Sam Corcos on Fundraising, Transparency and Intentional Time

Sam Corcos, co-founder and CEO of Levels, joins Harry Stebbings to discuss risk, transparency, fundraising, time and family. He recommends treating investor relationships as an ongoing learning process and making specific requests for help. Levels sells a health-related product, and his product claims and company figures are his own account from the December 2023 interview.

Corcos wanted to be a research scientist as a child. He later adopted a mobile lifestyle without a fixed home base and recalls arriving in New York without a place arranged, then staying with a friend he happened to meet. He sees room for spontaneity as a way to encounter experiences that a rigid itinerary might exclude.

Stebbings asks whether spontaneity conflicts with optimization. Corcos says the answer depends on the goal: a specific gallery may require advance planning, while meeting people and discovering unexpected places require open space. He describes a spontaneous road trip as an example of the latter.

Corcos argues that people often overestimate certain risks and offers claims about rare events and hitchhiking. The interview does not establish the statistics or a general safety conclusion; risk varies by circumstances. His broader point is that fear can make people avoid opportunities, but his examples are personal judgments.

He says his risk tolerance has increased as he has reexamined feared outcomes. Stebbings notes that founders can become more cautious as employees, customers and reputation depend on them. Corcos thinks success can paradoxically make people less willing to take risks, even when they have more resources.

On luck, he distinguishes random opportunity from preparation to recognize and use it. He believes both play a meaningful part. Looking back at prior companies, he also writes reflections on mistakes so that experience can inform the next venture.

At CarDash, Corcos was technical co-founder and CTO. He says he sometimes deferred to the loudest opinion on software decisions even when his own concerns proved valid. Over time he learned to trust his judgment while recognizing how much he still had to learn.

He also recalls disagreeing with the CEO and failing to fully support the chosen direction in front of the team. The CEO confronted him about it. Corcos’s lesson is to debate privately, then commit publicly or consider leaving if he cannot genuinely support the decision; Stebbings questions how easy that is for a co-founder on major matters.

Moving from technical co-founder to CEO changed the feedback loop. A developer can ship and see direct results; a CEO may influence many areas without a clearly attributable output. Levels holds a Friday forum to celebrate progress, which Corcos says supports morale even when his own contribution is less visible.

Levels has reduced titles to functional categories and adopted extensive transparency. Corcos says investor updates and all-hands meetings are public, while one-on-ones and performance information are shared internally by default; compensation is not. He explicitly says it remains to be seen whether the overall practice is a good idea.

Stebbings worries employees cannot be vulnerable when conversations are shared. Corcos says people can opt not to share personal discussions, but the default requires trust. The risks and benefits depend on the team and consent; his account describes Levels’ policy at the interview, not evidence it works for every organization.

They discuss Netflix’s “keeper test”: a manager imagines an employee leaving and asks whether they would fight to retain them. Corcos says uncertainty can prompt an honest conversation about improvement. The test is a management heuristic, not an objective measure of a person’s worth or a substitute for fair evaluation.

He recalls telling an employee that he would feel indifferent if they left, then sharing the conversation internally. According to Corcos, colleagues responded by describing the employee’s contributions and offering mentoring rather than turning against them. This is one company anecdote; it does not establish how the practice affects all employees.

Stebbings asks whether talent quality must decline as a company scales. Corcos suggests early startups may initially hire from a limited pool, attract stronger candidates with traction, and later face pressure to fill many standardized roles. He points to the keeper test’s use at larger organizations, but does not establish a universal scaling curve.

On rebuilding trust, he says it is possible but difficult. At an earlier stage Levels shared much information but kept individual performance and one-on-ones private; departures without clear context led some colleagues to distrust management, in his account. The company later increased internal sharing.

He acknowledges that broad sharing could fail in a political or low-trust culture and argues that hiring and maintaining good-faith relationships matter. Stebbings raises fundraising transparency as another risk: an announced round that takes longer can affect morale. Corcos believes credibility and consistent context help staff handle uncertainty.

He recalls a colleague’s experience at a company that disclosed little financial information and suddenly ran out of money. That story influenced Levels’ preference for informing employees about company finances. Corcos frames transparency as an experiment based on trust, rather than a risk-free policy.

Asked about retention, Corcos says the figure was around 25%, up from 20%, and that the company hoped to reach 40%. The interview does not specify the retention period or cohort, so the numbers cannot be interpreted as monthly or annual retention. He acknowledges it was not yet where the business needed it to be.

The product challenge, he says, is encouraging continued use after people learn initial patterns. He believes some long-term users value accountability and goal-setting more than individual discoveries. Levels sells the product, so his explanation is commercially interested and not independent evidence of health outcomes.

Stebbings proposes financial disincentives for missed goals. Corcos describes a wearable challenge in which participants committed money and could lose some when they missed a target, saying it had been effective in their experience. The interview gives no controlled outcome data or basis for a general health recommendation.

For fundraising, Corcos recommends casting a wide net rather than treating each investor contact as a single irreplaceable chance. An early meeting can establish a baseline, and a later meeting can show progress against a concern. His hypothetical increase from 25% to 40% retention is an illustration, not a reported achieved result.

He advises taking many pitches and using feedback to understand the market. Investors may have seen competitors a founder does not know. He distinguishes an inexperienced founder from a poor one, while emphasizing that meetings can reveal gaps in market knowledge.

Levels had approximately 100 angel and operator investors before institutional funding, he says. That broad network offered potential expertise but also overhead. He argues founders must actively seek help rather than assume investors know their priorities or criticize them for failing to read minds.

A useful investor update begins with knowing what the company needs. Levels includes an asks section, but Corcos finds targeted requests more effective than vague broadcasts. He tries to make each request specific, time-bounded and easy for the recipient to act on quickly.

For example, asking for introductions to the two best designers an investor knows may elicit a clearer response than asking for any good designer. Corcos estimates he sent more than 3,000 specific requests and received responses he counted as successful about half the time. Those are his internal estimates, not independently verified conversion rates.

Stebbings suggests a shared sales-target sheet with named contacts and ownership to make introductions easy and visible. Corcos adds that founders can identify mutual connections before asking. They see this as a way to reduce the work required of an investor, while acknowledging not every request will be fulfilled.

He recommends setting expectations before accepting capital: if the founder wants active help, say so and let an investor decline. Some investors promise engagement and then do not respond, he says. He believes some diligence is worthwhile, with its depth proportional to the check and expected role.

Corcos says Levels gained significant help from its angels and from a16z’s operating platform, citing access to more than 110 operating partners in an internal analysis. He praises a service directory that makes expertise discoverable, and also credits Trust Ventures for regulatory help. These are his assessments of investors with financial relationships to Levels.

His best early meetings were with investors who already understood the category. Corcos compares fundraising to finding people who share a vision rather than converting skeptics from scratch. He recalls sharing detailed material in advance and says some well-prepared investors read it closely.

Stebbings prefers keeping materials until a meeting so a founder can make a personal case before an investor rejects a category. Corcos prefers sending information ahead: time is finite, and an early no can save a meeting. Both are context-dependent approaches, not a rule that all investors will behave the same way.

On meeting associates or principals, Corcos says the value depends on how a firm actually makes decisions. He recalls a principal sharing material with a partner who later did not remember it. Stebbings distinguishes junior roles focused on research from those focused on sourcing; neither speaker claims this covers every firm.

With many small angel checks, Corcos wishes he had used a rollup vehicle to simplify signatures and administration. Minimum check size should reflect the fundraising target and time available, he says. A small check may still be valuable if the goal is a network of operators able to help.

He identifies early employees of post-IPO companies as useful angels: they may invest modest amounts but know talented people and have time to help. That potential only becomes useful if the company has a process for making requests and following up. His assessment comes from Levels’ experience.

Corcos recalls a pitch to Moshe Lifshitz in which his co-founder drank a green juice and showed a glucose increase during the meeting. The demonstration helped convey the product’s proposition to that investor, he says. A single demonstration is not clinical evidence about the drink, glucose data or health effects.

He recommends testing different pitches and learning which explanations resonate. The amount of theater depends on the business and stage: a future-oriented vision may benefit from a demonstration, while a later SaaS business may be judged more directly on economics and growth. He treats pitching as an iterative sales process.

Some investors did not provide the help they promised, according to Corcos. He says the company tracked requests and found some people had fulfilled none, but questions whether extensive diligence is worthwhile for every small check. For a board member, by contrast, he recommends much deeper references because the relationship is consequential.

Before accepting Jeff Jordan as a board member, Corcos says he spoke with founders who had worked with him and heard positive feedback. Stebbings asks whether experienced investors’ views can carry too much weight. Corcos says founders should listen carefully while remembering that both sides can be wrong.

On performance, Corcos first suggests delivering on stated objectives, then acknowledges that achieving the wrong goals is not necessarily success. He thinks fear of failure can lead people to choose goals that are too small. This is his framework for ambition, not a measurement of high performance.

His practical starting point for time management is to record where time actually goes. He discovered much more social-media use than he expected and used a blocker to interrupt the habit. A manual calendar audit or software can reveal gaps between intention and behavior.

He uses the calendar as a to-do list, blocking time for specific tasks. Because interruptions are inevitable, he recommends leaving buffer space and moving tasks forward when time becomes available rather than scheduling every minute. The right amount of open time depends on the variability of the job.

Preparing for parenthood, Corcos worries most about being present. He expects children to require more deliberate scheduling but acknowledges he has not yet experienced the change. He also describes a concern that a child can affect a marriage; the percentage he cites is unsourced in the interview and should not be treated as a reliable statistic.

He and his wife set aside a weekly date and a separate time for family priorities and unresolved logistics. Corcos sees this as intentionality, not a contradiction of spontaneity: sometimes the intention is to explore without a fixed plan, and sometimes it is to protect time for a relationship.

He says he once wrote a one-page description of the values he wanted in a partner, including open-mindedness. He treated dating as a matching process rather than a campaign to persuade someone. His six-month engagement-or-separation goal and early conversations were personal choices, not advice that every relationship should follow that timetable.

He recalls some dates ending quickly after he shared the document because both people recognized a mismatch. Stebbings jokes that he might instead try to persuade the other person. The example illustrates Corcos’s preference for directness and respect for time, as he describes it.

The same attention to fit matters for co-founders, he argues. He recommends knowing a potential partner through experience and references, and setting clear expectations about the CEO’s role. He recounts an unnamed allegation of misconduct in another company as an example, but the interview does not independently establish that account.

He does not believe equal founder equity is always required; the answer depends on contributions, expertise and what the group means by co-founder. He also recalls talking openly with co-founder Josh after feeling resentment over a message. The discussion revealed a misunderstanding and resolved it, according to his account.

Stebbings argues that difficult conversations should be timed thoughtfully rather than started when people are exhausted. Corcos agrees that “immediately” need not mean the same minute: waiting until the next day can be sensible, while allowing resentment to linger for months can be damaging. They agree on addressing issues within a reasonable time.

Corcos says being more present with his wife remains a personal challenge. He practices leaving his phone behind and returning attention to a conversation, even when work occupies his mind. He sees presence as a skill to practice, not a condition that must be perfect before spending time together.

In quick fire, he names strong hiring as one of Levels’ best resource decisions and poor hiring as one of its worst. He says its product organization became too large relative to engineering and has changed his mind about the role of product management. These are company-specific judgments, not a universal organizational ratio.

He worries about society’s difficulty agreeing on shared facts and says open-mindedness is important to him. Looking five years ahead from 2023, he hopes for a larger family, better retention and progress on Levels’ metabolic-health mission. Those are his aspirations as CEO, not demonstrated clinical outcomes.