Salesforce unveils AI agents for sales teams – here's how they help

As part of its Sales Cloud software platform, the company announced Einstein SDR Agent, which will “autonomously engage with inbound prospects to nurture pipeline 24/7,” said Salesforce. 

Also: Sales professionals spend 70% of their time on non-selling tasks – AI can help

An “SDR” — or, sales development representative — is the role of someone in sales who qualifies leads, before the main salesperson steps in to nurture those leads. 

“No sales team has enough pipeline,” said Ketan Karkhanis, Sales Cloud general manager, in a media briefing. “Everyone needs more tools to qualify the pipeline and engage with the right customers at the right time.”

The second AI helper is Einstein Sales Coach Agent, a video likeness of a person that can guide salespeople through things such as rehearsing their sales pitch. 

“Today, they do this in front of a mirror,” said Karkhanis of salespeople rehearsing their pitches. 

“Imagine you are working on a deal, the coach pops up and lets you role-play,” he explained. Following the mock encounter, the Coach Agent provides feedback to the salesperson regarding areas to work on.

The two agents are characterized by Salesforce not as a replacement for salespeople but as “your new team member,” something that will “get my job done faster,” said Karkhanis.

Pricing has not yet been revealed for the two products but will be announced “soon,” said Karkhanis. Salesforce is holding its annual Dreamforce user conference in San Francisco from September 17 to 19. 

Also: What Gartner’s 2024 hype cycle forecast tells us about the future of AI (and other tech)

The two new offerings go beyond agentic technologies, which are increasingly popular in AI applications for enterprises, claimed Karkhanis. Unlike agents in their typical form, he claimed, the SDR rep and the coach “are outcome-focused — the SDR gets you meetings, the coach helps you get better at your pitch.” 

The early adopters are likely to be tech firms, said Karkhanis, based on early feedback that Salesforce has. They are attracted to the SDR Agent by the prospect that “the whole sales cycle could be potentially done by the sales agent, right down to showing a demo link,” thereby automating much of the prospecting work. 

The appeal of automating roles in sales may resonate with smaller firms that have fewer human resources. “I’m surprised that we are getting even more traction from mid-market and lower toward this technology,” he said. “It’s not just large enterprises.”

The SDR Agent can do things such as engage in a chat with prospects to answer initial questions, and then schedule a meeting with the prospect by having access to the salesperson’s calendar. The company emphasizes that the bot is indicated in its interactions as automation, so that prospects know they are interacting with software, not a person, “for full transparency.”

The Sales Coach Agent will do things beyond role-playing, such as popping up suggestions in online meetings with a prospect. If a prospect mentions another customer, the software will bring up competitive data, for example, such as where the other company’s product lacks features. 

Also: Microsoft launches two new Copilots, adding AI-guidance for Service and Sales

The coach has attracted particular attention from chief revenue officers, said Karkhanis. “Sales enablement is a hot topic,” he observed. The “biggest challenge they [CROs] have is onboarding new people, and getting them productive very fast.” The coach technology is “likely to be deployed around new hires sooner,” he said. “Ten new people will get a coach by default.” 

There is “significant energy in that area, significant pull.”

Under the hood, Karkhanis said, the most important resource at Salesforce is the company’s Data Cloud software, which allows for a more sophisticated use of what’s called retrieval-augmented generation (RAG). 

The RAG method is an increasingly popular approach to generative artificial intelligence whereby a large language model (LLM)  is connected to an external source of truth in order to shape the answers to queries. “We can do RAG like no one else,” said Karkhanis.

Also: Make room for RAG: How Gen AI’s balance of power is shifting

Additional technologies such as Salesforce’s Agent Force provide for various LLMs and prompts that inform the products, he said. 

Asked which neural nets the programs use, Karkhanis replied that Salesforce is pursuing an “open ecosystem model,” with technology from OpenAI, Anthropic, and others. There are also internally developed AI models at Salesforce such as its XGen Sales. 

Setting up the SDR agent involves a salesperson creating a new instance of the bot in the software interface, assigning it permissions, personalizing it with one of various avatars that are offered, and choosing which leads to assign the agent. Bounds can be set, such as “all inbound leads under $10,000 in prospective value,” said Karkhanis.

An SDR Agent doesn’t need its own license for email and the like. “Technically, it is a user in Salesforce, you need compliance, etc. But our intention is that it is modeled around outcomes versus users.”