Salesforce's acquisition of Cimulate to boost AI shopping

Salesforce's planned acquisition of commerce technology startup Cimulate promises to expand Agentforce Commerce's capabilities in a world where search engines are replacing website clicks with AI summaries.

Cimulate includes a basket of large language model (LLM)-fueled AI features, including AI-powered search and browse, a product recommendation engine, AI content optimization, conversational shopping agents and agentic interfaces. Merchandising features also offer analytics insights.

The company's customers include blue-chip retailers such as Pacsun, Boot Barn and CDW. In a press release detailing the sale, Salesforce said Cimulate will provide customers with better "intent-driven" shopping experiences.

Understanding searcher intent is a general benefit of LLMs. Cimulate's CommerceGPT LLM-based operating system incorporates natural language into shopping search and product recommendations. For example, when a customer searches for "blouse," "shirt" or "top," the LLM knows what they're talking about, unlike rules- or keyword-based search engines that CommerceGPT would replace.

Futurum vice president and research director Keith Kirkpatrick said that Cimulate could help Salesforce keep up with evolving commerce experiences. Customers increasingly expect to be able to jump channels -- such as from messaging to web to SMS text -- while researching and buying, and Cimulate technology could fill gaps in the Agentforce Commerce -- formerly Commerce Cloud -- platform.

"It's very much like Buck Rogers, where everything, all the time, is at your fingertips," Kirkpatrick said. "That is how you're going to capture the customer in a world where everything is pretty much undifferentiated in terms of product quality."

Yet even the ideas around how shopping works outside the browser are still being invented. Digital marketing experts can't yet even agree on what to call the function of making something rank well in AI search summaries. Some call it generative engine optimization or generative experience optimization (GEO); others call it AI search optimization (AISO); still others, such as Cimulate, call it Answer Engine Optimization (AEO). Some have different definitions for those terms, while others use them interchangeably.

Whatever terminology the users of marketing and e-commerce technology settle upon, it's still in its infancy -- as marketers rush to find an edge in these new digital spaces. At the same time, AI search summaries gobble up what used to be customer clicks to their websites.

Cimulate will also bring technology to Salesforce that will enable the creation of autonomous shopping agents. While consumers may or may not adopt shopping agents -- opting instead for the dopamine hit of doing their own shopping -- there is an opportunity for shopping agents to automate B2B transactions and ordering, especially for standing or repeat orders of supplies or inventory.

In rules-based inventory management, it's pretty easy for a hardware store to look at, for example, snow shovel inventory. The store could figure out how many it sold last year, look at weather predictions for this year, and guess how many to order before the first snow. Kirkpatrick said that generative AI models will be able to process far more complex signals than rules-based analytics to determine how much shovel inventory to allocate and where to distribute it.

"Obviously, we're beyond [the basics of inventory management], but we're not at that level of 'Let's try to optimize production and supply chain, making sure that we're taking the right route and optimizing fuel,' because that stuff is going to take a bit more time to get right," Kirkpatrick said. "I certainly see agents as being a part of that."

The acquisition of Cimulate is expected to close in the next three months, Salesforce said in a press release. Terms of the deal were not disclosed.

Don Fluckinger is a senior news writer for Informa TechTarget. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.