Operators said

The Science of Scaling · 26 Mar 2025

Predictable, Scalable, Revenue w/ Ron Gabrisko (CRO, Databricks)

Listen to the episode

These are notes on the conversation, checked against its transcript. The episode itself has the full discussion.

In brief

Ron Gabrisko, CRO of Databricks, joins host Mark Roberge to describe how he built the sales function after joining nine years ago, when the company had under $1 million in sales and about 50 people. He covers discovery-led early selling, prospecting routines, a recurring film review of discovery calls, and hiring around 40 sales staff in the first quarter, including the interview volume, ramp periods and attrition that came with it. Mark adds frameworks on sales-leader references, quantitative performance improvement plans and turnover benchmarks. The core argument is that a predictable sales playbook, built on enablement, measurement and disciplined hiring, is what lets a technical company scale revenue.

For founders

  • The first sales hire should combine account executive skills in negotiating price and moving deals with product-manager skills in spotting patterns across customer conversations.
  • Before building a sales playbook, decide the target market and persona, then build discovery around them.
  • For technical products, Ron says sales has to teach customers how to use the product, so discovery matters more than the pitch.
  • A recurring film review of recorded discovery calls with the whole team, including engineers and product, is presented as a way to speed up market learning.
  • Ron says the early sales hiring cohort did not all work out, with roughly 60 to 70 percent succeeding after a year.

For revenue leaders

  • Mark's example model gives each AE their own sourcing share of pipeline, alongside what marketing and SDRs generate.
  • Ron says he interviewed 10 to 20 candidates for each sales hire and blocked a full day each week for interviews and LinkedIn outreach.
  • Mark says performance improvement plans should be quantitative, with the threshold announced to new hires on day one.
  • Mark targets sales team turnover of 10 to 20 percent a year, most of it involuntary, against an industry average he gives as about 40 percent.
  • Mark recommends a back-channel reference question that asks the referee to rank three named sales-leader traits from strongest to weakest.

What was said 26, most useful first

The first sales hire should combine account executive and product manager skills. Listen

Mark Roberge says the first sales hire needs to handle tough conversations, negotiate price and move deals to a commercial agreement, like an AE. They also need to reflect on about 30 customer conversations, see patterns and communicate them to engineering. He contrasts this with an early team of six or seven engineers and a product person having two or three customer conversations a week, and says the first hire can take that to 20 to 30.

“Your first sales hire is a combination of almost like an account executive plus a product manager.”
A recurring film review of recorded discovery calls can speed up how fast the team learns the market. Listen

Mark Roberge describes a session at 4 or 5 pm where the team, including engineers, product and design, listens to a recorded first customer meeting. The salesperson is assigned the call in advance. The order is self-assessment first, then positive feedback, then needs-improvement feedback, then open discussion, and the review is about understanding the buyer and opportunity rather than criticising the individual. He says the frequency of the review reflects how quickly the team is understanding the opportunity.

“So I always go self-assess, then positive feedback, then needs for improvement feedback, and then we open up to everybody in the room.”
Each AE should be accountable for sourcing part of their own pipeline, not just closing what marketing and SDRs provide. Listen

In an illustrative example, Mark Roberge describes a model where a third of pipeline comes from marketing MQLs, a third from SDR cold outreach, and a third from AE strategic outbound sourcing. He describes a rep who hits quota but sources only 10 percent, then 5 percent, of their pipeline in successive quarters, and says the rep should be held accountable for those sourcing numbers.

“Bob needs to feel accountable to his own outcomes of sourcing accountability.”
For a technical product, free training for open-source users generated many sales at Databricks. Listen

Ron Gabrisko says they invited about 30 open-source users to the office and trained them on the open-source product using Databricks' own product, so attendees saw the features and the team met those users. He says many of their sales came from this, and it was later expanded into an event called Spark Live. He suggests free training and certification for technical products.

“So we would basically invite 30 Open source users to the office”
Ron interviewed 10 to 20 candidates for each sales hire and blocked a full day each week for interviews and LinkedIn outreach. Listen

Ron Gabrisko says he knew what he was looking for and had a strong network, but still interviewed about 10 to 20 people for every hire. He blocked out the entire day on Thursday for interviews and for reaching out to people on LinkedIn. He says the team needed that capacity, because even if 30 to 40 percent did not work out, the hiring rate still had to be high enough.

“I personally would interview you know, probably 10 to 20 people for everyone I hired.”
To double revenue, a sales team may need to roughly double too, because you need the capacity and some hires will not work out. Listen

Ron Gabrisko describes a successful company with typical lands of about $100,000 and 17 salespeople. The CEO asked how to pay for hiring, since new reps have ramps of 12 to 24 months. Ron's answer was that if the goal is to double revenue, you should double the sales team, because you need the capacity to build the market and some hires will not work out.

“I go better double your sales team right”
Sales team turnover averages about 40 percent a year across the industry, and Mark targets 10 to 20 percent, mostly involuntary. Listen

Mark Roberge says the average turnover on sales teams across the industry is 40 percent a year, and that building a good business on that is hard. He says zero percent is not good either, because nobody is that good at hiring, so zero turnover suggests quotas are too soft and more could be had from the team. He targets 10 to 20 percent and wants as much of that turnover as possible to be involuntary, meaning the company controls it.

“the average turnover on a sales team across the industry is 40% a year.”
Make the performance improvement plan quantitative: missing goal two quarters in a row below 80 percent puts a rep on plan, with one quarter to get back above. Listen

Mark Roberge says he uses a performance improvement plan with clear numbers and announces it on the first day of the job, along with the CEO, product, playbook and compensation plan. If a rep gets above the threshold in the one quarter they have, they are off the plan. If not, they are terminated. He says this makes termination neither political nor a surprise, and avoids a fear-based culture.

“Just make that quantitative you miss your quarter you miss your goal two quarters in a row below 80% you go on a plan.”
About 60 to 70 percent of his early sales hires were still working out a year later, which he called a good rate. Listen

Ron Gabrisko says that after hiring around 20 AEs in the first quarter, a year later roughly 60 to 70 percent worked out, which he thinks is a good rate. He says the first year often had longer sales cycles, and most of the people who went into year two were pretty successful.

“let's say, you know, 60, 70% and we had, you know, I think a good rate”
A back-channel reference question that asks the referee to rank three named sales-leader traits from strongest to weakest gets past generic praise. Listen

Mark Roberge says the traits they look for in sales leaders are managing conflict, coaching, and being exceptional interviewers. The question asks the referee to rank the candidate on these three, which forces them to name strengths and average or below-average areas rather than give constant praise. You can then match the ranking against your own read of the candidate, for example on coachability.

“can you rank them from strongest to, for one, two, three for this individual?”
Early sales should be founder-led, with the first salesperson's main job getting the founders in front of customers. Listen

Ron Gabrisko says he would advise going after as many meetings as possible, and that the early salesperson's main job should be getting the founders in front of customers. He hired people from his network to set up enterprise meetings about security, reliability and speed challenges. He says these early conversations shape many of the things that go into the playbook.

“go get as many meetings as you can”
For technical products, early sales has to teach customers how to use the product, so discovery matters more than pitching. Listen

Ron Gabrisko says early on at Databricks his selling was less about the pitch and more about discovery and asking questions. He says messaging was too technical, so it had to help customers understand how to use the product rather than starting from a blank sheet. He describes a common belief among technical founders that the best product at the lowest price will win the whole market, and says that belief is wrong.

“sales people need to teach your customers how to use it and how to get value out of it.”
When building a playbook, define the target market and persona first, then build discovery around them. Listen

Ron Gabrisko says that when building a playbook, you first have to decide the target market and the persona. Much of the early selling is discovery, asking smart questions. He says reps need to know the personas, titles and company types so they know where to look on LinkedIn, and discovery should establish whether a prospect is the right fit and whether there is a potential project.

“first you have to decide like what's your target market? What's the persona?”
Not investing in an enablement team and a repeatable playbook up front is a mistake. Listen

Ron Gabrisko says he sees other CROs failing to invest in enablement early. He describes building a playbook so the approach is repeatable, with messaging that is certified and training that teaches people to use it. He says the key to the first zero-to-10-million or 10-to-50-million phase is touching as many of the right customers as possible, so reps need to be focused on the right personas.

“one of the things I think that is a mistake is not investing in that up front”
A weekly prospecting day for the whole sales org, with weekly meeting counts, makes outbound a shared routine. Listen

Ron Gabrisko says Wednesdays were prospecting day at Databricks, with everyone in sales, not just SDRs, prospecting all day. He measured reps every Friday on the meetings they had set, around five to ten a week, and ran contests to make prospecting fun. He calls prospecting the hardest part of sales but the most important part of early company sales.

“Wednesdays was prospecting day for us.”
Databricks hired about 40 people in its first quarter of sales hiring, roughly half account executives and half sales engineers. Listen

Ron Gabrisko says that after putting together a business plan to scale sales, the company hired 40 people, probably half sales people and half sales engineers. An engineer asked what a salesperson does, which he says showed they were starting from zero. He describes this as the beginning of a culture shift.

“we hired 40 people I mean probably half of them are sales people half of them are sales engineers”
Including engineers in customer calls helps a technical team see the sales process and shifts culture toward market needs. Listen

Ron Gabrisko says as professional sales hires arrived, the culture moved from engineering-centric to market-centric. Engineers joined many customer calls so they could see how the process worked, and he went down to the engineering floor to get people fired up about the market.

“so we would include engineers you know in a bunch of the customer calls too right so they could kind of see how it worked”
Ron hired sellers who already knew the big data space, and judged them on competence and effort, since those who put in effort eventually succeed. Listen

Ron Gabrisko says that hiring people from big data and from the same space removed a learning curve, so they did not hire people with no real experience. He says that by three to six months in, you have a good idea of who is putting in effort. He expects that people with effort who come from the space will eventually be successful, and most of those hires were pretty successful going into year two.

“So we hired people that came from big data and came from this space.”
Hiring from your own network or from people already in the industry shortens the learning curve. Listen

Ron Gabrisko calls hiring from his network the cheat code, because he knows what he is getting. He also says hiring people already in the industry helps, since they already know the market and the product.

“The cheat code is I'm hiring for my network so I know what I'm getting.”
Executive recruiters are worth the money for key hires because they widen the candidate pool beyond your own network. Listen

Ron Gabrisko says that after 25 years he still did not know everyone, so for key hires he likes to use executive recruiters, which he thinks is worth the cost. He still reaches out to people on his own LinkedIn to get them interested, but says recruiters give access to a bigger candidate pool.

“I love using executive recruiters for key hires”

This quote could not be matched to the transcript. Treat it as a paraphrase.

Two interview questions for sales leaders: how they would adjust their past hiring profile for your company, and how they coached a common skill gap. Listen

Mark Roberge says a sales leader's core job is to hire, coach and hold people accountable, so interviews need to surface that. His first question asks how they would adjust the hiring profile they used at a previous company for this one and why, and how they assess the traits they say they look for. His second asks about a common skill deficiency they saw among their reps and how they coached through it. He says strong candidates give deep, specific answers.

“Can you tell me about a common skill deficiency you see amongst the sales people you've managed over the years?”
He needs a rep who has made a million dollars, because strong sales candidates ask for that proof. Listen

Ron Gabrisko says his big thing is that he needs somebody to make a million bucks. He explains that good salespeople joining early-stage companies ask who the biggest customers are and what they buy, then ask whether anybody has made a million bucks and whether they can talk to them, because they want to understand the sales culture and how hard it is to sell.

“I need somebody to make a million bucks.”
Iterating messaging quickly on customer feedback helps a company grow faster. Listen

Ron Gabrisko says messaging was being refined at the same time reps were being trained, with updates almost every quarter, scheduled after a series of road shows and customer visits. He says reps would report back that a pitch did not really work, so the team had to iterate. He concludes that the faster the feedback loop is iterated, the faster you can grow.

“the faster you can iterate that feedback loop the faster you can grow”
Mark cites a typical span of about six enterprise reps per manager, and eight to nine per manager for inside sales teams. Listen

Mark Roberge says he thinks that across the industry enterprise teams run about six reps to one manager, while inside teams run about eight or nine to one. He raises this to ask how a team going from zero to 20 reps would manage with two or three managers, and whether they would break the ratio.

“for enterprise teams it's a six to one ratio of enterprise reps to manager and inside teams it's more like eight or nine to one”
The most important thing for a sales leader is the ability to hire, motivate and mentor great people, and without it the leader hits a ceiling. Listen

Ron Gabrisko says the number one thing for a good sales leader, at Databricks or anywhere, is their ability to hire great people and then motivate, inspire and mentor them. He calls this the Achilles heel of sales leaders, and says that once a leader cannot do it, they hit a ceiling.

“what's your ability to hire great people? You know, motivate them, inspire them, mentor them.”
Start the go-to-market section of an investor pitch deck with the words Predictable, Scalable, Revenue. Listen

Mark Roberge says that opening the go-to-market section of a pitch deck with those three words resonates with investors. He says the harder question is how to get there, and that a predictable sales playbook is a key ingredient. He argues many startups get the playbook wrong by thinking it is a sales deck or a pitch deck, which he says hurts the go-to-market system.

“Start the go-to-market section of your investor pitch deck with those three words.”