Skills Over Titles: How LinkedIn's Hari Srinivasan Rewired Hiring Around What People Can Do

Host Lenny Rachitsky opens by noting that LinkedIn has shifted from a cringey place to post to a feed he finds more interesting than Twitter, and a bigger traffic source for his newsletter. Srinivasan, who does not run the feed himself, says every decision at LinkedIn is judged against one question: does it connect people to economic opportunity? He relays that the head of the feed describes two jobs for it, keeping people close to the relationships that create opportunity and surfacing the knowledge and advice members say they want from outside their network.

He is careful not to declare victory. Machine learning and ranking algorithms have grown sharper as LinkedIn learned what members value, and generative-AI prompts that let people share their perspective are an early but exciting addition. He suggests content built on real knowledge and advice, which is what surveys keep surfacing, is the kind that works best.

Srinivasan then explains the Talent Solutions org he leads, which covers any product that helps someone get a job or learn a skill. It is organized as two marketplaces, hiring (job seekers, recruiters and hiring managers) and learning (learners and instructors), with teams for Recruiter, Jobs, LinkedIn Learning and instructor tools. Even the two film studios, in Europe and Santa Barbara, trace back to the Lynda acquisition.

Srinivasan acknowledges that the market has tilted from candidates to employers, with more seekers and fewer open jobs. But he thinks several pandemic-era changes are sticking. In March 2020, he recalls, customer service roles went unfilled while hospitality workers posted about losing their jobs, because people searched for particular titles and did not realize others could do the work.

LinkedIn responded with skills-first hiring: translating people's experience into skills so a recruiter can ask for someone who can negotiate or de-escalate a customer. Hospitality workers have about 70% of what customer service needs, and the rest can be trained. Roughly 47% of recruiters now explicitly use skills when searching, which he sees as a change that is holding.

Seekers also began looking for jobs by values or interests, such as a role aligned with their purpose or something in AI, which led LinkedIn to launch collections for filtering that way. Srinivasan stresses that the company did not simply decide to impose these ideas; it noticed members wanting them and amplified the signal quickly. He adds that the Open to Work signal evolved from a hidden flag for recruiters to a public frame as the stigma of unemployment eased, and that a new Open to Internal Work feature may follow the same path.

Asked about product management hiring, Srinivasan points to public Economic Graph data showing tech hiring down about 50% year over year, with PM trending slightly below software engineering. If you are struggling to land a PM role, he says, you are not alone. His first tip is to build relationships and make clear what you can do, even when you do not already have a network.

His second tip is to attach evidence to each skill listed on a job post, such as work products or recommendations, so a recruiter can see why to recommend you to a hiring manager. His third is to target roles where you have industry experience even if you lack functional experience, for instance a PM role in automotive tech if you have worked around cars. Recruiters, he explains, always look for two things: skills and capability, and intent and interest.

On intent, he recommends expressing interest in a company page even when it has no open roles, using Open to Work, and specifying the kinds of jobs you want. He says Open to Work carries a high signal-to-noise ratio and that skills show a correlation with people getting roles. Very senior jobs, he adds, are usually recruited rather than applied for, so LinkedIn often works through messages and long-term contacts.

Srinivasan tells the story of his first product review after LinkedIn acquired his company. He had never been a PM, followed a lot of advice, and was thoroughly criticized by the then-CEO, driving home convinced he might be fired. He was told to come back in a few weeks, and rebuilt his plan by working backward from the idea that connecting people to opportunity is what matters most.

He says this is why the phrase is not just a warm slogan: with engagement, hiring, marketing and premium businesses under one roof, it is the tiebreaker. His first product was the profile, where the question of what to prioritize became simple once you could explain how a feature helps someone do what they want. Keeping the idea alive, he says, comes down to leaders repeating it constantly until it is the culture's DNA.

Below the North Star, his marketplaces are measured by tangible outputs: the number of hires made and the number of people who learned a skill, which for learning is usually measured in time spent. Among LinkedIn's public values he singles out "members first" as the principle that establishes trust and clarifies which side of a two-sided ecosystem gets priority when interests conflict.

Srinivasan has long been drawn to complex products and says he thinks about products along a complexity curve. A book he wrote for kids was finished once published, but a change such as Open to Work ripples outward: how others perceive the member, how recruiters use the data, whether people use the platform more or less. You have to reason about second, third and even fourth-order effects and measure them as you go.

That complexity also demands decision machinery. LinkedIn uses RAPID, in which the R recommends, the A must agree, the I gives input and the D decides, and he insists the most important feature is a single name on the decision line. His personal rule of thumb is that after three emails back and forth you pick up the phone, and after 20 minutes on the phone you write to the decision-maker and go.

A five-day escalation rule adds a clock: if something has not been resolved, it moves to the next level of management. He also screens for systems thinking in interviews and promotions, because most talent systems reward only how a person's own area grew, whereas this ecosystem needs people who can see and decide at the level of the whole.

Responding to the stereotype that LinkedIn product teams only micro-optimize growth, Srinivasan notes that growing the network was essential early on, but that value has long been the underlying bet. He says he is proudest of less glittery systems such as skills, values, Open to Work and commitments, which he calls major social systems. These are hidden inside marketplaces and get less attention from influencers.

He then describes LinkedIn Learning, launched in 2015 on the thesis that seekers often lack the skills for the job they want. LinkedIn recruits top instructors, helps write scripts and films them in studios, and most usage runs through enterprise customers. He calls it one of the bigger creator systems in the world and a natural extension of the mission, since helping people build skills is part of connecting them to opportunity.

His own PM course grew out of an internal Product University, created after a survey in which people said they lacked the skills to do their jobs. The team learned that frameworks alone do not teach product management; you need case studies, including failures and postmortems. The course covers how to validate and prioritize ideas against acute pain points, plus growth lead Damian's approach of thinking in loops.

For people trying to break into product management, Srinivasan describes a triangle of skills: a Spielberg-style creator, a data scientist, and a general manager. He has never seen a great PM in the center; they live on the edges. His advice is to figure out where you sit and gravitate toward roles that suit that strength instead of trying to compensate for every other corner.

He also credits luck: he worked on the first hybrid SUV in the US at Ford, a product people drove hundreds of miles to see. Experiencing something people truly loved set a bar for his later work. For PMs he manages, he asks for extreme clarity that they own their product, and then for constant building, which he treats as a muscle that atrophies without use.

Srinivasan practices this on Mind of Hari, a site of side projects he keeps deliberately non-commercial. They include a bedtime-story notebook he wrote with his oldest child, which he says has about half a million readers, a board game called Parallel Universe, and honey-sweetened gummy bears with roughly 40% sugar and five ingredients, sold seasonally.

In the lightning round, Srinivasan recommends *Thinking in Systems*, *Tomorrow, and Tomorrow, and Tomorrow* and *An Immense World*, and says his favorite interview question asks what the most complex thing someone ever built was. He looks for two things: whether they gravitated toward something genuinely hard, and whether they could simplify and explain it. His favorite recent product is a kids' toothbrush game that turns a moment of annoyance into joy.

On process, he reviews how his team works every quarter or six months. One change is planning around "orange and red priorities," the big rocks set up front so teams plan from there rather than purely bottom-up. The other is an experiment to cut product reviews from an hour to 15 minutes so teams reach a problem statement faster and either find clarity or break and return.

His pro tips for LinkedIn users are to pay more attention to the profile's skills section as skills-first hiring grows, and to try LinkedIn Learning, which he thinks many people miss because it is sold mainly through enterprises. He closes by asking listeners for honest product feedback, noting that living inside such an abstracted ecosystem makes it hard to know what everyone's experience is.