快比正确更重要：Claire Vo 谈职业跃迁、CPTO 和 AI 会拿走的 PM 技能

Vo, who started out as a copywriter and rose to chief product officer at three companies, says her secret is being a curious, impatient builder who enjoys the work. On the practical side, her advice is to know what you want from your career, be clear and ask for it, and make it easy for your boss or champion to get you there. When her company's head of marketing left, she spent half a day drawing an org chart with herself on top, including who would report where, and pitched it to her boss. She got the job.

She tells bosses what her next role is, whether that was chief product officer at Optimizely or, when she joined LaunchDarkly, a future CEO role. But she says the conversation should be a tiny fraction of her interactions, and that high performers are promoted about as fast as an organization can support. She says she has almost never wished she had promoted someone earlier, and has sometimes promoted too quickly.

Vo advises learning your organization's talent calendar, since larger companies promote in cycles and she can't always promote whenever she wants. The more important shift is the framing. Instead of 'I want to be a director because I need direct reports', say that the leader has nine direct reports and needs leverage, and that you have credibility and a track record. That, she says, solves a problem for the company rather than a career issue for an individual.

At Color she moved from product into leading engineering, because the engineering organization needed to scale and she was confident she could do it. She tells PMs it is fine to look left and right of product, since her first director role was in marketing. Her recurring theme is that the universe is bendable to your will: in fluid startup and growth-stage organizations, structure can shift around highly motivated, talented people.

Lenny asks how to avoid taking on everything just because you are good at it. Vo says she believes in operating in your zone of genius, where you are exceptional and derive joy, because that makes work sustainable. She says breadth is fine as long as you can show up energized, and warns that spending most of your time on growth edges is unfair to both you and the organization.

To find it, she suggests going through the last month or quarter of your calendar and sorting activities into buckets from hated to loved, then moving toward the top bucket. Another test is to ask what you do that no one else in the organization can. For Vo that is fluency across product, engineering, design, data, operations and revenue, plus the ability to move between strategy and details. She protects 'builder time' through calendar management, and Lenny points to an energy-audit guide by Matt Mochary.

Vo says people assume she is hired to teach later-stage companies to act like big ones, but she is hired to remind them they can operate like a startup. Her first rule is not to let pace degrade to the pace of recurring meetings: if something will be decided in the next meeting, ask why not now or tomorrow. She calls that an artificial timeline introduced by the calendar.

Her second tool is asking leaders to bring the clock speed one click faster: if something is due this year, do it this half; this half becomes this quarter, then this month, this week, today. She says it works because it is tangible, and in the moment before naming a due date you check whether it can be pulled in. The third is a personal SLA, since a leader who is slow on decisions and approvals becomes the bottleneck.

On talent, Vo says to define the bar explicitly: state leadership principles, hold people accountable to them, and build specific, measurable career ladders, especially at senior levels where criteria tend to be soft. She asks PMs to put on their product-definition hat and write real goals for each level. She also stresses normalizing feedback, quoting Brené Brown's 'clear is kind'.

She says conflict-avoidant cultures degrade the talent bar, and that questioning an idea is not questioning someone's innate talent. When someone isn't a fit, moving quickly is a hard but necessary part of the job that keeps the team healthy. In her example, two product and engineering leaders could not work together, and she told each privately that they would not stay if nothing changed, while saying she believed they could. One became a very effective manager within six to nine months.

Vo says she doesn't want to be defined by being a woman or a mom in the C-suite, but after years in healthcare, where women held more leadership roles, she is struck by how different the ratios are in tech and engineering. She cites figures such as the share of women founders and software engineers and says it is simply math. She says it has not been easy even at her level, and stresses that this is not about imposter syndrome, which she has no time for.

Her approach is curiosity: asking what is structural, cultural, external or internal, and where there are points of leverage. She says she is repeatedly asked whether she is technical, despite having written the first year of her startup's code, run large engineering teams and shipping code on weekends. She thinks someone with a different name or gender would not get that question, and says the simplest remedy is normalizing seeing diverse leaders so people can imagine them.

She tells a story about selling her experimentation startup to Optimizely while very pregnant. After concluding the company had hit a market-size ceiling, she arranged meetings with two large customers in Seattle in the same week as a Microsoft experimentation day, sat in front of Optimizely's CFO and demoed the product. Soon after, Optimizely acquired it. She cites it as an example of scrappiness and getting yourself in the room, with the results to back it up.

Vo uses CPTO for a leader who runs product, engineering and design together, which she says is very different from a pure product role. She argues that you must be technical: she compares the PRD with GitHub during product reviews, and works on architecture, infrastructure and delivery speed. The role is also operational and demands org-design skills, since engineering teams are much larger than product ones, and it comes with being paged at 1 a.m. when a service goes down.

She gives two reasons for the role. Strategically, these functions are all builders of the product, and putting them under one leader lets the company optimize for the whole rather than a function, with no debate about what is best for product versus engineering. It also gives the CEO leverage, as R&D is expensive and one executive can be accountable for it. She adds that founders can play the role, and there is no perfect structure.

To prepare, she recommends early-stage startup experience, where one person does it all, and a mindset that there are no lanes. Her team's lanes are dotted: a PM can sketch a design and an engineer can write a spec. Lenny relates this to GitLab's 'short toes' value.

Vo says ChatPRD began when, at a previous company, she needed a spec for a complicated platform product and used ChatGPT to produce a serviceable PRD during a single meeting. She refined that prompt over several months, turned it into a custom GPT for her team, then hit the monetization limits of the GPT Store and built a standalone app. She says thousands of people now use it, and people create dozens of specs and PRDs every month.

The standalone app uses OpenAI's assistants API and customizes to each user's content, role and company, and it can create and iterate on the document itself. About 60% of use is turning an idea into a PRD, about 30% is improving an existing spec, strategy doc or roadmap, and the rest is brainstorming and internal PM work. Her main tip for builders is that prompts and context really matter, and she compares outputs against other tools as a PM would.

She says it brings in real money, though not venture scale, and her goal is for it to be a hundred percent fun, a hobby that gives her coding on weekends and customer support at night. Her stated ambition is modest: covering some of her kids' education costs.

Vo calls herself short-term pessimistic and long-term optimistic about technology. She does not expect PMs to be eliminated next year, but thinks the ratios between PMs and other roles will shift, as will the skills, and perhaps faster than expected. She distinguishes lowercase-c communication, the functional trading of information, which models are good at, from capital-C communication, being influential, convincing and bold, which is far harder to replace.

Lenny says that in his Twitter poll people feared communication most and strategy least, though he thinks synthesis-heavy strategy is something AI does well. Vo agrees that synthesis, decision-making and communication are exactly where the tools are strong; the harder parts are boldness, seeing a future that a model trained on priors cannot, and charisma. She expects an identity shift for PMs, from valuing their own ideas to asking whether they are building the right thing, quickly, and delivering.

Looking at Lenny's list of PM tasks, she says many of them are about 80% tractable for AI, so the question is whether to hand over the keys or use tools to make a skilled person three or ten times better. She mentions a team that required a week of trying to automate a role before opening the job description. Her advice is to learn to build non-deterministic AI products, as PMs who jumped on mobile had their pick of jobs, and to practice with outside-in product teardowns, no-code tools and fun creative tools.

Vo says ChatPRD users report saving dozens of hours, and one said they would not need to hire another PM. Lenny notes that is hard for job seekers to hear, and Vo replies that sustainability in organizations is a leader's responsibility, since inefficient hiring and unsustainable costs help no one. She adds that a small startup that cannot afford another PM can extend its runway by not growing the team.

In Contrarian Corner she describes herself as an unabashed sales-led product apologist. She argues that sales-led motions have built tremendous businesses, and that listening to the market does not mean not caring about craft; many companies would love to be SAP with a better product. In the lightning round she recommends *High Growth Handbook*, *Scaling People* and *Circe*, praises Waymo's product and service design, and asks candidates how they would improve the company's business model.

Her life motto is 'fast beats right': making a decision and executing beats noodling forever. On TikTok she advises consistency and treating content as documentation of what you think about at work. She closes by asking listeners to help each other and to appreciate the privilege of a job where you create products by typing into a box.