AI shopping searches surged 200% in one year - and it's a top priority for commerce leaders now

Follow ZDNET: Add us as a preferred source on Google. ZDNET's key takeaways 86% of commerce leaders say AI is raising expectations. Agentic AI use in shopping grew 200% year over year. Only 27% of companies have fully unified customer data.The use of agentic AI search as the first step in shopping online grew by 200% in the last 12 months, according to Salesforce's State of Commerce report, a survey of 3,450 commerce professionals and 4,690 consumers, plus behavioral data from more than 1.5 billion global shoppers. AI is raising the bar for customer expectations, making the expansion of AI for e-commerce the No. 1 priority for commerce leaders. Here are the key takeaways from the report: AI and data are the top priorities for commerce leadersThe top priority for commerce leaders is the implementation and expansion of artificial intelligence capabilities. The second-highest priority is improving data quality, accessibility, and management, followed by automating and streamlining processes. Eighty-six percent of commerce professionals believe that AI is raising the bar for customer experiences, and 61% say that meeting rising expectations is harder than ever. Also: Google says AI agents spending your money is a 'more fun' way to shopThe top challenge for commerce leaders is the implementation and expansion of AI capabilities -- a top opportunity and challenge. Managing supply chain challenges, fragmented data across systems, evolving privacy and data regulations, and staffing gaps are also top challenges. Scaling AI is the main focus for commerce leadersBusinesses are no longer interested in AI pilots, so 2026 is the year to focus on operational scale over experiments. Thirty-five percent of agentic AI users are now focused on agentic expansion and are no longer working on proofs of concept. The report found that 71% of commerce leaders agree that AI is a core part of commerce operations. The top five use cases for agentic AI in commerce include autonomous customer service resolution, AI shopping concierges, autonomous replenishment purchasing, autonomous fraud decision-making and intervention, and autonomous merchandising optimization. The challenge is not just scaling the adoption of AI agents but also knowing whether they're working as expected. Also: How to shop with AI: 6 ways I find deals, price track, and let agents buy for meMore than 6 in 10 commerce leaders cite poor data integration, a lack of a defined AI strategy, and poor data quality as barriers to scaling AI. Security and trust, a lack of AI skills within organizations, and change management are also barriers. Commerce leaders are seeing major improvements from AI agents. The top three outcomes for business-to-business organizations using AI agents are revenue growth, employee productivity, and organizational efficiency. Other benefits include customer satisfaction, customer loyalty, and hyper-personalization of offerings. A surprising part of the report is that only 32% of commerce leaders say their organizations have fully defined AI success metrics and key performance indicators. The lack of performance metrics is not slowing down self-service and automation for B2B commerce organizations. Nearly 7 out of 10 report positive benefits from automation, and more than half of B2B buying tasks -- browsing and configuring products, placing and tracking orders, managing reorders, and accessing contract pricing -- are handled mostly with AI agents. Agentic AI adoption success requires removing data silos and a system integration strategyCommerce leaders are expanding their vendor ecosystems to address fragmented data and disconnected systems. Only 29% of commerce leaders find that their commerce platforms meet only basic needs. To expand their capabilities, commerce leaders are looking to add AI-powered personalization (46%), order management (40%), and content management (39%). Only 27% of commerce organizations say their customer data is fully unified across sales, service, marketing, and commerce teams; this is a major obstacle to scaling AI agents. The challenges caused by disconnected data include slow or ineffective responses to customer issues, duplicate or conflicting customer records, difficulty measuring the impact of marketing or commerce investments, fragmented customer experiences, the high cost of maintaining disconnected systems, and missed upsell or cross-sell opportunities. Also: How small businesses can survive AI shopping: 7 essential stepsTrapped data results in trapped value-creation opportunities for commerce organizations. The cost of disconnection is measurable, but so are the payoffs of unification. More than 4 out of 10 commerce professionals cite the ability to better leverage data and AI as a primary driver of their platform strategy, followed by improving integration between core systems. The top five outcomes realized from unifying customer data are better alignment between sales, marketing, and commerce teams; improved outcomes from AI or automation; improved customer retention and loyalty; more effective personalization across channels; increased conversion rates; and improved issue resolution. AI-influenced shopping behavior is outpacing organizations' ability to keep upCustomers are using AI to discover and research products before making purchasing decisions. They expect the same level of personalization online as they do in stores, and 84% of B2B buyers agree that customers expect the same level of personalization online and from their account teams. The most common omnichannel failures are inconsistent pricing and promotions across channels, inventory data that is not synchronized across multiple systems, ordering channels that are siloed across multiple systems, cross-channel transactions that are not well supported, and customer information that is not recognized across channels. All these points of failure lead to poor customer and employee experiences.Also: Why Amazon really doesn't ...

Indexed and credited by AIPROPX. Originating outlet: ZDNet. Open at source →