Startups operating in today’s hyper-competitive technology landscape are under more pressure than ever to achieve rapid growth and sustainable scaling. For many early-stage companies, reaching that next tier requires landing lucrative contracts with large enterprises—the deep pockets of the global IT world. Yet, as any veteran software founder or enterprise sales leader will readily attest, securing these high-stakes contracts is notoriously difficult, filled with bureaucratic hurdles, lengthy sales cycles, and fierce competition from established market leaders.
One of the most critical and frequent mistakes made by younger, less-experienced startups is the persistent belief that their technology will practically sell itself. This insight comes from Arsalan Tavakoli-Shiraji, co-founder and senior vice president of field engineering at Databricks, a powerhouse in the big data storage and artificial intelligence sector.
Today, Databricks stands as an AI and data storage giant, boasting a staggering valuation of $190 billion and servicing more than 70% of the Fortune 500. However, the enterprise was once a fledgling startup itself, and Tavakoli-Shiraji was directly responsible for helping navigate those turbulent early days and bringing in the company’s foundational enterprise deals.
Reflecting on those formative experiences during a recent appearance at TechCrunch Disrupt, Tavakoli-Shiraji addressed the common misconceptions held by founders trying to break into the enterprise market.
"Everybody when they’re selling to enterprises at startups… we think it’s all about the tech and having the best tech," Tavakoli-Shiraji told attendees. While he acknowledges that underlying technology must function exceptionally well to survive, he quickly warns that having a superior product is simply not enough. Through trial and error, he learned the hard way that "people and processes matter the most."
Tavakoli-Shiraji vividly remembers the harsh realities of early enterprise sales. In the company’s infancy, he once secured a coveted meeting with a major corporate prospect, only to discover upon arrival that the executive across the table was the exact same decision-maker who had recently signed a multi-million-dollar mega-deal with Databricks’ primary competitor.
"He has no interest in listening to me because he bet his career on that competitor," Tavakoli-Shiraji recalled, illustrating how brand loyalty and career risk management dictate corporate purchasing decisions far more than minor technical advantages. He also remembers numerous other early prospective deals completely fizzling out at the final stages simply because the startup was entirely unprepared to navigate complex corporate legal requirements, rigorous contractual demands, or stringent security questionnaires.
Drawing from his journey of transforming Databricks from a small enterprise into an industry titan, Tavakoli-Shiraji offered several strategic principles for startups aiming to win over massive corporate customers.
Sell under a bigger tech company’s umbrella
In the early days of Databricks, the startup attempted repeatedly to secure a introductory meeting with UBS, the massive Swiss multinational bank, but repeatedly hit a brick wall. The doors simply would not open for an unproven company.
The turning point arrived when Databricks executed a strategic partnership with Microsoft, integrating its platform to become an official Azure cloud partner. This partnership fundamentally changed the mechanics of their sales motion. By aligning with Microsoft, Databricks’ sales agreements were suddenly covered seamlessly under Microsoft’s established enterprise terms and compliance frameworks.
Crucially, it also meant that Microsoft’s massive, trusted sales force actively booked meetings and advocated for the startup’s technology to their existing enterprise clients.
"All of a sudden, that meeting wasn’t hard to get," the founder recalled. "It was a huge boon to us that we were able to partner with somebody who had that access and distribution."
With Microsoft’s formidable sales engine helping drive momentum, Databricks found itself pulled into boardrooms across the globe. While this wave of demand created entirely new operational challenges—such as prospective clients demanding localized features for regulatory environments in countries where Databricks had not yet expanded—they were undeniably positive problems to encounter on the path to hyper-growth.
Over the years, Databricks has evolved to play that exact empowering role for a new generation of smaller startups. The company now acts as the established tech giant that works hand-in-hand with early-stage innovators, selling their specialized products directly under Databricks’ own trusted enterprise terms and rigorous cybersecurity setups. Notable examples of this ecosystem approach include Hightouch, a company that provides advanced AI-powered marketing and customer-data tools and in which Databricks has made strategic investments; Addepar, which offers specialized data analytics tools for investment professionals; and Tecton, a machine learning startup that Databricks ultimately acquired.
When selling AI, address the "four Cs"
As enterprise focus shifts heavily toward artificial intelligence, startups must navigate a distinct set of corporate fears and requirements. Tavakoli-Shiraji categorizes the primary concerns and objections raised by enterprise buyers into four distinct pillars he refers to as the "four Cs": context, choice, cost, and control.
Context refers to the enterprise demand for AI systems that can be trained and customized effectively—a process he likens to "bringing on a new employee." Enterprises require AI that can seamlessly learn the organization’s proprietary terminology, internal processes, unique business priorities, and historical data patterns without exposing sensitive information.
Choice represents the enterprise need for flexibility and the freedom to switch between different AI models rather than being locked into a rigid, single-provider ecosystem that could leave them vulnerable to sudden price hikes or technological stagnation.
Cost focuses squarely on financial predictability, requiring clear mechanisms for monitoring and controlling how much an AI product will actually cost to run at scale across thousands of employees.
Finally, Control encompasses the essential guardrails, compliance frameworks, and security safeguards that govern corporate AI behavior, ensuring strict adherence to regulatory standards and the absolute protection of proprietary and customer data.
Attack the expensive legacy systems enterprises think they’re stuck with
During his discussion, Tavakoli-Shiraji offered a humorous assessment of where major corporations truly stand in their artificial intelligence journeys. When he asks Chief Information Officers to rate their organization’s AI adoption on a scale of one to five—where a rating of one means employees merely have casual access to consumer tools like ChatGPT, and a five means the company has deeply integrated transformative AI capabilities that fundamentally altered its business processes—the responses are revealing.
"And everybody loves picking, ‘I’m three leaning towards four,’" Tavakoli-Shiraji noted with a laugh. However, he quickly adds the sobering reality: "99% of the ones I talk to are in bucket one and two."
According to his observations, large enterprises have largely failed to figure out how to deploy autonomous AI agents in ways that generate genuine business value beyond allowing internal software engineering teams to write code slightly faster. He predicts that it will take many years before traditional enterprises are able to properly and safely deploy AI in a deeply meaningful, enterprise-wide capacity.
This lag presents a massive market opportunity for nimble startups capable of delivering tangible value by modernizing deeply entrenched, expensive legacy systems that corporations assume they are permanently stuck with.
"If you go to any bank today, they still have [an IBM] AS/400 and mainframes," Tavakoli-Shiraji described, pointing to decades-old computing architecture still running core financial infrastructure. "Nobody wanted to touch it because nobody knew what that code did."
However, the advent of modern AI agents has completely changed the equation. Today, advanced AI systems "can go in and understand and actually decompose what that code does," he explained. This technological leap opens the floodgates for innovative startups to build products capable of performing everything from layering modern, intuitive user interfaces onto ancient, text-based software to completely rewriting and replacing legacy mainframes altogether.
