Salesforce Q2 ‘25 earnings came out last night, and in true Salesforce fashion, the hype is real. Salesforce beat Wall Street estimates and maintained their growth trajectory for this year, resulting in a 4% uplift in share price.
But the main story was AI, or more specifically, Salesforce’s new platform, Agentforce, which was mentioned 39 times throughout the earnings call.
Public companies like Salesforce have to report their earnings once a quarter to show investors exactly how the company is performing. However, the earnings call which is hosted by Salesforce executives such as CEO & Founder Marc Benioff, COO Brian Millham, and Amy Weaver CFO, is there to build hype around Salesforce and attract investors. So, it goes without saying that everything should be taken with a pinch of salt.
Nonetheless, the executive team at Salesforce did a fantastic job at this and they laid out their vision for the next evolution in AI.
The Next Evolution in AI
Salesforce has been drip-feeding its vision for a few months now, initially with the announcement of its new Service Agent product, and subsequently, its Sales Agent product.
Over the past eight years, we’ve moved from the Einstein product released in 2016 that was classed as predictive AI – that is, looking at past trends and predicting future events. Things have moved from Lead Scoring, Opportunity Insights, etc. towards generative AI, which, as we all know, allows us to create text, images, video, or sound from a text prompt.
Salesforce sees the next evolution of AI as autonomous agents. While predictive and generative are great for certain use cases, it’s not exactly the future we’ve been promised with AI. With so much time and effort being invested into AI, the leading tech companies of the world clearly have a vision where our productivity and company output are taken to new heights. While summarizing a Slack conversation is helpful, it’s not going generate the trillions of GDP added to the world’s economy like we’ve been promised.
Salesforce’s Sales & Service Agents are completely autonomous agents that are designed to run 24/7, handling basic queries that can be handed off to a robot.
Initially, the Sales Agent can answer prospects’ basic questions (potentially looking to replace part of a BDR/SDR role), as well as book meetings for a sales rep.
The Service Agent seeks to upgrade current chatbots by allowing humans to have fully formed conversations with them instead of following a pre-designed chat tree that can only help with a pre-determined amount of questions. The Agent is also multi-modal, which allows it to understand text, images, and videos.
What is Agentforce?
So, what exactly is this Agentforce that the Salesforce executive team couldn’t stop talking about on the earnings call? Well, as you might have guessed, it’s the platform that the agents are built on. As CX Today guessed a few months ago, it seems to be based on Airkit.ai, an acquisition that Salesforce made nearly a year ago.
“We’re going to be building custom agents for so many of you as well with Agentforce because it’s a development platform as well as this incredible capability to radically extend your sales and service organizations.” Marc Benioff
So, like much of the Salesforce platform, Salesforce is providing out-of-the-box products such as Sales and Service agents, and then allowing us to build our own custom agents. It’s hard not to get swept up in the excitement, with Marc Benioff stating that “we’re going to make a quantum leap for AI”, with these new breakthroughs.
Benioff also took a big swing at Microsoft, stating that so many customers were disappointed with copilots they bought from Microsoft, “because they’re not getting the accuracy and the response that they want. Microsoft has disappointed so many customers with AI.”
This summarises my thoughts on AI as a whole at the moment (which includes Salesforce). Whilst you can sense that a breakthrough is close, AI chatbots and copilots tend to be cool, but underwhelming.
Benioff goes on to say in the coming months, they’ll be releasing agents for other roles, such as industry-specific agents and health agents.
Salesforce are already teasing some of their major success stories with this product, such as Royal Bank of Canada, ADP, and OpenTable, stating that some of these are achieving 90% case resolution through Agentforce.
Summary
Whether this is the real deal or just another hype cycle for AI is yet to be seen, but the vision and the potential are exciting. Salesforce is unlikely to completely rebrand its vision and platform at Dreamforce, as Data Cloud is still such a core part of AI evolution.
Salesforce made a point to state that the number of Data Cloud customers grew 130% last quarter and will probably be the fastest product to $1B in revenue, as well as the fastest to $5B and $10B.
Salesforce don’t often tease their big announcements for Dreamforce, but this time, it’s different.