loading
loading
Ron Gabrisko joined Databricks as CRO when it was doing under $1M in ARR. Databricks is now a $190 billion company doing $7B+ ARR growing at a pace like we have never seen before. Ron sat down with Sam Blond, co-founder of Monaco and former CRO of Brex, on the SaaStr AI Podcast. They last recorded in October 2024, and this episode covers what has changed in sales and go-to-market in the two years since: hiring, demand gen, seller time, post-sale, and pricing. Databricks is 13 years old. Monaco started in 2024. Sam compared Databricks to an oil tanker turning around, and Monaco to a company built AI-native from day one. Their answers line up on most topics, and split on one. 1. Every Databricks AE Candidate Builds a Live Demo in Genie During the Interview Ron says the basics of great selling are the same as two years ago: prep, relationships, account research, grit, EQ and IQ. The addition is that every seller has to be AI first. Databricks tests for that in the interview loop: - Pick an account. Any account. - Give a CEO-level AI pitch for that company. It has to be industry specific: how Databricks would transform that company and its industry. - Build a demo in Genie on the spot. Then walk the panel through it and the outcomes it would drive. Back channels and referrals from known networks still factor in. Every candidate also has to be met in person. After hiring, the same skills get certified. Every rep has to be certified on Genie and on the CEO-level pitch, and certification runs throughout the year. Databricks AEs build custom demos for specific accounts, such as Pepsi and Rivian. The company runs Databricks on Databricks, and reps use Genie on their phones. Ron, a former baseball player, expects reps to practice 10x more than they play. At Monaco, Sam sees some of the same top performers from Brex still at the top of the leaderboard. He also sees new names there, sellers with what he calls a go-to-market-engineer quality, who walk customers through the technical internals of the product in demos. Sam said that historically, neither he nor the top sellers he worked with could do that. His takeaway from the conversation: Monaco should add a technical test to onboarding that reps pass before they’re allowed on a demo. For founders: Databricks evaluates sellers by having them build with the product in the interview. If your loop is mostly deal stories and role plays, you’re testing for a different job. 2. Databricks Kept SDR Headcount Flat While Adding AEs. Reply Rates Went From 1-2% to 6-10%. Databricks still has a large SDR team, but its size has stayed roughly flat while the company keeps adding AEs and solutions architects. Ron attributes that to productivity per SDR. Databricks has automated outbound, signals on prospects and contacts, and campaigns, with personalization down to the individual. The old model, in Ron’s words, was loading 5,000 contacts into “the automated spammer” and getting 1-2% response rates. With outreach customized to each person, Databricks now gets 6-10%. If he were starting a company today, Ron said he would build demand gen and the tooling for it, and keep the team small. Monaco’s numbers: - One SDR (Sam’s nephew) books 400 outbound meetings a month - No one in marketing until two weeks before the recording - Account research, buyer identification, signal tracking, message customization, and LinkedIn and email outbound are all automated Monaco still runs a few things by hand: gifting campaigns, high-end events, and cold calling. Sam says cold calling still needs a human. The improvement is in targeting: the system tells the rep who to call, when, why, and with what script. Sam thinks a bot cold calling on your behalf would be too dystopian today, and expects that could look different in two years. For founders: At Databricks, AI didn’t shrink the SDR team. It meant the SDR team no longer had to grow in step with the closing team. At Monaco, one SDR covers the whole company. 3. Ron Keeps the Human SDR Team Because It’s Where His AEs Come From. But Do You Still Need Them? Many Databricks SDRs get promoted to AE. Ron’s concern with cutting the team to zero: then nobody is teaching new sellers how to sell Databricks. He expects the team may get smaller over time, but he wants an academy of some kind for younger, less experienced sellers, with a path into AE or management roles. Without it, every AE has to be hired from outside and no one is mentoring the next generation. Ron also said some of his best sellers today came up through those ranks. They grew up with AI and build their own programs and apps using Genie and LLMs. Sam and Ron both started as SDRs. Sam pointed out that the SDRs who graduated around 2024 entered the workforce after AI tools were already in place, and two years later many of them are top sellers. What makes them good is different from what made their predecessors good. For founders: Automating SDR work and keeping an SDR team are separate decisions. Databricks automated the work and kept the team as a training pipeline. 4. 60% of Rep Time in Front of Customers, and Demos Built on the Customer’s Industry Data Seller time. Ron’s rule is “sell from your feet, not your seat.” He thinks a lot of teams lost that during the shift to virtual selling. He pushes for at least 60% of rep time in front of customers, with this weekly split: - Monday and Friday: prep and admin - Tuesday through Thursday: on the road with customers, or more Databricks has built agents for account research, presentation prep, exec briefing docs, and proposals. That time goes back to customer meetings. Ron also expects AEs to multithread into new departments, people, and stakeholders themselves rather than leave it to SDRs. His line: “You’re the CEO of your account.” At Brex, Sam’s best AEs generated a lot of their own pipeline through outbound. At Monaco, AEs spend so much more time with customers that the top closers have no time left for outbound. The buyer. Databricks used to sell mostly into IT. With AI now a CEO-level priority, Databricks sells to CFOs, CMOs, and other business leaders. Ron says the time saved by automation is what made covering those buyers possible. Demos instead of slides. Ron says customers now expect to see what you’ll actually deliver. That may not be on production data, but it should be on dummy data that resembles their industry and company. Databricks has built and then deployed use cases like supply chain predictions and stockout prevention for large retailers and pharma companies. Ron’s view is that most vendors’ messaging sounds the same, so buyers decide on what you can build for them, how fast, and whether they trust you. Monaco has replaced the traditional CSM role with forward deployed account executives. In calls or Slack, the rep shows the customer something already built, explains how they’d deploy it, and offers to deploy it for them. Under the old model of quarterly check-ins and usage monitoring, the customer did that work. For founders: Databricks and Monaco both have reps build a working version of the customer’s use case before the customer commits. Plan for the time and tooling that takes. 5. Databricks Charged $1,000 Per User. When It Made Users Free, Consumption Jumped. Ron calls building the company on consumption pricing one of Databricks’ best decisions. Revenue grows as customers’ data and queries grow. He now expects millions of agents to query data, and says Databricks has more data under management than any other company. Early on, he told several AI coding companies they would never get big on per-user pricing. They have since moved to usage pricing based on tokens. His pricing framework: - Find your base unit. It should scale over time and track the value the customer gets. LLM companies use tokens. Databricks uses DBUs. - Price additional value as multipliers or add-ons to that unit. Examples: security, accelerators, premium support. - Match price to value. Price too high and no one buys. Price too low and you leave money on the table. Databricks once added a per-user fee and a platform fee on top of consumption. When Ron asked customers why they weren’t ramping faster, they told him each user cost them $1,000. He asked whether they’d open it to everyone if users were free. They said yes, and consumption went up sharply after the change. Two other mistakes Ron sees: - Requiring a commit before customers can use the product. He wants the lowest possible entry point: free tiers, free trials, PLG, or a credit card. Once customers like the product, usage spreads from one or two users to a department, then several departments, and then you ask for a commit. - Pushing for a large deal on day one. Early-stage companies need logos and new customers first. Ron thinks companies that started on the wrong pricing model can still change it. The worse mistake is failing to capture value from customers. For founders: Choose the pricing unit early, since changing it later is hard. Remove seat fees, platform fees, or upfront commits that limit how much customers use the product. A Few More Interesting Learnings - Sellers now need both technical and business skills. Ron’s reps build demos and pitch business outcomes to CEOs. - Every customer-facing role needs AI proficiency. That includes SDRs, AEs, and FDEs. Sam’s advice to sellers who haven’t made the shift: use the consumer AI products and deploy agents yourself. - Both companies sell on their own product. Monaco runs on Monaco, and Databricks runs on Databricks. - Cold calling is still done by people. Both see it as the one demand gen activity that hasn’t been automated, though targeting has improved. - Ask customers where adoption is stuck. Ron’s simplest pricing advice is to find the friction and remove it. Where Databricks and Monaco Land on the Same Playbook Despite an 11-year age gap, both companies: - Have candidates build with the product during interviews - Use AI for research and outbound, and put reps in front of customers - Show customers a working build on realistic data - Price on a unit that grows with customer usage, with as few caps as possible They differ on SDRs. Monaco gets 400 meetings a month from one SDR. Databricks keeps a full SDR team, because that team produces its next generation of AEs.
document fetched 2026-10-29 03:04 · canonical https://saastr.com/databricks-cro-ron-gabrisko-5-interesting-learnings-on-selling-at-a-200b-ai-company-flat-sdr-headcount-6-10-reply-rates-and-free-seats-that-unlocked-consumption