John Somorjai on Salesforce’s Acquisition Strategy in the AI Era

Alex Immerman asks John Somorjai what it is like to work with Salesforce CEO Marc Benioff. Somorjai describes Benioff as curious and forward-looking, combining internal innovation with acquisitions. When Benioff sees the industry changing, he asks the team to find companies that could help Salesforce move with it.

Somorjai says acquisition ideas start with customers: what do they want, and where are Salesforce products missing something? The ventures team also meets thousands of companies and brings information about emerging demand and technology back to the product organization.

The company's internal framework considers strategic and customer fit, whether Salesforce can accelerate and integrate the business, and whether the price can create shareholder value. Technology, talent and cultural fit matter, but he says a reasonable valuation is especially difficult in the current market.

Somorjai describes Salesforce as a trusted foundation for enterprise applications that can be accessed through different interfaces. In a more headless model, he says, customers can interact with the underlying platform through products such as Slack and other AI interfaces.

Immerman asks how Salesforce captures value if AI lets customers accomplish more with fewer seats. Somorjai says the company already uses several pricing models, including consumption-based pricing. He sees a shift toward paying for outcomes such as resolved customer-support cases rather than only paying per user.

He points to Fin's short deployment time as supporting outcome pricing. Salesforce piloted it on its help site and, according to Somorjai, got it running in 14 days. He says it handles over 40% of traffic and resolves 65% without a human; those are the company's reported results.

Immerman suggests that an AI product can capture part of a labor budget while lowering a customer's overall cost. Somorjai cites Qualified's Piper agent as an example of answering website visitors' questions and arranging meetings, tasks associated with sales development.

Salesforce has made large acquisitions such as Slack and MuleSoft as well as buying newer AI companies. Somorjai highlights distribution: the company has about 15,000 sales representatives, so a small startup's product can potentially reach a much larger customer base.

For founders weighing partnership against a sale, he says Salesforce maintains a broad partner ecosystem and sometimes invests in partners. Selling offers the chance to use an existing global sales channel; remaining independent means building that channel and financing growth oneself. He frames the choice around timing, distribution and dilution.

Acquisitions also bring people and entrepreneurial practices into Salesforce. Somorjai says integration teams focus on onboarding and career growth, and that four founders of acquired companies have gone on to become chief product officer.

He gives Vlocity and its founder David Schmaier as an example. Its industry-specific focus helped Salesforce expand its vertical approach: it brought six verticals, and Salesforce subsequently added seven more. Schmaier later held senior product and strategy roles.

Faster software development changes the build-versus-buy calculation but does not eliminate it. Somorjai says product teams examine a three-year roadmap, available expertise and the difficulty of hiring. An acquisition can accelerate delivery, but the entire process through integration can itself take six to nine months.

Listen Labs stood out to him for both technology and ease of use. His team was able to send a survey within ten minutes without training. Customer feedback and renewal prospects remain important parts of diligence, and the team used the product itself to survey customers.

Fin brought capabilities Salesforce felt it could not build quickly enough: an AI-native support product, outcome pricing, quick implementation and an experienced team. Somorjai says the team and its founder Owen were also valuable because they could challenge established practices inside Salesforce.

The original Slack thesis was to create a collaborative interface for Salesforce products and applications. Somorjai says the later rise of agents made that interface even more relevant, although the company did not foresee agents in their current form when it acquired Slack.

Slack was a well-liked product that faced distribution challenges. Salesforce negotiated the acquisition during the pandemic, with much of the diligence conducted remotely. Somorjai says integration took time to find its stride; he reports that Slack revenue has since nearly tripled.

Immerman describes Slack as a record of how people communicate and make decisions, giving AI useful organizational context. Somorjai sees the combination of that context and a conversational interface as an advantage for agents used in everyday work.

The speakers joke about dealmaking in Hawaii. Somorjai says a meeting there between the CEOs helped start the process and establish alignment around the announced Braze acquisition.

Salesforce Ventures has expanded beyond close ecosystem partners, though its underlying goal remains finding companies that can deliver enterprise value. Early investments included partners such as DocuSign and Twilio; later it backed areas including data infrastructure and cybersecurity.

The firm created an AI fund and invested across model providers and infrastructure, including Anthropic, Mistral and Cohere, as well as voice and video companies. Somorjai says the investment mandate now also includes physical AI, robotics and drone delivery.

Somorjai divides his time among finding acquisitions, working with integration teams and overseeing ventures. He is interested in improvements in robot mobility and dexterity and points to portfolio companies across humanoids, robotics software and logistics, including investments outside the United States.

Asked whether Salesforce would invest in a potential competitor, he says it should stay close to important technology even when the eventual relationship is uncertain. Salesforce Ventures has invested in more than 750 companies and acquired 29; most investments therefore do not become acquisitions.

The value Salesforce Ventures offers startups, he says, includes introductions to enterprise customers and operational advice from the company's experience. A robotics investment need not be on the immediate Salesforce product roadmap to benefit from those relationships.

Immerman asks where a new company could challenge Salesforce. Somorjai is skeptical that the many new CRM startups are sufficiently differentiated to become very large, though he allows there may be room at the smaller end of the market. Immerman suggests some could start with small customers and grow alongside them.

Somorjai worries that some venture valuations are unsustainable. He argues that very few companies go public and the pool of potential acquirers becomes small above roughly $10 billion. A high valuation can limit exit options and ultimately hurt employees if the company cannot grow into it.

He distinguishes a handful of exceptional model companies from the much larger number of startups seeking similar valuations. Immerman is more optimistic that the scale of top outcomes has expanded. Somorjai says seeing many unsuccessful companies seeking buyers shapes his caution.

On what can stop a deal, Somorjai starts with preparation. Buyers need accessible financial and customer-contract data, and a focus on annual recurring revenue alone can hide customer attrition. High churn may point to unhappy customers or product-quality problems.

Security can veto an acquisition. Salesforce conducts its own penetration tests, and Somorjai advises founders to test their systems before diligence. Some vulnerabilities can be fixed before signing or closing; in one severe example, testers reached source code and a CEO's laptop, and the deal could not be salvaged.

Immerman closes by asking about Somorjai's father, a Holocaust survivor and chemist. Somorjai recalls growing up around scientists, dinner-table conversations and a father who taught constantly. He credits that environment with encouraging curiosity and attention to the future.