What I Realized From Watching Salesforce Pitch AI to 50,000 Folks

Pricey subscribers,

Final week, I went to Dreamforce and needed to share my expertise visiting the largest convention in tech.

I spent an entire day watching Marc Benioff pitch AI brokers to 50,000 folks. So listed here are my ideas on:

1. How Salesforce constructed their companion flywheel

2. Marc Benioff’s “Agentic Enterprise” imaginative and prescient

3. High takeaways from periods with Dario (Anthropic) and Sundar (Google)

4. Whether or not Salesforce can efficiently transition to AI

Strolling round Dreamforce was surreal. Salesforce coated your complete Yerba Buena space with synthetic grass, constructed a working waterfall, and turned San Francisco into what felt like a company Disneyland.

However why would 50,000 folks wish to attend this convention?

The reply lies in Salesforce’s companion ecosystem. Right here’s the way it works:

1. Appeal to builders: 6,000+ apps have been constructed on Salesforce’s AppExchange market, together with a number of 7-figure companies.

2. Develop clients: Extra apps convey extra enterprise clients. 89% of Fortune 100 now use AppExchange apps.

3. Construct careers: Salesforce’s free Trailhead coaching program has helped 5M (!) folks land jobs as Salesforce admins and builders.

4. Share income: Companions pay Salesforce round 15% of their AppExchange income in alternate for infrastructure, distribution, and help.

This flywheel is Salesforce’s actual moat however can it thrive within the AI period?

Marc Benioff definitely desires it to. His “agentic enterprise” imaginative and prescient is one the place people work along with AI brokers.

I sat two rows from Marc and he lives as much as his fame because the consummate salesperson. As a substitute of standing on stage like a typical CEO, he walked across the venue with a robotic digital camera following him. He continuously went off script to greet folks within the viewers. He was additionally surprisingly self-deprecating.

I can see why corporations would wish to companion with him. Marc additionally spent a lot of the keynote inviting CEOs from non-tech corporations like Williams Sonoma and Pandora to talk. My takeaway is that:

Salesforce is betting that conventional enterprises will undertake AI by way of their present Salesforce infrastructure moderately than constructing from scratch with AI-native rivals.

It’s a sensible wager given their put in base.

Talking of AI-native rivals, Marc additionally interviewed Dario (CEO of Anthropic) and Sundar (CEO of Google). Listed below are my prime takeaways from every interview:

1. Belief beats options when promoting AI to enterprises. Anthropic is profitable offers by being “the corporate that’s reliable and behaves responsibly.” Enterprises are not looking for flashy demos, they need security and reliability first.

2. Income is at $7B annualized with 10x annual progress. Dario admits that “It could actually’t proceed like this for for much longer. The straightforward 10x progress years are ending, however we’re nonetheless within the explosive part.”

3. AI curing ailments makes Dario probably the most excited. Regardless of the present hype for AI coding, Dario sees healthcare as the actual killer app for AI (he has a biology PhD). “We have a look at these powerful ailments like most cancers, like Alzheimer’s — I simply suppose AI is the fitting factor to treatment them.”

1. Digital tremendous intelligence can be our collaborator. Sundar has an optimistic view that “digital tremendous intelligence can be a collaborator” dealing with advanced duties with people. He doesn’t suppose AGI will exchange people as a result of…

2. We (people) are extra adaptive than we predict. Sundar put his dad in a Waymo self-driving automobile — “it felt like science fiction” to his father however his children “acquired on their telephones and so they had been like no matter.” We continuously underestimate our capacity to adapt.

3. Google Glass may very well be coming again. The primary model failed as a result of the expertise wasn’t there. However Sundar is bullish on the way forward for wearable AI: “Now that you simply really have seamless AI with intuitive interfaces — voice, gestures, imaginative and prescient — Google Glass goes to be again and I feel it’s going to be nice.”

However let’s convey it again to Salesforce…

I feel it’s an open query. Salesforce has a number of tailwinds working in its favor:

1. 89% of Fortune 100 already use Salesforce apps. This large base creates switching prices and provides Salesforce a captive viewers to upsell AI options to.

2. Slack is a pure human and agent collaboration interface. Their stay demo confirmed the brand new Slackbot answering questions and connecting to firm knowledge.

3. Wealthy CRM knowledge from gross sales conversations, help tickets, and extra can energy AI brokers which might be extra personalised for every enterprise.

Nevertheless, there are additionally a number of headwinds:

1. Agentforce AI adoption is rising however nonetheless small. Solely 5% of Salesforce’s 150K clients have adopted Agentforce as of this writing.

2. Salesforce just isn’t an AI-native firm. The corporate has 76K staff and won’t be capable to transfer as quick as AI-native rivals.

Total, this does seem to be a wager the corporate effort for Marc and his workforce, even when the corporate is a $240B CRM juggernaut.

I acquired an opportunity to interview Salesforce VP of Product Dan Fernandez, on the convention about Agentforce Vibes — their new vibe coding software for enterprises.

Yeah you learn that proper, Salesforce is launching a vibe coding software within the extremely aggressive AI coding market.

Naturally, I needed to ask Dan why. His reply:

“Whenever you’re vibe coding a Salesforce app, we do plenty of the infrastructure for you. You don’t must arrange digital machines, arduous drives, or routers — that’s all taken care of.”

It seems that Agentforce Vibes isn’t competing with Cursor or Lovable. It’s particularly made for constructing Salesforce apps by way of:

• Browser-based IDE that launches straight out of your Salesforce interface.

• Semantic catalog that understands your present Salesforce elements and APIs.

• Slack integration coming quickly for team-based growth workflows.

This product matches Salesforce’s broader technique: Leverage their ecosystem benefits moderately than compete head-on with AI-native startups.

It’s all in regards to the companion flywheel that I shared at first of this submit.

I went into Dreamforce skeptical that Salesforce can execute on its AI imaginative and prescient.

However I left extra optimistic after seeing the corporate’s companion flywheel in motion. I feel it’ll all come down as to whether the corporate can transition its ecosystem of 6,000+ apps and 5M+ builders to embrace AI brokers.

The subsequent 12-18 months will inform us if Salesforce’s AI transformation goes to work. It’s going to be an enchanting case examine to observe.