Gong.io Ltd., one of the most well-funded startups in the sales software market, today introduced a set of generative artificial intelligence models for its platform.

The models are designed to make sales teams more productive by speeding up tasks such as understanding why a deal has stalled. According to Gong, its AI technology can reduce the amount of time required for some chores by a factor of 20.

San Francisco-based Gong has raised $583 million from investors since launching in 2016. Its namesake platform, which is used by more than 3,500 companies, helps salespeople close more deals. It uses AI to analyze a sales team’s discussions with potential customers and point out areas for improvement.

Gong can, for example, highlight which product features should be emphasized the most during deal negotiations to win over a prospect. It’s also capable of providing other pointers, along with higher-level information such as whether a pricing change decreased or increased market interest in a certain offering.

The AI models detailed by Gong today will expand the number of tasks that its platform can streamline for users. Many of the new models focus on easing information retrieval tasks. In particular, they’re designed to help salespeople more quickly find the deal details they need to understand buyer requirements.

According to Gong, its models can automatically analyze a sales call and generate a summary that highlights key details. Sales managers can then consult the summary to identify the best course of action. For example, if Gong detects that a potential buyer expressed interest in two different offerings from a company’s product portfolio, the sales pitch can be adapted accordingly.

In addition to generating call summaries, the new AI models enable users to retrieve data about deals using natural language queries. A sales manager can, for example, ask whether a customer has requested a product demo in the past week. Gong says the feature significantly reduces the amount of time required to understand the status of an outstanding deal.

Another task that the company’s new AI models promise to speed up is detecting market trends. The models can spot when prospective buyers in a given industry start raising a new concern about a company’s products. Gong could, for example, detect if customers begin mentioning that a competitor offers more favorable pricing. 

The company’s models can organize buyer concerns by frequency as well. Using that information, companies can adjust their sales pitch to more effectively address the issues raised most often by customers. 

To ease the task of adapting a company’s sales pitch, Gong is rolling out a text generation feature called Engage. It automatically creates sales emails and customizes them based on a sales team’s previous interactions with a prospect. According to Gong, the tool can address any concerns that the prospect may have raised during earlier phases of the deal negotiations. 

“We have been working with LLMs for over a year, and see it as a major breakthrough,” said Gong co-founder and Chief Executive Officer Amit Bendov. “But highly accurate, domain-specific AI that delivers a deep understanding of what’s going on within a sales organization and the next steps needed to win deals is more elusive.”

The company says its new AI features are powered by a mix of internally developed and general-purpose language models. According to Gong, its engineers trained the models on tens of billions of sales interactions. It claims its AI software can perform some tasks three times more accurately than off-the-shelf neural networks not optimized as extensively for sales tasks.

Image: Gong