Databricks Co-Founder Reveals Essential Strategies for Startups to Conquer Enterprise Sales

Startups today face immense pressure to achieve rapid growth, a trajectory often dependent on securing contracts with large enterprises – the deep pockets of the IT world. These organizations offer not only significant revenue streams but also invaluable market validation and long-term stability. However, as any company that has attempted to penetrate this highly competitive segment will attest, winning such high-stakes enterprise contracts is notoriously challenging. The sales cycles are protracted, procurement processes are intricate, and the inherent risk aversion of large corporations means that a startup’s offering must be exceptionally compelling and secure.

A critical mistake many younger startups make is the misguided belief that their cutting-edge technology will, by its sheer brilliance, practically sell itself, warns Arsalan Tavakoli-Shiraji, co-founder and senior vice president of field engineering at Databricks. Databricks, which helps businesses store and analyze vast amounts of data, is now an AI and big data storage giant, boasting an impressive valuation of $190 billion and counting over 70% of the Fortune 500 companies among its customer base. Its platform, known for pioneering the data lakehouse architecture, enables businesses to unify their data, analytics, and AI workloads, addressing complex data management needs at scale. Yet, this formidable presence too began as a tiny startup, with Tavakoli-Shiraji playing a pivotal role in securing its foundational enterprise deals.

While acknowledging that excellent technology is a non-negotiable prerequisite, Tavakoli-Shiraji emphatically emphasized at TechCrunch Disrupt last week that simply having a superior product is "not enough." He learned through direct experience that in the complex world of enterprise sales, "people and processes matter the most." This insight underscores that beyond technical specifications, building trust, fostering relationships, understanding intricate organizational dynamics, and navigating established internal procedures are the true determinants of success. He recalled a particularly disheartening early meeting with a high-value prospect who had no interest in listening because he had "bet his career" on signing a mega-deal with Databricks’ primary competitor. Furthermore, other promising early deals ultimately faltered not due to technical shortcomings, but because the nascent startup was unprepared for the rigorous legal, contractual, or security questions and requirements that are standard in enterprise procurement, including stringent data privacy regulations, service level agreements (SLAs), and comprehensive security audits.

Drawing from these profound experiences, Tavakoli-Shiraji presented his top strategies, offering a practical roadmap for startups aiming to secure lucrative enterprise contracts and accelerate their growth trajectory:

1. Sell Under a Bigger Tech Company’s Umbrella

One of the most impactful strategies Databricks employed in its nascent stages was leveraging the established credibility and expansive reach of a larger technology partner. Tavakoli-Shiraji recounted how Databricks initially struggled to even secure a meeting with UBS, the large Swiss banking giant, due to its lack of established reputation.

The turning point came when Databricks formed a strategic "partnership with Microsoft," becoming an Azure cloud partner. This alliance offered several critical advantages. Firstly, it meant that sales agreements for Databricks’ services could be covered under Microsoft’s expansive, pre-approved legal and contractual terms, instantly alleviating many of the security and compliance concerns that typically derail startup deals. More importantly, Microsoft’s formidable sales force actively stepped in, booking meetings and advocating for Databricks. "All of a sudden, that meeting wasn’t hard to get," Tavakoli-Shiraji recalled, emphasizing the immediate access and trust that Microsoft’s brand conferred. This partnership proved to be "a huge boon to us that we were able to partner with somebody who had that access and distribution."

With Microsoft’s powerful sales team acting as an accelerant, Databricks found itself "getting pulled everywhere" into new enterprise opportunities. While this rapid expansion brought its own set of "good problems," such as prospects requesting features for countries where Databricks had not yet established a presence, these challenges were indicative of genuine market demand and a clear path to growth. This co-selling model allowed Databricks to rapidly scale its customer base and validate its technology within the enterprise ecosystem, significantly de-risking its offering for potential clients.

Today, Databricks has grown into the very "big company" that can offer a similar umbrella to other promising startups. Through its robust partner ecosystem and strategic investments (like Databricks Ventures), it now enables emerging companies to sell their products under Databricks’ own established terms and comprehensive cybersecurity framework. Notable examples include Hightouch, which provides AI-powered marketing and customer-data tools (and in which Databricks has invested); Addepar, offering data tools for investment professionals; and Tecton, a machine learning startup Databricks later acquired, integrating its capabilities into its own platform. This full-circle journey exemplifies the power of strategic partnerships in overcoming early market entry barriers and accelerating enterprise adoption.

2. When Selling AI, Address the "Four Cs"

In the burgeoning and often complex field of Artificial Intelligence, enterprises harbor a distinct set of concerns that startups must proactively address. Tavakoli-Shiraji neatly categorizes these primary objections into what he calls the "four Cs": Context, Choice, Cost, and Control.

• Context: Enterprises require an AI solution that can be seamlessly integrated into their specific operational environment, much "like bringing on a new employee." This means the AI must be capable of learning and adapting to the company’s unique terminology, internal processes, strategic priorities, and specific data landscapes. A generic AI solution that fails to understand the nuances of an enterprise’s business will quickly prove ineffective and generate irrelevant outputs, leading to frustration and disadoption.
• Choice: Vendor lock-in is a significant apprehension for large organizations. They demand the freedom to switch between different AI models, frameworks, or even foundational providers without incurring prohibitive costs or operational disruptions. Startups must demonstrate that their solutions offer interoperability and flexibility, ensuring that customers are not irrevocably tied to a single vendor, thus preserving their long-term strategic agility and negotiating power in a rapidly evolving technological landscape.
• Cost: Beyond the initial purchase price, enterprises are acutely focused on the total cost of ownership (TCO) for AI products. This includes not only the running costs associated with computation and data processing but also the expenses related to implementation, integration, maintenance, and the potential for scaling. Startups must provide clear, predictable, and transparent cost structures, demonstrating how their solution offers superior value and efficiency compared to alternatives, and provide a clear return on investment (ROI).
• Control: Paramount among enterprise concerns is the need for robust control over the AI’s behavior, data security, and ethical considerations. This involves implementing stringent safeguards that govern the AI’s decision-making processes, ensuring compliance with regulatory requirements (such as GDPR or HIPAA), and, critically, protecting sensitive corporate and customer information from breaches or misuse. Enterprises need absolute assurance that they retain full governance over their proprietary data and the AI models operating on it.

By directly addressing these four fundamental concerns, startups can build trust, mitigate perceived risks, and position their AI solutions as viable, responsible, and strategically sound investments for large organizations.

3. Attack the Expensive Legacy Systems Enterprises Think They’re Stuck With

Tavakoli-Shiraji observes that many CIOs tend to overstate their company’s actual AI adoption levels. He often playfully gauges their progress on a scale of one to five, where one signifies mere employee access to tools like ChatGPT, and five indicates a profound transformation of business processes through new AI capabilities.

"And everybody loves picking, ‘I’m three leaning towards four,’" he noted, "But 99% of the ones I talk to are in bucket one and two." This stark reality reveals that despite widespread interest, most enterprises are still in the very early stages of deploying AI agents in ways that generate substantial value beyond basic engineering tasks like code generation. He predicted that "It’s going to take many, many years before people can actually properly deploy AI in a very meaningful way in the enterprise."

This extended timeline, however, presents a massive untapped opportunity for innovative startups: modernizing the expensive, entrenched legacy systems that many enterprises mistakenly believe they are stuck with. Tavakoli-Shiraji points to the persistent presence of antiquated systems in major institutions. "If you go to any bank today, they still have [an IBM] AS/400 and mainframes," he described. These mission-critical systems, often decades old, are characterized by complex, undocumented codebases that pose significant modernization challenges. "Nobody wanted to touch it because nobody knew what that code did," he added, illustrating the paralysis that often grips organizations when faced with such technical debt, fearing disruption to mission-critical operations.

Crucially, the advent of advanced AI agents has changed this paradigm. These agents possess the capability to "go in and understand and actually decompose what that code does," leveraging techniques like natural language processing on code and sophisticated pattern recognition to map out system functionalities. This breakthrough opens the door to a new generation of products that can do everything from adding modern, intuitive interfaces to these ancient systems – making them accessible and usable for contemporary workforces – to facilitating their complete replacement with cloud-native, AI-driven solutions. By offering pathways to unlock the value trapped in these legacy environments, startups can provide tangible, transformative benefits that directly address a major pain point for enterprises, accelerating their digital evolution beyond perceived limitations.

Tavakoli-Shiraji’s insights underscore that while groundbreaking technology is