Kareem Zaki - Small Ideas Attract Competition - [Invest Like the Best, EP.392]
What this covers
Kareem Zaki has been a general partner at Thrive Capital for a decade, has co-founded three healthcare businesses worth over a billion dollars each, and has led the firm's investments in Ramp and Robinhood. The conversation comes as Thrive announces a new $5 billion fund. It starts with the arithmetic of scarcity: roughly 120,000 companies founded, around 2,000 worth over $1 billion, and only 100 to 200 truly category-defining. From there Zaki describes how that shapes Thrive's habits, including about 15 names holding most of a fund's capital and a small team whose trust lets partners seek candid challenge in the investment committee rather than permission.
The middle of the episode covers how he judges companies. He looks at market quality, asking whether a company is its customers' top vendor, and at founders with one uncompromising, almost mutation-like trait. He also checks unit economics to see whether a growth story survives the cheap-capital era. He argues that platforms compound through product surface area and data gravity, that sector allocation fails because industries move at different speeds, and that hard capital requirements can now be a feature. His AI forecast is contested: it will disappoint in the near term, as factories took decades to rework themselves around electricity.
The healthcare section is the most contentious. Zaki cites the system's cost, its local monopolies, and three PBMs controlling 80% of the market. He argues that startups should meet the system where it is, build big rather than ship small MVPs, and make committed bets. He also proposes ending employer-chosen health plans, following the pension-to-401(k) shift.
Kareem Zaki argues Thrive's edge comes from concentrating on the rare category-defining companies, which requires generalist, lifecycle, concentrated, and builder structures. He also argues that healthcare needs big products built inside the existing system and more consumer-level competition, not startups that work around it.
- Only about 100 to 200 of 120,000 companies founded in a period become category-defining, so spreading attention across good companies dilutes returns.
- Markets change at uneven speeds, so pre-allocating capital to sectors is less effective than staying flexible across tailwinds.
- Healthcare fails for small startups because of fragmentation and slow adoption, so big products that meet the system where it is are required to reach scale.
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Only about 10 to 20 companies a year can reach the scale needed to be category-defining, and this scarcity makes a good deal the biggest trap in venture because a good deal can look viable on paper while business building proves much harder.
“one of the worst traps in venture is a good deal, because you look at it and you think you can put it on paper”
Startups are disadvantaged in capital, resources, people, and time, so they need the market to be changing so incumbents with calcified advantages have less ability to compete; this is why investors should look for the change catalyst in established markets, as Amazon's advantage over Barnes and Noble and Netflix's over Blockbuster illustrate.
“They need change, and so we really pay attention to what is the change catalyst that's going to drive it in a given market.”
The most important healthcare change would be to remove the employer from choosing health plans, a arrangement that originated in World War II wage freezes, so that consumers can choose individual plans and create competition.
“the biggest thing we can do is honestly get the employer out of healthcare.”
In the short to medium term, AI will disappoint because the world is not ready for it, just as electricity took decades to reshape manufacturing, since factories built around steam power needed to be redesigned into individual workstations before the new technology paid off.
“I think in the short to medium term, I think it's inevitable that we're disappointed. And I think the big reason is because the world's not ready for AI.”
In the cheap-capital era, many companies spent a dollar to earn ninety cents, and no amount of growth covered that loss, so checking unit economics, retention, and cross-selling is needed to test whether a growth narrative holds.
“these businesses famously were spending a dollar to get ninety cents, and no amount of growth was going to cover what was going to happen there”
US healthcare consumes over 25% of the federal budget, growing faster than inflation, costs employers more than R&D, and is the leading cause of consumer bankruptcy, yet quality is not better than the best countries in the world despite far higher spending.
“I think we spend over 25% of the federal budget on healthcare, and that continues to grow faster than inflation.”
Healthcare is local with two or three insurers and one to three health systems per market, three PBMs control 80% of the market, and IT providers are concentrated, so the absence of competition removes urgency to change and fails to reward good behavior.
“when you don't have competition, you don't have an urgency to change.”
Tech companies that build a platform from a dominant position compound value through product surface area, data gravity, and workflow engagement, and traditional finance investors often underappreciate this strategic placement compared with reading a P&L.
“the opportunity for a lot of these tech companies to take a product and build a platform”
A high-quality market has customers for whom the company is their most important vendor rather than number five or seven on a priority list, because being a top priority gives a company the right of first refusal to build more products over time.
“Are you the most important thing that these customers work with? Or are you number five or seven on a priority list”
Because US healthcare is a $4.5 trillion system, roughly the size of Germany's economy, it cannot be rebuilt from first principles on a startup timeline, so healthcare companies should build within the system, get paid like it, and evolve it over time.
“Go build within the system, go understand the system, and then help it evolve over time.”
The classic startup MVP approach, which launches a small product to test a market, does not work in healthcare, because a small piece of a fragmented process cannot change the consumer experience, so companies must build big products that carry the full experience.
“the only way in this moment of time to move the needle on healthcare is to go after big problems and build big products.”
Venture's accepted math, in which 70% of companies fail and 30% pay for them, does not work in healthcare because health systems and insurers cannot integrate with startups that may disappear in two to three years, so healthcare investors must make larger, committed bets.
“if 70% of the startups they work with are gonna be gone in two to three years, you just can't get any momentum of innovation.”
Moving employer health benefits to individual accounts, analogous to shifting from pensions to 401(k)s, would make consumers aware of costs; giving someone a $15,000 health budget rather than a $2,000 deductible on a $15,000 plan would show them what care costs and let them choose hospitals accordingly.
“they went to 401(k)s. So employers used to be responsible for your retirement—that was pension funds.”
Investors underappreciate the scale that category-defining companies can reach, as shown by the fact that a trillion-dollar company was unimaginable a decade ago and a $3 trillion company was dismissed five years ago, yet three such companies now exist.
“I think we're underestimating the impact and the transformation and the returns to scale of a lot of these companies”
Technology does not transform every industry at the same speed or at the same time, so pre-allocating capital to sectors is a mistake; social networks produced major companies from 2005 to 2012 and then few large ones, while consumer fintech boomed from 2010 to 2015 and then slowed.
“we don't think technology moves uniformly through every industry at the same speed at the same time”
Concentration produces higher conviction and better decisions than the classic venture approach of making 100 bets and hoping three work, and in most Thrive funds the vast majority of capital sits in about 15 names.
“we think concentration drives conviction, which is better decision-making”
Investors' gut reactions to $50 billion, $100 billion, or $200 billion valuations anchor them to the past, while the true underwriting question is how early technology transformation still is; SpaceX opened a new frontier with Starlink, and OpenAI should not be capped just because its current products are impressive.
“we're anchored to the past. And the reality is we're underestimating the ability for technology to transform every industry.”
Fast growth is the first signal of a potentially great market, but durability of the tailwind and quality of the end market determine whether it lasts, because a company cannot escape the genetics of its market.
“you can't escape the genetics of the business at the end of the day. And the genetics of the market.”
Category-defining founders often have an unreasonable, almost mutation-like uncompromising trait in one dimension, such as Stripe's push for simplicity, Ramp's 5-hour customer-request turnaround, Costco's value focus, and Robinhood's refusal to bundle robo-advisors without shock value; but having too many such traits makes a founder hard to work with.
“they almost have an unreasonableness, like a mutation in them that makes them uncompromising around some dimension”
Because building software became so easy and capital was cheap, many companies were funded for trivial products, but the market is now shifting toward real problems and capital-intensive businesses, where a hard capital requirement can be a feature because it creates less competition.
“we're realizing that a hard thing, a capital thing, might actually not be a bug but a feature of a great opportunity”
Investment committees are often more political than truth-seeking, but Thrive's small team with deep trust lets partners go into the room knowing they can make a deal, so they can invite candid, unfiltered feedback rather than seek permission.
“we're inviting participation in our journey to go make the best decision”
Of roughly 120,000 companies founded in the period since Thrive's founding, about 3,000 are worth $500 million and about 2,000 are worth over $1 billion, but only 100 to 200 are truly category-defining; spending equal time on the rest would mean spending about 95% of effort on non-defining deals.
“if you spend equal time trying to chase those opportunities as those 100 or 200 companies, you'd spend 95% of your time on those deals”
Embedded portfolio support such as interim CFOs or growth officers is most valuable for the fastest-growing companies, which outgrow their systems and people, rather than for struggling companies, because chaos from rapid growth is the common problem.
“We found the opposite—it's the fastest-growing companies who need this the most.”
Technology companies that start with one product can compound by layering products onto their surface area, data gravity, and workflow engagement; GitHub's dominance in developer workflows made it a strategic platform for later cloud products, even though developer growth was forecast at only 5%.
“you just sat in a place where code was being created and collaborated on. What are all the downstream things that are going to be created from that?”
Moving health insurance from employer-selected plans to individually funded accounts would follow the pension-to-401(k) shift, in which employers stopped guaranteeing retirement and gave workers money to manage, which could create competition and new companies.
“I think if we can move to a world like that on the healthcare side, I think it would be one of the biggest unlocks”
If an idea does not resonate with a generalist audience outside healthcare or financial services, the founder is probably thinking too small, because the best ideas are big and simple enough to captivate people who have not spent their lives in a market.
“if I'm pitching something and it's not resonating with a generalist crowd, it probably means I'm thinking too small.”
A pitch deck shows a company at its most polished, but being a partner inside the company reveals the real core engine; when metrics are strong despite operational imperfection, that signals a market pulling the company forward faster than it can execute.
“when you're looking at a pitch deck, you're looking at a company at its prettiest moment.”
Fast growth is the first warning of a potential overvaluation trap, and the real test is durability of the tailwind and quality of the end market, since many tech companies were automatically assumed great and some were funded on the hope that scale would eventually produce money.
“what's the durability of this tailwind, and what's the quality of the end market?”
A category-defining founder combines a big vision that rewires how investors see the world with the ability to lay out concrete steps, since a big vision without a plan is just a fantasy, as SpaceX's two-decade sequence from orbit launches to Starlink shows.
“And a big vision without a plan is just a fantasy.”
Small ideas that look tractable attract many competitors and lots of capital, so big ideas with few competitors are more defensible, a point Zaki repeats throughout the conversation.
“I know I keep repeating it, but the small ideas attract a lot of competition.”
When a company's metrics look amazing even though operations are imperfect, that signals a strong market pull that gives the company margin for forgiveness, so the market will carry it even if the founders are not perfect.
“when things aren't going perfectly but the metrics look amazing, that to us is a great signal because there's such a pull in the market that you didn't have to be perfect”
Thrive's GitHub investment rested on the insight that owning the developer community where code is created and collaborated on would generate many downstream products, even though forecasts projected only about 5% developer growth.
“Everyone was forecasting, I don't know, 5% growth for developers.”
No company is big enough to create its own wave, so great companies ride big enduring tailwinds that can last decades, such as Stripe riding the 30-year e-commerce trend.
“no company is big enough to create its wave or create a market; great companies ride these big, enduring tailwinds”
Thrive's impact, finance, talent, data, and communications teams add richer context to investment decisions, and executive chairman Nitin's experience predicted that the corporate-card market would consolidate to three players, one acquired, which happened with Brex, Ramp, and Divvy.
“It's going to be three companies, and one of them is going to get acquired.”
Thrive wins deals by reaching founders early, building relationships, and adding value before investing, because established firms like Sequoia, Kleiner, and Benchmark will beat a newcomer who shows up only at a late stage.
“If we showed up in a process with Sequoia and Kleiner and Benchmark and they're meeting us for the first time, zero chance.”
Being fine with meaningful bets on good-but-not-iconic companies, such as those aiming for 2.5x or 3x returns, dilutes an investor's thinking and attention, which reduces their ability to identify the rare iconic companies that will reshape an industry.
“if you're fine just investing meaningfully in a company because it might be a 2.5x or it might be a 3x, it actually dilutes your thinking and your aperture”
A pitch deck shows a company at its most polished moment, while being a partner gives investors access to the raw operating reality, so lifecycle relationships reveal information that new investors cannot see.
“when you look at a pitch deck, you're looking at a company at its prettiest moment”
Early AI opportunities based on being first to apply new tools to existing problems are fleeting arbitrage moments that disappear as others learn about the technology, so they are unlikely to create lasting defensibility.
“Those are, I think, fleeting arbitrage moments. And to your point, if we think about the arc of time, might not be defensible.”
Thrive hires people with a maverick mentality, meaning they challenge the status quo, combined with high horsepower, meaning they work hard, have high IQ and EQ, and deliver what they promise, because the firm needs to mirror the mavericks it backs.
“If you've always done the default thing in your life, if you've never really challenged the status quo, you're probably not a fit at Thrive.”
Small, tractable ideas look attractive but attract heavy competition, while big ideas with large market potential face far fewer competitors, so the firm pursues ambitious ideas where only perhaps 10 to 20 companies a year can reach scale.
“sometimes these small ideas, while they seem really tractable, small ideas have lots of competition.”
A company's DNA is set in its founding moments when the team, vision, and prioritization are chosen, so investors need to understand the whole lifecycle of a business to be useful thought partners to founders who plan decade-long arcs.
“the DNA of the company is really set in those founding moments where you set the team, you set the vision, you set the prioritization”
Thrive's approach rests on four pillars: being generalists across sectors, being lifecycle investors across stages, concentrating capital in a few companies, and adopting a builder mentality that includes co-founding companies and building Thrive itself.
“it falls into four pillars. The first is that we're committed to being generalists; we're committed to being lifecycle investors; concentrated in our investments; and builders.”
Revenue growth does not by itself mean a company is solving a real problem, so investors should ask how the world changes if the company succeeds and whether an outsider would be excited by it.
“revenue doesn't mean you're solving a problem”
Multidimensional people who connect different fields, such as Roger Federer's soccer background or Tom Brady's baseball background, tend to produce elite and novel outcomes, which David Epstein's work in Range supports, so Thrive looks for generalist range in hires.
“If you want to be predictably good, specialize. If you really want to be the elite, the cutting edge of things that have never been seen before, it usually comes from these generalists.”
Specialists are predictably good, while generalists with varied experience are more likely to reach the cutting edge; David Epstein's Range argues that Nobel laureates are 20 times more likely than other scientists to have serious hobbies, and that connecting worlds lets people see new opportunities.
“If you want to be predictably good, specialize. If you really want to be the elite, the cutting edge of things that have never been seen before, it usually comes from these generalists.”
Investors undersized Uber's market because its rapid growth showed it was not constrained by an assumed billion-dollar taxi market, and the broader transportation market was far larger, so sizing new markets by existing spend is often wrong.
“everyone undersized the market, because, again, these new markets are hard to predict”
Watching a team execute across multiple acts, meaning successive product launches and scaling, gives investors much more confidence than a first-time meeting or a pitch deck can provide.
“The second thing is usually multiple acts. Seeing how the team's executed, how they're moving towards those, their ability to launch new products, scale new products”
Healthcare founders should build for the rest of America, such as Ohio, rather than New York and San Francisco, because only products serving about 100 million people can move the needle, and an early Headway data point showed that the MTA, the largest New York employer, signaled a problem common to all workers.
“We got to build for Ohio, where I grew up. We got to build for the rest of America.”
Zaki's family immigrated from Egypt because as Christians they faced restrictions that limited his father, a leading surgeon, from advancing to professorships, which shaped Zaki's belief that one should not put a ceiling on ambition.
“My dad was a leading surgeon and doctor, and you can do a good job as a Christian, but you can't be the best.”
A venture firm that claims people are its most important asset must develop its own talent by hiring people with zero to four years of experience and molding them, rather than outsourcing talent development and hoping to poach experienced people later.
“we really focus on hiring people who have zero to four years' experience, who have the raw talents and abilities”
An investor who does not deeply understand a business during diligence will not do so after investing, so the depth of an investor's diligence questions about a founder's market and roadmap is a reliable signal of future partnership.
“If an investor doesn't take the time to really understand your business in a diligence process, I can guarantee they're not going to do it after they make the investment.”
The big breakout AI businesses in the short term will be new categories that lack the friction of transforming existing workflows, and Thrive partner Vince believes the first $100 billion AI company will be a consumer company.
“Vince, our partner, thinks the first hundred-billion-dollar AI company is going to be a consumer company for that reason.”
Transforming traditional sectors with AI requires combining technology with operations, processes, and incentives, so Thrive partnered with founders to acquire accounting firms and rework their workflows rather than only selling software.
“It's going to be operational, it's going to be incentives, it's going to be processes that are actually going to drive the change.”
Thrive co-founds companies internally, with Kareem personally involved in starting four companies, and this builder mentality makes the firm better thought partners to founders because its investors have lived through company creation rather than only reviewing decks.
“I've personally been a part of starting four companies here at Thrive, and that's really helped me mature and be a better thought partner”
Thrive's New York base produces an East Coast and West Coast blend of independent thinking and creativity, and investment conversations should imagine the company five to ten years out, asking how the S-1 would read and how incumbents would respond.
“Even though we're looking at seed and Series A companies, we're trying to imagine the world five or ten years from now”
Thrive's increase in fund size allowed it to lead a $7 billion round for Stripe after COVID with nearly $2 billion invested between the firm and its LPs, and to lead a Ramp round, showing that the strategy stayed constant while capacity grew.
“we were able to support and lead a $7 billion round right after COVID; we invested nearly $2 billion between us and our LPs”
Ambitious young founders should choose problems that matter and that they would work on for a decade, rather than something buildable in a month, because deep problems attract better talent and create lasting value.
“what do I want to work on for a decade? What do I have the best shot of recruiting the best talent to go execute on?”
Zaki's Christian faith drives his intensity because he believes God placed him in his opportunities, and grace toward imperfect colleagues makes intense work more enjoyable and agreeable.
“if I believe that this is where God's put me, this is the opportunity He's put me in, then it's on me to make the most of it.”
The firm's narrow focus on category-defining companies allows it to reject most opportunities, and its concentration lets it be a committed partner, so founders respect that Thrive is not making many investments and sees its yes as a commitment for the whole journey.
“it's a sign of our passion and commitment. And I think they want to be picked in that way.”
Stripe holds roughly 2% to 3% global market share in a payments market where e-commerce is only about 20% penetrated, so Stripe could plausibly grow its share to 5% or 8% as online commerce expands.
“e-commerce's 30-year trend—we're 20% penetrated”
Thrive audits its calendar each quarter, and roughly 80% of meetings are with early-stage companies, because the firm must spend most of its time early to discover breakouts even though most capital is deployed at later stages.
“usually every quarter, probably 80% of our meetings are early stage”
A founder's primary motivation is to solve a problem rather than to optimize the P&L, so Thrive stays focused on product and customer narrative even at later stages, which keeps founders engaged and makes concentrated commitment resonate.
“They didn't start their company because they loved the P&L. They started the company because they wanted to solve a problem.”
An eighth-grade teacher who insisted that Zaki attend a different school, St. Xavier in Cincinnati, changed the trajectory of his life, leading to college and everything after.
“I do think it honestly changed the trajectory of my life and set me on a different path”
Category-defining companies often build quietly for years, as SpaceX did for two decades and OpenAI did since 2015, and then have breakthrough moments that leave competitors scrambling to catch up.
“companies that have a long-term vision, they're going after these crazy problems”
Robinhood's founders rejected bundling robo-advisors into their product because they believed any new offering must shock consumers, which led to zero-dollar trading and later a 3% cash back credit card.
“no, if we're going to do something, it's got to shock the consumer.”
Thrive's deliberation starts qualitatively with a ten-year narrative, imagining S-1 filings and how incumbents would respond, then tests whether the investment returns justify its risk quantitatively.
“How's the S-1 going to read? How are the big incumbents going to respond?”
When an investor treats a company as one of many bets, the founder feels that the investor is not fully in it, whereas a concentrated investor shares the highs and lows and is committed in a way that goes beyond capital.
“to have an investor who's like, 'This is one of a bunch of bets; if it works, amazing; if it doesn't, I'll be fine,' I think being concentrated binds us a little bit closer to the founder”
Moving through eight schools growing up taught Zaki that new environments do not faze him, and that the first month without friends is temporary, which forced him to become proactive about building connections.
“I ended up going to eight different schools growing up.”
Strategy without structure does not allow a firm to push forward on its mission, so Thrive has built organizational structure to execute its category-defining strategy.
“Strategy without structure, I think, doesn't allow you to actually push forward on that mission.”
In its early days Thrive had to hustle to get meetings with founders, including emailing Stripe founders and showing up in Indianapolis, and sent a 100-page deck to GitHub, sometimes investing about a month of work to get one hour with a founder.
“You'd put almost like a month's work in just to get an hour with the founder.”
Thrive uses a data and analytics team to track new companies, traction, and consumer downloads, augmenting partners' judgment about trends and where founders are, a development from its earlier reliance on relationships.
“We have a whole data and analytics team that's really analyzing where founders are.”
The most surprising thing about Thrive's journey from a $40 million first fund to a $5 billion fund is how consistent its strategy has remained, centered on backing category-defining companies across stages.
“the most surprising thing about the journey is just how consistent the strategy's been”