Live by the cloud, die by the cloud.
Salesforce entered the year on a high, with a rising stock and heaps of positive sentiment about what was around the corner. But Q2 brought a rough earnings call, which led to the stock plummeting 20% and whispers about the beginning of the end.
A pioneer of the cloud era that became one of its biggest winners, Salesforce is now grappling with the beginning of a new era – the era of AI – and questions about its ability to maintain its dominant perch. How does a company growing just 8% YoY – akin to running backward in tech – compete with the radical startups promising a whole different future?
Salesforce was already dealing with all that as execs prepared for this week’s Dreamforce. And then came the Klarna bombshell: the Swedish fintech let the world know it no longer needed Salesforce; instead, it was going the DIY AI route.
It felt like an “Emperor’s New Clothes” moment: how many firms would follow suit, and reconsider what was considered a sacrosanct part of the tech stack?
Enter Agentforce
Salesforce needed its leader to do something big, and Marc Benioff answered the call. He announced that the company was doing a “hard pivot” and betting the farm on Agentforce, its suite of autonomous AI agents. HubSpot did the same at their Inbound conference this week. The two largest GTM tech companies in the world are betting on AI Agents.
They are encouraged by the fact that we now have larger multi-modal AI models with PHD level reasoning, allowing us to create AI agents - pieces of software that can achieve complex, multi-step goals requiring various tools. The types of things that humans have traditionally outperformed machines in.
Both companies have broad platforms that their customers use to enable and track multiple customer touchpoints, and this gives them valuable context with which to equip these AI agents. This also minimizes the hallucinations that currently make AI adoption difficult for the enterprise.
The demos were definitely interesting, but it’s time to dream a bit. Jensen Huang, Nvidia’s leather-jacketed luminary, claims that AI is moving at “Moore’s law squared.” So what could this mean for AI Agents in 12-24 months?
“Hi, my name is…”
Until recently, I assumed all my future colleagues would be human. Now, I’m pretty confident that the most successful companies in the future will be a blend of humans and AI Agents. Daunting, yet exciting.
The top expense for a tech company is human capital. It’s also one of its chief limiting factors, as attracting, hiring, and retaining elite talent is a hard problem that gets harder at scale.
Every founder I know is always deciding which projects to prioritize and which tradeoffs to live with, and the limiting factor is generally the stable of elite individuals available to them. AI Agents will mean far fewer tradeoffs.
They will still have to make tradeoffs but of a different sort. Maybe the CAC a company can tolerate is very different when the technical support department is run by AI Agents. Or when a company can provide personalized customer success to every single customer, even those paying only $100 a month.
Now, it’s not certain that Salesforce and HubSpot will be winners here. They have the advantage of being broad platforms, with data across various customer touch points. They also have the advantage of having a large customer base and expertise in distribution.
However, our ecosystem is built on the back of companies we don’t know today, who are going to shape the future. A platform shift is a great birthing ground for these startups. They have the advantage of baggage-free first-principles thinking. They don’t have the customers or their data today, but they will have AI Agents to help bridge any gap.
It actually doesn’t matter who wins the AI arms race - Salesforce, HubSpot, or a company we’ve never heard of. What matters is for us to understand how to use AI Agents to do more and figure out how best to deploy capital in this new world. All while holding on to the humanity in our organizations.
