The playbook for building high talent density teams | Adam Ward, Head of Talent at Cursor
In a Nutshell
The core message is to abandon the "funnel of doom" approach of sourcing hundreds of candidates and instead identify the world's top 50 people for a specific role, then relentlessly pursue them through deep network cultivation, targeted scoping, and personal relationships. The most important takeaways are that work samples and extended onsite experiences provide the highest signal quality, treating each candidate as the atomic unit rather than optimizing for volume, and that recruiting excellence requires founders who treat talent as the highest priority with the same rigor as revenue or uptime metrics.
These notes were generated by AI and may contain inaccuracies.
Companies think about recruiting like a funnel. They reach out to 100 people, expecting 20% to apply. By definition, that is not the top 20%. That is just the 20 people caught on a bad day. The way companies recruit is they weed people out at each stage and hire the remainder. Adam Ward describes this as the "funnel of doom."
The alternative approach is to identify the 50 people in the world who can do a specific role, then relentlessly pursue and focus energy on activating those people, rather than trying to find them out of the funnel of doom.
The mistake most companies make is asking "Hey Lenny, who's the best product engineer you know?" The better question is "Hey Lenny, of all the product engineers you've worked with, who stands out to you as the person who's most collaborative with designers? Or who can translate a framework into a product better than anyone you've ever seen?"
Recruiters are excellent at 90 to 110% capacity and are terrible outside of that. If they only have half a req load, they are useless. They are great at the edge of adrenaline rush busy. Find those people.
Adam Ward is known in the Valley as the most effective recruiting leader at building elite high talent density teams. He co-founded an independent recruiting firm called Growth by Design, which Cursor acquired fully last year. Adam is now Head of Talent at Cursor. Prior to starting his own firm, Adam led global recruiting at Pinterest, growing the company from 200 to over 2,000 employees. Before that, he led a global team of 65 recruiters across 10 offices at Facebook, helping them scale Facebook's tech org from 1,000 to over 10,000 employees.
The hiring market is at an 11 on a scale of 10. There are similarities to other markets, like when companies were moving to mobile and there were no mobile engineers. The difference now is that everything is compressed—days and weeks instead of months or years.
It's a tale of two cities: back-to-back headlines of NBA salary offers to a new grad PhD and a blue-chip company laying off 10% of workforce. There is a resettlement of talent and labor and how work gets done, with a large gap that will come back closer together.
Trending up: Forward deployed engineers—technical people who can partner with sales teams and customers to deploy technical products, helping companies move from token maxing to optimizing. This is a real supply-demand issue.
Waning: Specialized technical people who are very narrow, very deep, and overly specialized. The market is squeezing together engineer, product, and design, preferring design engineers or engineers with stronger product sense.
New grads who haven't had enough experience are also waning, as products like Cursor can do some of that work.
The person who was a full stack engineer three to five years ago is now potentially a forward deployed engineer. These people like moving out of the terminal and into the company, customers, and clients. For the right person, this unlocks another chapter in their career.
These roles are high in demand. There is a return to the craft of things. AI can create slop in design, but there is hope it will help get back to the craft and appreciation for it.
There has been a rise of the power IC over the last 12 months. Taste, decision-making, problem-solving, and curiosity in all the good ways are coming back to the surface.
Adam first heard the term as an intern in 1998 at Trilogy. His first manager Jeff Daniel coined the term during a campus tour talking to CS students.
Candidates come from three buckets: they apply, are activated or sourced, or are referred. Referrals and sourcing of passive talent is the focus for talent density. When sourcing talent, companies reach out to 100 people, 20 people reply. By definition that is not the top 20%—it is just the 20 people caught on a bad day.
This leads to remainder hiring, hiring the person who drops to the bottom. This approach regresses to the mean over time as volume increases.
Comparing recruiting to a sales funnel is detrimental because in a sales funnel, you are selling a widget or piece of software that is relatively static and rational. In recruiting, you have an irrational unit of a hiring manager and recruiter, and another rational unit of a candidate. Recruiting as a craft is only 40-50 years old, compared to sales which has been around for hundreds of years. Humans try to relate recruiting to sales, which has been somewhat detrimental to thinking about it strategically.
When doing a senior specialist hire or leadership role well, there are three main steps. The first is scoping—getting really clear on what the top 1% means before talking to any candidate. This sets up the entire process: sourcing strategy, pitch, assessment process, and close.
Pair scoping with really good market mapping, intelligence, and signals to give confidence of who the very best people are. Identify the 50 people in the world who can do the role, then relentlessly pursue and focus energy on activating those people.
Scoping involves thinking about skills and experiences needed, stack ranking them, and getting clear objectively before talking to candidates. This helps craft the sourcing strategy, pitch, assessment process, and close. These are objective, somewhat transferable things—not asking what companies candidates are from.
Do not focus on logos on resumes. There are vintages of companies when a team did fantastic things, and possibly because a person was on that team at that time, they are great. Most companies jump to that almost immediately.
Start with a set of people who are trusted in terms of having high taste in talent. The mistake most companies make is asking "who's the best product engineer you know?" The better question is asking about specific characteristics: "who stands out to you as the person who's most collaborative with designers?" or "who can translate a framework into a product better than anyone you've ever seen?"
When asking the right targeted questions to people you trust, you start to triangulate around the same names.
Map characteristics like "curious, systems thinker" to companies that share that approach to how they hire. Do a Boolean or AI string to look at job openings among 100 companies to see which ones also say these characteristics matter. Build company lists that are not the normal list, validated through the things objectively identified as needed.
Systems thinking is a trending attribute. It comes down to the fundamental question of what problem are we trying to solve, then breaking the larger problem into really small pieces. Engineers excel at this—taking something
Systems thinking sounds strategic but is actually extremely tactical. The best systems thinkers take big problems and break them into very small tactics, removing each blocker one by one while staying anchored to the overall goal.
Finding top talent involves sitting with people and jamming in a document to capture recommendations of people and companies to look at. While tools like LinkedIn exist, the most valuable approach is the most time-intensive: spending time digging into people's networks. At Cursor, there are people whose entire job is to dig into current employees' networks.
This is done through one-to-one, hand-to-hand cultivation of signal. The approach involves sitting down every other week for six weeks and diving into people's networks. One person shares three names, the other checks them out and asks follow-up questions about additional people. Notes are sent after each conversation, and there's accountability around follow-up.
While these signals go into a system and database, the extraction itself is a hugely personal and effective approach.
When founders say talent is their number one priority, there should be no obstacles in terms of effort, resourcing, or time put in to get there. If talent is the most important thing, you can't outsource it or hand it over to a recruiting organization to do.
The very best companies have founders who keep the company on track around talent. What gets measured gets done. Companies measure talent the same way they measure uptime, customer satisfaction, or revenue, and hold hiring managers and leaders accountable to hiring.
A major pet peeve is having recruiters sit in a position where hiring managers aren't deeply involved in the process. Many companies have recruiters own the hiring goal, but recruiters aren't making hiring decisions—hiring managers, leaders, or committees are.
Recruiters' goal should be helping provide confidence to the hiring manager to make the right decision. They're a confidence engine, not a decision-making engine. No one should care more about a hire, and no one gets the joy or pain of a good or bad hire, than the hiring manager and team.
Coda's approach to hiring involves a Slack channel called "hiring ideas" where people on the team share people they might want to hire. The channel is super active—someone comes across something impressive like a product, a post, or overhears something at a barbecue, and they drop that person's profile or X post into the channel.
It creates a rallying effect where people share what they've heard, what they know, and what they think is compelling about the person. The team creates a quick strategy for how they might engage that person. This creates social credibility in the company to surface great people, and then others come in and help get them into the process.
This approach recognizes that top people aren't going to apply through a traditional funnel—they're not making it through the funnel, and often you don't even know what you're hiring for.
Rather than having people apply and work through a funnel, the approach is to figure out who are the best top 50 people in the world for a role, then relentlessly pursue them. Not all hiring ideas are great or right for the company, but it becomes a tangential injection into the hiring process that often leads to great outcomes.
Relentless pursuit recognizes that this can feel like a slow burn at day zero, but over multiple days you're planting seeds with potential top 50 people who will come to harvest at different rates. Stories exist of working on candidates for a year.
When people say it's not good timing, the response isn't to ask them to interview or consider a role—it's to ask for a next conversation. Even if they can't join until late fall, you can still connect them and have conversations on a weekly basis. The continual touchpoint approach means being willing to engage over days, weeks, months, and sometimes years.
When reaching out to someone at Cursor (now part of SpaceX), the brand helps, but for less interesting or successful brands, the key is spending time upfront to get a sense of what could be engaging for them. Almost as much effort goes into finding who can create a warm intro as goes into the actual messaging.
"Caring is free"—when building recruiting teams, the principle is that you can care more about a candidate than anyone else. Candidate satisfaction scores consistently show the number one thing is that the company generally wanted them and made them feel like they belonged there. Caring is an unfair advantage that exists within everyone.
Rather than getting caught up in the numbers game of reaching out to 100 to get to 20, the focus should be on hiring one person for the role, not 10. The approach is building relationships with dozens of people who could be the one.
This also means being comfortable with bespoke and ad hoc hiring rather than overly consistent approaches. Instead of letting fate or availability decide who has the next conversation, the savvy recruiter or hiring manager matches people based on shared interests mentioned by the candidate.
The most important conversation is the hiring manager's first conversation with the candidate, not the onsite interview. This person is who the candidate would work with and who makes the hiring decision. They should be asking the most important questions and pulling out the most important motivators.
Hiring managers might meet with a candidate three or four times before that person is fully in process.
The atomic unit of hiring is the person, not the job spec or role. This makes the process feel like a decision tree that could go many different ways based on each conversation. When not caught in the numbers game, it feels less scary to focus on how to get that person there versus whether they're the right person for a specific job.
Breaking bread—having a meal with someone—is one of the most human and leveling things you can do. Sharing a meal across the table is a disarming and joining act. Tactics center around food and drink as much as possible, whether coffee, lunch, or breakfast.
For people who might be famous or well-respected with expertise, having them come in and give a talk to the team or do a fun side project demonstrates increased interactivity and gives them a realistic preview. The goal is to play "home games" by bringing people into the office when possible, though some prefer neutral sites like coffee shops early on.
The goal is for people to feel a gut reaction when walking into the office—like the feeling when walking onto a university campus and knowing it feels like home. This can be a defining moment, and it's valuable to suss out early if it doesn't hit.
Cursor's office has a warm, soft feel like a Hogwarts library, with lots of chalkboards creating a disarming, warm feel that evokes Midwest vibes—very hardworking and ambitious, but understated. The work speaks for itself.
Research shows work samples are the highest predictor of success at a company, yet most companies still spend time doing one-to-one across-the-table assessments. Cursor is known for doing long onsite experiences where candidates do a project and work side-by-side, giving strong signal both ways.
Most companies get the onsite wrong by treating it as an assessment where they're extracting information without giving much back. The power of self-selection serves good inputs and data, so the onsite experience should feel more two-way.
The onsite is carefully curated: who candidates meet, work samples and project work, who's greeting them, who's sitting beside them for lunch. Everything is curated to make it a high-quality experience that gives candidates data while the company gets data from people most likely to connect with that person.
Recruiters primarily run point on candidate experience as the quarterback, while hiring managers contribute the most insight into people. Recruiting coordinators handle significant execution including scheduling, candidate shepherding, and logistics like moving candidates between workstations, lunch, and project work, as well as helping set up environments.
The worldview that every candidate is special and rare transforms every touchpoint in the journey, creating a fundamentally different approach than typical recruitment processes.
Once candidates have a great experience, they become part of delivering that experience for future candidates. This creates a flywheel effect once the company gets over the initial hump. Current employees who went through the process recognize and acknowledge the value, then help deliver it for the next batch of people.
Not all exceptional candidates will join the company, but every one can walk away as a net promoter. These promoters carry more weight because they're influential, great, and known to be good. Even if they never return, they may tell others, creating groundswell and building a stronger talent magnet.
Cursor tried to eliminate the work trial due to the significant time commitment and candidate resistance, but found it tanked their signal quality and had to bring it back. The work trial provides high-confidence decision making and is worth the investment if talent is the number one priority.
While not everyone is suited for work trial work, team members can contribute in various ways - some handle lunches, others do greetings, creating multiple participation opportunities.
Different functions have tailored versions of the work trial. Go-to-market teams work through customer challenge scenarios often one-to-one with the hiring manager. Product teams do similar project-based work. The goal is to create function-specific but uniform approaches across different roles.
Candidates can choose from a set of problems and the company is flexible about right-sizing based on availability and time constraints. They sometimes use Saturdays to accommodate working candidates. While there's initial conflict around time and effort that can feel slow, once candidates are ready to move, the process accelerates significantly.
Debriefs happen as immediately as possible - typically same day or next day. Each function has slightly different processes, but hiring is manager-led rather than through hiring committees. The hiring manager gathers input from the team but holds the group accountable for providing the right inputs to make the best hire or no-hire decision.
Offer decisions go to founders for final approval. When facing competing massive offers from companies like Anthropic, the approach is to sell throughout the entire process rather than at the close. The craft of recruiting involves upfront clarity on motivations and ensuring the entire experience addresses those factors, making the offer feel like a natural culmination.
The focus stays on intrinsic factors - the role, team, people, and work - rather than extrinsic compensation factors. This becomes challenging when competing with large compensation packages, but the goal is attracting people who genuinely want to be at the company. Sometimes candidates discover new motivations during the process, and discovering misfit early is preferable to redoing the work in six months.
Companies like Anthropic and Tesla don't allow negotiation on offers. While this approach demonstrates commitment and has been done by various companies over time, it's not necessarily right for every organization.
Candidates should present offers with clear logical alignment - walking through the reasoning behind requests makes it easier for recruiters to understand and communicate to decision makers. Inconsistent or add-on requests create problems, while being clear and consistent from the outset about the four most important factors facilitates productive engagement.
Current compensation levels are extraordinarily high, with open questions about where the market goes from here. Time has become the most fixed and precious asset - a race against the clock to secure model talent. The shift has been dramatic for AI researchers who were previously undervalued in academia but are now the most in-demand talent.
Closing happens throughout the process by removing objections and understanding motivations. It's treated as a team sport with daily stand-ups around high-profile candidates, involving hiring managers, recruiters, and former peers to ensure candidates feel cared for and that they belong.
Personal gestures make significant impact. A candidate who was a classically trained violinist from a performing arts high school received a unique Turkish stringed instrument as a gift during their offer dinner, demonstrating thoughtfulness and uniqueness that stands out in the candidate experience.
Each candidate gets their own Slack channel for strategy, updates, and coordination, creating the same rallying effect during the close phase.
The fundamental approach treats each person as the atomic unit rather than optimizing for funnel efficiency. This doesn't scale to processing hundreds of candidates per role, but works when narrowing to a small percentage of high-potential people where groundwork has been laid over time.
The thesis is that companies at the cutting edge of technical revolutions were recruiting in outdated ways. With commitment to time and effort as a collective force, organizations can gain unfair advantage in talent. Company success often comes down to execution, which depends on having the right talent at the right time.
Working with founders who truly believe recruiting is the most important thing is both a gift and occasionally frustrating. The gift is that convincing founders to participate in recruiting activities isn't a problem. The challenge is that founders are deeply involved and want to do the work themselves, sometimes creating confusion about ownership, but this is considered extremely net positive overall.
Recruiters should potentially be the highest paid team because of the tremendous impact and the supply-demand crisis around strategic recruiters who can function at this level. The somewhat transactional nature of traditional recruiting has created scarcity of recruiters with this approach and skill set.
Talent density isn't just about individual 10x performers unless the team consists of one person. Great leaders function as puzzle builders, assembling complementary teams that help each other excel. Talent density is fundamentally collective in nature, though recruiting one human at a time can create over-focus on individual contributors rather than exponential team impact.
A 10X person cannot exist on a bad team. Excellence gets throttled without a strong and supportive team environment. The bar for team quality is non-negotiable, similar to how client service providers cannot succeed without great clients.
The most common early-stage hiring mistake is attempting a "silver bullet hire" - trying to hire a first head of talent who must simultaneously solve two competing skill sets. Founders typically hire when already behind on hiring, creating debt of 5-10-15 unfilled roles. They want someone who can both execute immediate hires and build scalable systems. These are competing interests that should not be collapsed into one role.
Rather than combining system-building and immediate hiring execution into one role, these functions should be decoupled. One approach is hiring someone to build the foundation and partnering them with a contingent recruiter to make short-term progress. Alternatively, bring in someone to establish the foundation first, then hire a high-powered IC recruiter who can execute. This prevents the new hire from being forced to either make hires without systems or build systems without making hires.
The most important quality for a first recruiting hire is being curious and business-focused. They need to understand how the business works and how each hire will change the slope of the business. The second critical attribute is high motor - someone who understands the grind and pace of early-stage work and thrives at the edge of overwhelm. Recruiters are excellent at 90-110% capacity and struggle outside that range.
At Cursor, the team composition differs from previous roles because the company and product demand different types of recruiters. The Cursor approach requires recruiters who understand leadership recruiting motions, won't get caught up in funnel numbers, and have strong networks they can tap into. Seasoned recruiters' networks can be as influential as technical leaders' networks.
The three attributes sought in team members are:
- Excellent humans who are excellent at what they do, with no willingness to trade off either quality
- People with a chip on their shoulder who have something to prove and love the craft, demonstrating internal drive and motivation
- People who put team over themselves, choosing to work on the craft of recruiting rather than pursuing middle management
These attributes can be identified through network referrals by asking for people who are excellent humans, excellent recruiters, have natural drive, and put team over themselves.
Effective recruiting leadership requires balancing operational involvement with strategic perspective - being in the weeds while also lifting up to paint the bigger picture.
Renegs on offers are becoming more common, particularly for senior roles where there's often a gap beyond a couple weeks between acceptance and start date. After offer acceptance, companies should maintain a second motion focused on ensuring the candidate shows up and accelerating their ramp through pre-boarding. This includes dinners with other accepted candidates, creating community, sending company laptops, and regular touch points to help them visualize their future role.
A growing trend is the rise of the talent engineer - a function within recruiting teams potentially filled by former engineers who build tools. The perspective is that every recruiter should be a talent engineer, as tools like Cursor now provide recruiters with capabilities previously requiring engineering resources. The goal is for recruiting teams to feel like talent engineers rather than relying on others to build needed tools.
Two frequently recommended books are "Emotional Intelligence" by Daniel Goleman, which emphasizes the importance of EQ in the AI age, and "The High-Growth Handbook" by Elad Gil, valued for its non-linear reference format allowing readers to jump to specific chapters for answers on recurring company-building problems.
"The Pit" is recommended as a series that makes high-pressure work feel less frantic by comparison, provides perspective on the gravity of decisions, and captures the ethos of Western Pennsylvania's blue-collar grinder mentality.
Granola is valued for its continued innovation and excellent product decisions with strong feedback responsiveness. WhisperFlow is used for quickly capturing notes to team members while on the go.
Three principles to live by and instill in children:
- Work hard, recognizing the value of hard work
- Do great work, distinguishing between effort and quality
- Be helpful, making time for founders seeking advice and answering emails
Raising children is humbling as they challenge you in unexpected ways and serve as mirrors reflecting your own behaviors. Despite the challenges, it represents the greatest joy, supported by an amazing partner who handles the unfair load.
The most enjoyable career moments often occur when slightly over your skis - thrown into unexpected situations and figuring them out. These periods involve ups and downs but become career highlights when viewed retrospectively.
The J-curve describes how the best roles involve joining and initially performing poorly before improving, creating the best professional experiences. This aligns with the pattern of growth through initial struggle.
Cursor's design and experience are exceptional, with the standout feature being the ability to switch between models extremely easily. This allows users to test new models immediately without switching products. The agent feature enables choosing the right model for specific tasks, addressing both performance optimization and cost considerations as models are released more frequently.
Cursor is training state-of-the-art models on SpaceX's data centers, representing the world's largest data center and unlocking new model capabilities.
Keep Lenny's Podcast in your library
Save the videos and channels worth coming back to, and find them again in one place.
