unified across sales, service, marketing, and commerce teams.
For a sector talking increasingly about real-time personalization and autonomous resolution, that is a serious issue.
Among B2C organizations, 46% report duplicate or conflicting customer data, while 45% point to the cost of maintaining disconnected systems. In B2B, 37% say disconnected data slows responses to customer issues or makes them ineffective.
Those gaps risk turning an AI service interaction into a faster way of delivering the wrong answer.
A customer service agent, human or automated, cannot provide a smooth resolution if inventory, order, pricing, promotion, and customer information sit in separate systems.
The result is conflicting answers, unnecessary transfers, repeated explanations, and support teams forced to manually bridge gaps in the tech stack.
More than six in ten respondents also cited poor data integration, a lack of a defined AI strategy, and poor data quality as moderate or major barriers to realizing AI value.
Indeed, just 32% said their organization has fully defined AI success metrics and KPIs.
Schwartz described the tension as “a market at an inflection point,” adding:
“AI is both creating the pressure and offering the solution – but only for those ready to act on it.”
Self-Service Becomes a Bigger B2B CX Test
The report also offers a useful reminder that self-service is not solely a consumer expectation.
More than half of B2B buying tasks are now handled mostly or fully without rep involvement. That includes browsing and configuring products, placing and tracking orders, and managing reorders, each at 54%.
Fully self-service B2B organizations are more likely to say they can meet customer expectations than those that remain mostly rep-assisted, at 82% versus 64%.
That does not mean B2B businesses should remove people from the buying journey. Complex purchases, exceptions, and high-value relationships will continue to require experienced account and service teams.
However, the data does suggest that customers increasingly expect to complete routine tasks without waiting for someone to respond. They want visibility of inventory and pricing, order updates, and relevant recommendations when they need them.
The problem is that many organizations are still struggling to provide a consistent experience across channels. Salesforce identifies inconsistent pricing and promotions, as well as inventory data that does not synchronize in real time, as key omnichannel obstacles.
Those may start as commerce problems, but they quickly become service problems when the customer has to call to find out which price is correct or whether an item is actually available.
Discovery Is Changing Too
With the rise of AI search, customer experience now often begins before a shopper reaches a brand’s site.
Salesforce found that 79% of organizations have seen an increase in traffic attributed to LLM-powered search. It also reported that consumer reliance on AI assistants as the first stop in the shopping journey grew 200% between May 2025 and May 2026.
Meanwhile, 86% of respondents believe LLMs will be essential to product discovery within the next year.
For CX leaders, this means that providing accurate product, policy, pricing, and availability data is becoming essential not only for agents and websites, but also for AI-powered discovery experiences beyond a company’s direct channels.
Salesforce’s findings do not suggest that AI will solve every commerce and service problem overnight. In many ways they suggest something more demanding: AI will make weak customer data, fragmented journeys, and slow service processes harder to hide.
