How to Win Deals & Grow a $250B Business | Nikesh Arora, Palo Alto Networks

Summary

Nikesh Arora describes how Palo Alto Networks expanded from a largely firewall-focused company into a broad cybersecurity platform. His central business insight is that enterprise growth depends on selling more products to existing customers: customer acquisition is expensive, so a company with only one product has limited opportunity to grow each customer relationship. Expansion works only if customers trust the company and its products perform well.

Strategy and operating model

• Build around customer expansion: Identify adjacent products that existing customers will trust the company to provide, then use the broader offering to deepen relationships with major enterprise customers.
• Prioritize product quality and leadership: In cybersecurity, customers want the best product, not a second-best alternative. Arora focuses on acquiring or building businesses capable of leading their category.
• Treat M&A as a tool, not a strategy: Acquisitions should support an existing strategic objective—such as entering a category when the internal team is not yet ready—not become an end in themselves.
• Keep a portfolio mindset: Product bets can fail because the market develops differently than expected. Arora says to cut losses when evidence changes rather than continue investing to defend an earlier decision.
• Maintain momentum after acquisition: Acquired founders and teams need room to keep building while learning to use the larger company’s resources—especially its sales organization—to scale their products.

Acquisition playbook

• Target category leaders: Arora generally seeks the number-one or number-two company in a market. He argues that top players tend to capture a disproportionate share of category profits, while buying a weaker competitor leaves the leader in place.
• Evaluate more than apparent market rank: Early leaders can look similar, sometimes with only 30–40 customers each. Assess product-market fit, product direction, business model, execution, market trajectory, team quality, and founder leadership.
• Match the target to the company’s roadmap: Before signing a definitive term sheet, Arora requires the founder to agree on a joint product roadmap. The founder and Palo Alto Networks’ product team have 30 days to work through the plan. If they cannot align, the deal should not proceed.
• Negotiate directly with founders: Arora prefers to agree on the deal with the founder rather than have bankers negotiate between people who will need to work together.
• Use diligence to test alignment and substance: The company has walked away from two deals after deeper discussions revealed likely management conflict or weaker technical substance than initial diligence suggested.
• Keep the team that created the product in charge: Arora says the acquired founder should generally lead the product rather than hand it to an incumbent manager whose prior effort did not produce the same result. The founder and team are part of what the buyer is acquiring.
• Retain key founders: Founders are typically re-vested for three years, ideally four, to preserve continuity while the product is still being developed.
• Integrate without treating the acquired company as conquered: Arora wants acquired teams to feel like first-class members of the larger company. Their job is to motivate and leverage the parent company’s resources—not to assume that everyone in the larger organization is an obstacle.

Examples and lessons

• Opera: Palo Alto Networks expected the market to move toward IP-based segmentation, but it did not. The company cut its losses after recognizing the market was not developing as expected.
• Cloud-security acquisitions: Palo Alto Networks had acquired Evident.io before Arora joined, then bought RedLock, which worked across all three major cloud environments. He describes the earlier acquisition as possibly too early and the later one as further along.
• Google and YouTube: Arora recalls arguing at Google that Netflix’s capabilities could be built internally using YouTube’s platform and engineering resources. He says the “we can build it” argument often underestimates the time needed to hire a team and catch up with a company already ahead.
• Founder-led execution: A small startup team may be far ahead of a large company because it has worked together intensely and moved quickly. Replacing that team with standard corporate processes can erase the advantage the acquisition was meant to buy.
• Price and strategic value: Arora weighs price, development stage, and conviction in the market. He says a strategic asset may justify a premium if the buyer believes it can become worth $5–10 billion to the company over the following five to ten years. He also notes that companies can be much more expensive in a strong funding environment than they were a few years earlier.

Metrics, scale, and timelines

• At Google, Arora says the company grew from about 2,000 to 60,000 employees and from about $2 billion to $63 billion in revenue during his tenure.
• He had been at Palo Alto Networks for eight years at the time of the interview. The company had expanded from a firewall-focused business to more than 70 products; Arora says it had bought 35 companies.
• When he joined, Palo Alto Networks had an approximately $18 billion market capitalization. He initially considered how to grow it to more than $100 billion.
• Arora estimates that a strong technology product often takes four to seven years to build.
• His enterprise-cost insight: Companies often spend 15–18% on product development and 6–10% on administration. He argued that sales, marketing, and customer support account for a large share of remaining costs, and that reducing those costs could make a business much more profitable. These are strategic observations, not reported Palo Alto Networks targets or results.

Sales and customer economics

At Google, Arora sold advertising without controlling either the price—set by advertiser bids—or the volume—determined by search activity.

His sales approach was to work with customers on ROI and conversion economics: estimate the value of a search or lead, determine what the customer could afford to bid, and help them spend up to that level while the economics remained attractive.

He also describes the early friction of selling Google ads in Europe, where customers could only pay using a U.S. credit card in U.S. dollars. The broader lesson was to remove obstacles to adoption and make the system easier to scale.

Cybersecurity and AI

• Arora frames cybersecurity in three broad areas: blocking threats at the perimeter, maintaining security hygiene and analytics inside the environment, and detecting and ejecting attackers who get through.
• More connectivity increases the attack surface, and AI can make it easier to discover vulnerabilities. He sees that as increasing the need for security products and services.
• In response to a newly identified vulnerability, he says Palo Alto Networks checked its own systems, blocked and patched the issue, built a customer patch overnight, and announced it publicly when ready. The example emphasizes rapid operational response and customer communication.
• He describes virtual patching as a way to protect customers around vulnerabilities while they patch underlying systems. He also argues that endpoint inspection can be delivered at low marginal cost, whereas AI services that charge by usage may become expensive at large volumes.
• Arora expects AI to change how companies organize work: tasks now split across specialized teams could be handled by smaller groups supported by capable AI tools. He predicts company structures could look fundamentally different within five years, while noting that AI security products and practices are still developing.
• To learn in a fast-moving field, he says he spends several hours a week with AI founders, talks with roughly 5–15 cybersecurity startups per month, reads, and follows industry discussion online.

Leadership principles

• Focus on execution, not worry: Arora says leaders should do everything they can control, then accept that they cannot control every outcome. In his view, stress itself does not change results; coordinated action does.
• Use the organization as an amplifier: A large field-sales team can help an acquired product reach customers, provided the product leader can work effectively in the larger company.
• Develop a learning loop: Study why prior acquisitions succeeded or failed, identify controllable risks, and adjust the process rather than assuming every deal will work.
• Build a public voice selectively: Arora says he began posting on X after being encouraged to share his own views. He prefers posting when he has something to say rather than forcing a daily schedule.

Presenters and sources

• Nikesh Arora — guest; chairman and CEO of Palo Alto Networks.
• David Senra — host and interviewer.
• People referenced in the discussion include Egon Durban, Mark Andreessen, Ari Emanuel, Brendan Iribe, and Palmer Luckey.