Databricks on Enterprise Sales: What Most Technical Founders Get Wrong
.jpg)
Getting unfiltered GTM advice from the people running enterprise sales at Databricks is not something most founders get. When Barry Dauber, VP of GenAI GTM, and Andrew Ferguson, VP of Databricks Ventures, came to Tel Aviv for an evening with Hetz portfolio founders, the conversation went into playbook mode.
What followed was a candid, playbook-style conversation about the gap between building something technically impressive and actually getting enterprises to buy it. The clips below capture the highlights.
From tech specs to business problems
The most common GTM mistake technical founders make is leading with the technology. Barry Dauber knows this firsthand -- even Databricks, a 10,000-person company, watches eyes glaze over when the conversation stays at the level of features and functionality.
The shift that matters is learning to connect a capability to an actual business problem someone loses sleep over. Dauber's practical advice for Israeli founders: start by selling to startups, not enterprises. The feedback loops are faster, the procurement cycles are shorter, and it's far better to make early mistakes with a Series B company than with JP Morgan. Once the pitch is sharp and the product has proven itself, graduating to large enterprise accounts becomes a different and more winnable conversation. And when you make that move, having a GTM person physically in the United States is not optional -- enterprise buyers buy from people they can meet.
Aligning IT and business buyers
Enterprise deals fail less often because the product doesn't work and more often because the wrong people are in the room. IT teams and business units frequently operate in opposition -- not hostility, but misaligned incentives and different definitions of success.
The practical implication for founders: instead of trying to convince both sides simultaneously, the goal is to help the IT person feel like a champion of the business case rather than a gatekeeper to it. Dauber's framing -- don't boil the ocean -- also applies to scope. The first deployment doesn't need to solve everything. It needs to demonstrate enough value, at low enough complexity, to earn the right to do more. That proof point is what moves a deal from pilot to enterprise license.
Pricing your first enterprise deals
Lighthouse accounts and design partners are worth pursuing -- but not for free. Dauber's position is direct: you want early customers to have skin in the game. A zero-dollar pilot doesn't create urgency, accountability, or organizational commitment on their side.
The sweet spot is a price point that feels real to the customer but doesn't require fifteen layers of procurement approval. Something in the range of a senior director's signing authority -- a $20,000 POC, for example -- is enough to put someone's neck on the line without creating a bureaucratic obstacle. The deal structure matters too: framing the POC cost as something that folds into the eventual enterprise license gives the customer a reason to move quickly if the outcomes are met.
Customer proof points beat partnerships
Startup partner programs at large platforms are, by design, long tails. Databricks alone has thousands of software partners and a small team to manage them. The math of mindshare is brutal: most partnerships get very little active support, no matter what the contract says.
What actually moves the needle is a shared customer. If a company like Coca-Cola -- already a Databricks customer -- tells their Databricks rep that a specific startup is solving something for them, that conversation changes entirely. The startup is no longer one of thousands in a partner directory. It is attached to a named account the sales team already cares about. Chasing the case study and the proof point is almost always a better use of time than chasing the partnership itself.
Getting enterprise-ready through the Databricks AI Accelerator
The Databricks AI Accelerator is an invite-only program for seed and pre-seed companies building data and AI-intensive applications. It is designed to do a specific thing: help early-stage startups get their technical architecture right, become security-compliant and enterprise-ready, and build the customer references they will need to raise a Series A.
Hetz Ventures is the only Israeli VC partner in the program's network. For portfolio companies that qualify -- and for founders thinking about what enterprise readiness actually requires before they need it -- this is the most direct on-ramp available. The program provides product credits, architecture guidance, and a path into Databricks' broader sales motion for companies that complete it.
This event was part of Hetz Ventures' ongoing program of bringing senior operators and platform partners to Tel Aviv for direct conversations with Israeli founders.
