Talent Density First: Netflix CTO Elizabeth Stone on Candor, Keepers and High Standards

Stone says the biggest change from running Data and Insights to being CTO is the amount of context switching and the sense of having more to learn. The engineering organization has more people, more problem spaces and more technical depth than she is familiar with, and the role carries more consequence. She hasn't checked whether she is the first economist to be CTO of a Fortune 500 company, but admits it is probably unusual.

She argues that tech companies should hire more economists, because economics is a flavor of data science and a way of framing problems so they feel tractable. When she moved from economics into tech, before the data science frenzy, that argument was harder to make. Few people at Netflix think of her as an economist, but she says it shows in how she thinks.

The most concrete benefit is thinking about incentives and unintended consequences. That applies internally, when a leadership team clarifies priorities and motivates a company, and externally, when thinking about what Netflix is to consumers and who it competes with. She notes it overlaps with psychology and helps with reasoning about cause and effect.

Lenny notes she moved from associate to vice president in three years, from data science manager to COO at Nuna in two, and from VP to CTO at Netflix in three. Stone rejects the idea of a secret sauce but lists some factors. She is dedicated to her work and teams and does the best work she can, less for her own ambition than because she is part of a team that depends on her.

She also cares about setting others up for success and acting as a translator between technical and non-technical people. She traces that skill to Analysis Group, where she had to communicate quantitative work to judges and juries. As a relatively introverted only child, she observes closely and learns from what others do, deciding what fits her own style.

Dedication, she clarifies, is not about weekends or sacrificing vacations. It is about holding herself to a high standard: replying quickly, following through, and being on time so nobody waits on her. Her example of translation is Netflix's move into live content, including WWE, which requires the content and technology organizations to build a joint strategy. She also acknowledges the Love is Blind live reunion problems, which the company has learned from.

Stone recalls her mother saying that the last 5% is the part that matters. She says she thrives in company cultures where that expectation is shared, so nobody feels alone in pushing for excellence. Lenny adds a similar story about asking whether you are doing world-class work.

With her reports she works in three steps. She models the standard herself, since otherwise there is no reason for others to follow. She gives direct and specific feedback when work falls short, because expectations are not clear unless stated. Finally, she helps fill the gap, for example by working in a document alongside the author.

She tries to deliver critical feedback privately, afterwards, so that it feels like a safe space and not a stage. Asked about burnout, she says the bar is not about time. In one example, she would rather a team spend time shaping the conversation for a quarterly business review than another 20 hours polishing the document.

Lenny describes three pillars of Netflix culture: high talent density, candor, and freedom and responsibility. Stone says talent density is intrinsic to the company and reflects founder Reed Hastings's belief that a different kind of company could make people thrive. She stresses that it is a means to an end, because none of the other cultural elements work without it.

Sustaining it requires behavior that doesn't come naturally. Managers must give uncomfortable feedback, receive it themselves, and make timely calls when someone isn't meeting the bar. The keeper test is a mental frame: if a person told you they were leaving, would you do everything to keep them? If you would be a bit relieved, you should have the tough conversation.

Stone says the test is useful because nobody wants to think that way. It only works when earlier feedback means the conversation is no surprise, which is the ideal even if real life is messier. She says it is a mental model, and people regularly ask her whether they are passing the test, which turns it into a lighter way to ask how they are doing.

Stone says Netflix has no performance reviews or ratings. There is an annual 360 cycle in which people request feedback from many colleagues, but it is not an input to any output. There is also an annual compensation cycle, and performance feeds into promotions and pay, so it has to be part of the day-to-day operating rhythm.

She admits the ideal is hard to keep. It is easy to lean on the 360, and she gets around 300 pieces of feedback, some about things from six months earlier that she wishes she had heard at the time. Ideally the 360 is only an annual check-in to distill themes.

Lenny worries this breeds a hunger-games anxiety. Stone says she has felt more at ease having these conversations than not having them, since in earlier roles not knowing how she was doing was more stressful. The high bar and the amazing colleagues can still be nerve-wracking, but hearing the manager's view, positive or critical with concrete suggestions, helps people relax.

Netflix pays what Stone calls personal top of market. The intent is to be highly competitive without the pay acting as golden handcuffs. She says they hope not to rely on compensation to attract people or to judge whether they will thrive.

Given that, hiring asks whether a candidate will help identify the right problems or new ways to solve existing ones, rather than whether they fit a box with specific skills. She looks for additive skills and perspectives, people who push the team's thinking and raise the level. Interview questions therefore probe for what could make someone exceptional as well as for baseline skills.

Lenny sums it up as not looking for someone merely great but for someone who raises the bar for the team, and Stone agrees. Lenny adds that the best people want to work with the best, and Stone says gaps in skills or behavior can be toxic for others on the team, so tolerating them has a cost.

Asked which Netflix practices other companies shouldn't copy, Stone points to freedom and responsibility. It works because the company isn't prescriptive about how people solve problems or what scope they tackle, and has few processes and guardrails. That would be hard or dangerous without high talent density, because it relies on people with strong judgment.

On her own team, innovations in content delivery, encoding, and discovery and personalization often came from individual contributors with good ideas, not from leaders declaring priorities. The challenge now is finding the sweet spot to operate efficiently at scale without snuffing out that core of the culture.

For candor, Stone describes her approach of context, not control. She takes notes in leadership meetings and shares them with her whole organization, including reflections on what isn't going well. Two years ago she introduced levels for individual contributors, a sacred departure from Netflix norms, and said the rollout was a rollercoaster. Recently she ran a candid retrospective, preferring to say it hasn't all gone perfectly than to claim every objective was met.

On Chaos Monkey, Stone says Netflix no longer runs it unbridled, because the company carries too much responsibility for the member experience to inject pain randomly. It still runs many intentional experiments to test resilience, and for new efforts such as cloud games and live content it designs low-profile tests of system limits.

Many companies embed data teams in business lines or split them by function. Netflix kept a centralized team that is functionally diverse, including data engineers, data scientists, analytics engineers and consumer researchers, and works on nearly every area of the business. This demands extraordinary partnership, because the people working on data problems don't report to the teams relying on them.

Stone says the benefits are functional excellence, better career mobility, and cross-pollination of ideas. Most important is objectivity: the team's job is not to tell the story someone wants to hear but to be truth tellers. Consumer Insights, which covers user and consumer research, and data science and engineering were combined about two years ago, which she calls unusual and a superpower in problems such as recommendations that mix attitudinal and behavioral evidence.

To stay close to people far from the top, Stone runs biweekly office hours in 20-minute slots that people book months ahead, and holds Ask Me Anything sessions with teams of varying sizes. She says it wouldn't happen unless she made it a priority, and she responds quickly so people feel they can message her.

Asked how she stays present, she says one-on-one conversations are the ones she treats as sacred, while large meetings are where she admits to multitasking. Many of her closest friends are people she met professionally, so she advises investing in others in what is a small community.

She reflects daily in quiet early mornings, asking why she feels anxious or excited, without journaling or formal meditation. In the lightning round she names Murakami's What I Talk About When I Talk About Running and A Fine Balance, asks interviewees what they would do differently in her job, and says endurance sports taught her mental resilience and recovery.