a16z partner David Haber reviews: the selection criteria for the $1.7 billion Apps fund, institutional moats, and AI collaboration systems

David Haber, a partner at a16z, in-depth interview: From a $1.7 billion fund to a century-old investment institution, how do top venture capitalists choose entrepreneurs, build moats, and seize wealth opportunities in the AI era?

Why are some venture capital firms able to sustain for decades, even across multiple technological revolutions, while others decline with the departure of star investors? Why do top VCs increasingly resemble tech companies? When AI can analyze tens of thousands of companies and identify investment opportunities, how will the value of connections, experience, and business judgment change?

On January 9, 2026, Silicon Valley's top venture capital firm Andreessen Horowitz (a16z) announced the completion of a new fundraising round exceeding $15 billion.

Of this, $1.7 billion is allocated to application layer (Apps) investments, $1.7 billion to infrastructure, $6.75 billion for growth investments, and the remaining funds invested in national strategic technology, healthcare, and other venture capital strategies.

This is not only a massive financing.

It also represents an important change in the venture capital industry: top investment firms are no longer satisfied with relying on the judgments of a few star partners to make money, but are trying to establish organizational systems that can continuously generate investment advantages.

In an exclusive interview published on October 5, 2026, on "The Luba Show," a16z general partner David Haber shared his entrepreneurial and investment experiences and proposed a viewpoint worth considering for all investors, entrepreneurs, and founders:

Building a sustainable investment institution is more important than managing a successful fund.

He calls this concept "Firm over Fund," meaning the institution is more important than the fund.

This viewpoint actually involves several core issues in the modern financial industry:

How does capital form a long-term competitive advantage?

Why must top entrepreneurs possess strong sales capabilities?

How do personal relationships transform from individual resources into organizational assets?

Can AI convert decades of accumulated experience of an investment institution into a sustainable intelligent system?

More importantly, can ordinary entrepreneurs and investors find practical methods suitable for themselves from the operational logic of these top institutions?

1. David Haber's career experience: Why do top investors often need to span multiple industries?

David Haber grew up near the U.S.-Mexico border in San Diego, California.

His family has a complex history of multinational migration. His ancestors experienced wars and persecution in Europe, and family members later moved through Central America and Mexico, eventually establishing a life in the United States.

The intertwined experiences of different languages, cultures, and business environments exposed him to cross-regional and cross-community relationships early on.

This background does not guarantee business success, but Haber later developed the ability to build connections across industries into a professional advantage.

He attended Harvard University, studying biochemistry.

During college, he interned in an early technology team at Facebook and worked with Nathan Blecharczyk, who later became a co-founder of Airbnb.

At that time, Airbnb had not yet become a globally recognized accommodation platform.

Looking back, this experience reminds us of an easily overlooked issue:

Major business opportunities often do not present themselves in the form of "major business opportunities" before they arise.

Today, people consider companies like Airbnb, Uber, and Stripe to be obvious success stories.

But in their early days, there was significant uncertainty regarding product demand, business models, regulatory environments, and market sizes.

Identifying early opportunities requires not only knowledge but also sensitivity to emerging behavioral changes.

After graduating, Haber joined a team related to Royalty Pharma and later became a venture capitalist at Spark Capital.

During his time at Spark Capital, he participated in early investments in the fintech sector, including Plaid.

In 2014, he founded the fintech company Bond Street, providing online financing services for small and medium-sized enterprises in the U.S.

In 2017, Bond Street was acquired by Goldman Sachs.

Subsequently, Haber worked at Goldman Sachs in corporate strategy, partnerships, new business, and mergers and acquisitions.

In 2021, he joined a16z as a general partner and participated in promoting the firm's development in New York.

This career trajectory is worth studying.

It spans investments in biopharmaceutical assets, venture capital, fintech entrepreneurship, large investment banking, and top VC firms.

On the surface, these experiences belong to different fields.

But they actually revolve around the same question:

How are capital, technology, talent, and customers organized to create enterprises with long-term competitiveness?

At Royalty Pharma, he was exposed to asset cash flows and long-term capital allocation.

At Spark Capital, he learned how to identify the growth potential of early-stage tech companies.

At Bond Street, he personally took on entrepreneurial risks, facing customer acquisition, credit risk, and financing costs.

At Goldman Sachs, he encountered the organizational systems and capital operations of large financial institutions.

And at a16z, he began to think at a higher level about how to build investment capabilities into a system that can continuously expand.

Truly valuable cross-industry experience is not simply having multiple job titles, but being able to apply mature operational methods from one field to another that has not yet fully developed.

2. Why can a Royalty Pharma with only a few dozen employees establish a financial business worth hundreds of billions?

A very important part of Haber's career experience is his exposure to Royalty Pharma's business model.

Royalty Pharma was founded in 1996 and focuses on biopharmaceutical royalty revenue investments.

Traditional pharmaceutical companies need to invest huge amounts of money in drug research, bearing the risks of clinical trials, regulatory approvals, and commercialization.

Even if research and development are ultimately successful, it may take years to achieve stable income.

Royalty Pharma adopts a different business model.

It acquires or invests in drug royalty revenues to obtain a portion of the future sales revenue of related drugs.

Assuming a biotechnology company has the revenue-sharing rights for a certain drug.

According to the contract, the company can receive 5% of the drug's sales over the next ten years.

If the expected annual sales of the drug are $2 billion, the corresponding annual royalty revenue would be $100 million.

However, these revenues are not without risk.

Drugs may face competition, patent expirations, declining sales, regulatory changes, or new treatment technologies that replace them.

Royalty Pharma needs to value future cash flows and purchase related revenue rights at reasonable prices.

If the payment price is lower than the risk-adjusted value of future cash flows, it may create investment returns.

This is different from ordinary stock investments.

It is closer to specialized capital allocation around the future cash flows of specific assets.

Haber mentioned in the interview that Royalty Pharma once supported a business with a market value of about $25 billion with a team of about 50 people.

This figure reflects a specific historical stage discussed in the interview, rather than a fixed number of employees or current market value.

What is truly important is not how many employees there are, but why the company can manage a large amount of financial assets with a relatively lean team.

The answer comes from three aspects.

First, the assets themselves can generate cash flow.

Second, a professional team is responsible for asset selection, risk assessment, and capital allocation.

Third, the company can continuously apply accumulated experience, funding sources, and investment capabilities to new transactions.

Here is an important financial concept:

Capital-intensive businesses do not necessarily require labor-intensive organizations.

Some financial institutions can manage hundreds of billions of dollars in assets without needing to employ large numbers of production staff like manufacturing companies.

However, a large asset management scale does not mean that shareholders will necessarily receive high returns.

If investment prices are too high, funding costs rise, or underlying cash flows deteriorate, large-scale assets can also incur losses.

The important insight that Royalty Pharma brought to Haber is that a financial institution can establish long-term competitive advantages around unique asset acquisition capabilities, professional judgment, and capital structure.

This also became an important foundation for his later thinking on "Firm over Fund."

3. The real difference between Fund and Firm: Funds focus on single returns, while institutions focus on long-term compounding

On January 12, 2026, Haber published an article titled "Firm > Fund," systematically explaining his investment institution philosophy.

Here, it is necessary to first understand the basic structure of venture capital funds.

Traditional VCs are usually managed by general partners (GPs) and funded primarily by limited partners (LPs).

LPs may include pension funds, university endowment funds, family offices, foundations, and other institutional investors.

GPs are responsible for finding startups, conducting due diligence, investing, and post-investment management.

After a successful investment, GPs typically receive management fees and performance compensation.

Performance compensation is usually referred to as Carry, which is a certain percentage of the investment returns that meet distribution conditions, as stipulated in the fund's contract.

For example, if a fund raises $1 billion.

Assuming the total distributable value returned to investors reaches $3 billion, temporarily ignoring management fees, return distribution thresholds, and other contractual arrangements.

The nominal investment appreciation created by the fund is $2 billion.

If a simplified Carry rate of 20% is applied, the performance compensation for the GP may be $400 million.

The actual distribution still needs to be calculated according to the fund agreement.

This model drives fund managers to focus heavily on investment returns.

This in itself is not a problem.

Because the main responsibility of GPs is to create reasonable risk-adjusted returns for LPs.

But Haber argues that many investors only focus on how to improve the current fund's returns, neglecting whether the investment institution itself can accumulate competitive advantages over the long term.

He believes that a true Firm needs to achieve two goals simultaneously.

First, create excellent returns for investors.

Second, establish competitive capabilities that can continuously enhance over time.

The difference between the two is profound.

1. A fund may rely on individual talent

A certain investor possesses excellent business judgment, able to identify potential in early-stage companies and achieve high investment returns.

But if all investment decisions rely on this person, the institution faces obvious risks.

Once he retires, leaves, or makes a judgment error, the fund's future competitiveness may decline rapidly.

This is known as key man risk.

The venture capital industry has long relied heavily on individual experience, connections, and judgment.

Because early-stage startups often lack stable profits, mature financial statements, and long-term operational data.

Investors need to assess the entrepreneurial team, technology trends, product demand, and highly uncertain future markets.

Such decision-making is difficult to fully standardize.

But if an institution can only rely on a few people's intuition, it is hard to form truly replicable organizational capabilities.

2. A Firm must have competitive advantages that transcend individuals

Haber suggests that enduring financial institutions need to continuously build moats.

For example, Apollo accumulates competitive advantages through long-term capital structures, credit investment capabilities, and insurance-related businesses.

Goldman Sachs forms a complex business system through institutional client relationships, wealth management channels, capital market services, and global trading networks.

Quantitative investment institutions like Renaissance Technologies and D. E. Shaw invest in data, computational technology, and research systems over the long term.

The competitive advantage of these institutions does not come from a single outstanding employee.

It also comes from the long-term accumulation of assets, processes, talent, and technology within the organization.

This is the core distinction between a Fund and a Firm.

The former may excel at investing.

The latter not only excels at investing but also understands how to continuously improve its investment capabilities.

An excellent fund creates one or more exceptional returns; an excellent institution must strive to establish a system that can continuously generate excellent returns.

Of course, scaling is not a natural advantage.

Large institutions may also experience bureaucracy, slow decision-making, conflicts of interest, and imbalanced incentives.

Therefore, institutional management only has economic value when it increases competitive advantages in the long term, rather than simply increasing costs.

4. Why is a16z different from traditional VCs? Because it treats the investment institution as a technology company.

Andreessen Horowitz was founded in 2009 by Marc Andreessen and Ben Horowitz.

Marc Andreessen was involved in creating the early internet browser company Netscape, and later co-founded the software company Opsware with Ben Horowitz.

Both have real experience in running technology companies.

This also influenced a16z's organizational design.

Traditional VCs often consist of a few investment partners and a streamlined support team.

In contrast, a16z chooses to build a large platform organization to provide recruitment, marketing, sales, operations, and other support services to its portfolio companies.

Haber mentioned in an article in January 2026 that a16z then had about 400 platform personnel dedicated to helping portfolio companies grow, and stated that the platform required hundreds of millions of dollars in operating expenses each year.

These belong to its publicly described institutional operating conditions and cannot be directly interpreted as investment performance or service effectiveness that has been independently verified.

Why would an investment institution be willing to bear such high operating costs?

Because it attempts to change the competitive landscape of the VC industry.

Traditional VCs primarily compete on capital, judgment, networks, and brand.

a16z hopes to further sell a comprehensive service:

capital plus talent networks, capital plus business resources, capital plus market influence, and capital plus organizational capabilities.

Suppose two VCs both wish to invest in an excellent startup.

The first VC is willing to invest $10 million.

The second VC is also willing to invest $10 million but can provide a broader recruitment network, customer referrals, market dissemination, and operational consulting.

Given similar valuations, control, and other terms, entrepreneurs may be more inclined to choose the second one.

Because for entrepreneurs, funding is just one resource for building a business.

Customers, talent, technology, and brand are equally important.

If a VC can help a startup grow faster, it may increase its chances of entering excellent projects.

This creates a potential positive cycle:

The stronger the institutional platform, the more likely it is to attract outstanding entrepreneurs.

The more willing outstanding entrepreneurs are to collaborate, the more opportunities the institution has to obtain high-quality investment projects.

The richer the investment cases and industry relationships, the more the institution's brand and network may be further enhanced.

This is a network effect built on organizational capabilities.

However, this cycle does not automatically establish itself.

If the platform team is large but cannot actually improve the operational results of the invested companies, then high operating costs may erode the profits of the fund management company.

Therefore, whether a16z's platform model is successful ultimately still needs to be verified by investment returns and real business results.

5. Why do top VCs increasingly need to become technology companies with their own products?

In the past, the core production tools for investors were mainly phones, emails, financial models, meeting minutes, and personal experience.

A large amount of investment judgment existed in the memories of partners.

For example, a senior investor may have known the founder of a certain company ten years ago.

He knows that this founder has failed several times, understands the capabilities of their team, and is aware of significant changes that have occurred in a certain industry in the past.

But other partners may not necessarily have this information.

This creates a typical organizational problem:

The company possesses a wealth of knowledge but has not formed reusable knowledge assets.

This issue is particularly severe for VCs.

Because an investment institution may come into contact with thousands of startups each year.

But the companies that are ultimately invested in only account for a small portion of that.

If the information about non-invested companies is not effectively preserved, the institution may lose its ability to observe long-term development trends.

For example, a startup may have only 20 employees in 2023 and has not yet found a stable business model.

In 2024, revenue begins to grow.

In 2025, it gains important corporate clients.

In 2026, its product enters a rapid expansion phase.

If the investment institution can continuously record and associate these changes, it may be able to identify opportunities earlier than investors who rely solely on the latest financing news.

This also explains why Haber began to try to build an AI system.

6. Vantage: How David Haber turned 750 meetings into an AI investment system?

In this exclusive interview, Haber introduced the AI tool Vantage that he developed.

According to the interview, he used records from about 750 meetings as the early data foundation to transform a large amount of unstructured information into searchable and analyzable organizational knowledge.

These meetings may involve entrepreneurs, business operators, industry experts, and other business professionals.

In the past, this information was scattered across meeting minutes, recordings, emails, and personal memories.

Now, large language models give institutions the opportunity to automatically organize this content, extracting people, companies, viewpoints, relationships, and key business changes.

For example, the system can attempt to answer:

Which partner has previously interacted with this company?

How did the founder previously explain their business model?

What changes occurred in a certain industry over the past year?

Which companies are showing new signs of growth?

Which internal expert knows the most about a specific niche market?

The value of such a system is not just to help investors save time reading meeting minutes.

The deeper value lies in gradually transforming the business experiences that were originally scattered in individual minds into information resources that the entire institution can access and utilize.

1. From individual memory to institutional memory

Suppose a VC has 20 investment personnel.

Each person participates in 300 effective business communications per year.

Then the entire team may accumulate 6,000 communication records each year.

In ten years, that’s 60,000.

Not all of this data has the same value, nor can it all be legally preserved and shared.

But if authorized, filtered, and governed, some of it may form a very valuable historical knowledge base.

AI can help establish connections between companies, people, industries, and events.

For example, has the strategic viewpoint proposed by the same entrepreneur changed over time?

Has a company achieved the growth targets it previously committed to?

Has a technology trend moved from the laboratory to the commercialization stage?

This has potential value for long-term investment research.

2. Why might internal data become a new competitive barrier for VCs?

Two VCs can purchase the same business database.

They can also use the same large language model.

But they may not have the same historical communication records, investment cases, industry relationships, and professional judgments.

An institution that accumulates high-quality internal information over the long term may possess context that is difficult to replicate from public databases.

This type of advantage can be understood as organizational knowledge capital.

However, it must be emphasized that data can only generate reliable value if it is properly governed.

Otherwise, meeting minutes may contain errors.

Personal evaluations may be biased.

Historical viewpoints may be outdated.

Some information may also be subject to confidentiality agreements, privacy requirements, or internal access restrictions.

Therefore, a successful institutional AI system not only requires a powerful model but also strict data permissions, source tracking, and auditing mechanisms.

3. AI can discover opportunities but cannot automatically prove that opportunities are worth investing in.

Suppose a system detects that a startup's number of employees, customer count, and financing activities are all increasing simultaneously.

It may flag this company as worthy of further research.

But this does not directly prove that the company has investment value.

Employee growth may indicate business expansion, but it may also mean cost overruns.

Increased financing may reflect investor confidence, but it may also indicate that the company continues to rely on external capital.

An increase in customer count may come with a decline in revenue quality.

Therefore, AI is more suitable as a research support system rather than an unmonitored final investment decision-maker.

Ultimately, due diligence, financial analysis, market research, and legal review are still required.

The most valuable lesson from Vantage is not that AI can make decisions for investors, but that AI may significantly enhance the ability of investment institutions to utilize historical information.

7. AI is changing the economic structure of venture capital: Will information advantages disappear?

The traditional VC industry has a clear information asymmetry.

Investors discover startups that ordinary market participants have not fully focused on through interpersonal networks, industry experience, and professional research.

But AI and business databases are lowering some of the information search costs.

For example, in the past, investment analysts might need several days to organize information about competitive companies, financing information, and product situations in an industry.

Now, AI tools can assist in generating preliminary research frameworks in a shorter time.

This means that the ability to organize basic information may become increasingly difficult to form a long-term moat.

However, investment opportunities do not necessarily disappear as a result.

Because the key to venture capital is not just acquiring public information.

It also includes obtaining investment opportunities, assessing business quality, winning founders' trust, and completing transactions at reasonable valuations.

VC competition can be broken down into four different stages.

First, discovering excellent companies.

Second, assessing the company's future development potential.

Third, persuading entrepreneurs to accept their investment.

Fourth, helping the company grow and achieving an exit at the appropriate time.

AI may reduce some costs in the first two stages.

But the third and fourth stages still heavily rely on interpersonal relationships, business resources, and long-term trust.

There may even be a new industry change:

As the information analysis capabilities of all investment institutions gradually improve, high-quality entrepreneurial projects will attract more competitive capital.

When more investors discover the same excellent company, the investment price may be driven up.

This means that increased information efficiency does not necessarily increase the investment returns of all VCs.

Sometimes it may even compress excess returns.

For example, if a company's future exit value is $1 billion.

If investors initially invest at an $100 million valuation, the potential return space is significantly different from investing at an $800 million valuation.

Even if the same successful company is ultimately invested in, the purchase price will affect the investment outcome.

Therefore, top VCs must possess the ability to discover, judge, and secure deals simultaneously.

AI can improve research efficiency but does not necessarily replace the latter two competitive advantages.

8. How do top VCs assess founders: Why is "Why you?" more important than a business plan?

Haber discussed the qualities of entrepreneurs that he values in the interview.

Among them, the most worthy of study are clarity of thought, historical awareness, mission-driven motivation, and proactive execution ability.

1. Why you?

When investors evaluate a startup, they cannot just ask whether the market is large enough.

Because a huge market usually attracts a lot of competitors.

The more important question is:

Why can this team achieve a competitive advantage in this market?

For example, suppose an entrepreneur wants to develop a medical AI product.

If his explanation is simply, "AI is very important for the future, and the medical industry has a huge market size," that does not prove he has a special advantage.

However, if the entrepreneurial team has ten years of experience in hospital information systems, understands clinical workflows, can legally obtain high-quality data, and has already established partnerships with medical institutions, the investment logic becomes more specific.

This involves an important concept:

Founder-Market Fit, which refers to the degree of fit between the founder and the market.

An excellent founder does not necessarily have to come from the industry in which they operate.

But they need to possess a unique ability that can help them solve specific problems.

This ability may come from technology, experience, customer relationships, or a deep understanding of user needs.

2. Why now?

The success of entrepreneurship depends not only on ideas but also on timing.

Historically, many great business ideas have failed because technology, costs, and infrastructure were not yet mature.

For example, early internet business models were constrained by internet penetration rates, payment systems, and logistics capabilities.

The large-scale development of mobile internet required the simultaneous maturation of smartphones, mobile networks, app stores, and payment technologies.

AI entrepreneurship is no different.

An AI application that could not be commercialized in the past may not have been due to a lack of demand, but rather because the model costs were too high, accuracy was insufficient, or there was a lack of necessary data interfaces.

When technological constraints change, business models that were previously unviable may become feasible again.

But entrepreneurs must clearly explain:

What key condition has changed?

Is it the decrease in computing costs?

Is it the improvement in model capabilities?

Is it regulatory approval?

Or are customers now ready to pay for new products?

If they cannot answer this, they may fall into the trap of chasing popular trends in entrepreneurship.

3. Why did past people fail?

Haber emphasizes the importance of entrepreneurs understanding the history of their industry.

This is very important.

Some entrepreneurs like to claim:

"Previous companies did not understand this market, and our approach is completely different."

But past companies may have had excellent teams, ample capital, and rich experience.

They failed possibly because there were real structural barriers in the market.

For example, high customer acquisition costs, complex regulatory requirements, unviable unit economic models, or consumers simply unwilling to change their existing behaviors.

If new entrepreneurs do not study the reasons for these failures, they may repeat the same mistakes.

Truly mature entrepreneurs should know:

Who has tried to solve this problem before?

What methods did they use?

Why did they fail?

What key conditions have changed today?

What difficulties still exist?

This is the importance of historical research for entrepreneurship.

History not only provides stories but also helps identify industry constraints that are easily underestimated by new entrepreneurs.

Nine, the lessons from Bond Street's failure and success: Huge financing needs, why is customer acquisition still difficult?

Haber's experience founding Bond Street is an excellent case for understanding fintech business models.

Bond Street aimed to improve loan services for small and medium-sized enterprises (SMEs) in the U.S. through technology, data, and a more user-friendly experience.

Traditional SME loans may involve a lot of paperwork, approval processes, and credit assessments.

Entrepreneurs believed that internet technology could reduce process costs and improve efficiency.

This direction has clear demand.

But Haber later reflected in an a16z article that one of the major challenges Bond Street faced was customer acquisition.

The problem was not just how to help customers obtain loans.

More importantly, it was how to accurately know when customers needed loans and whether they met the loan approval criteria.

1. Loans are products with obvious timing characteristics.

Many SMEs do not need loans every day.

Financing needs may only arise during expansion, equipment purchases, replenishing operating capital, or addressing short-term funding gaps.

If a fintech company cannot determine when customers have needs, it can only continuously advertise, hoping to reach them when they seek financing.

This increases customer acquisition costs.

For example, a company attracts 1,000 potential loan customers through advertising.

The cost per lead is $50, with a total expenditure of $50,000.

But only 100 of those customers actually meet the loan criteria, and ultimately only 20 customers complete the loan.

Then, the advertising expenditure per actual customer transaction is $2,500.

This is just a hypothetical example to illustrate the mechanism and does not reflect Bond Street's actual operating data.

If the net income generated from loans cannot cover customer acquisition costs, credit losses, funding costs, and operating expenses, the business will struggle to form a healthy unit economic model.

2. Why is understanding customer transaction intentions so important?

If a platform can observe customers' real transaction activities, it may be easier to identify financing needs.

For example, if a company starts purchasing equipment in large quantities or if accounts receivable significantly increase.

These changes may indicate that the company needs new funding support.

If fintech products can provide services in scenarios where customers have real needs, it may be more efficient than simply purchasing advertisements.

But this model involves data usage authorization, privacy protection, fair lending, and other regulatory requirements.

It cannot be simply understood as acquiring more customer data being better.

3. Why is understanding approval likelihood equally important?

Loan businesses must not only identify customers with needs but also assess which customers have reasonable repayment capabilities.

If customer demand is strong but credit risk is too high, financial institutions may still be unable to provide loans.

This means there is a gap between customer demand and the serviceable market.

For example, if ten thousand companies want financing, it does not mean that all ten thousand companies meet the loan standards of a particular institution.

Excellent fintech companies must simultaneously understand:

Who has demand?

Who can receive services?

What are the service costs?

How much revenue can customers bring?

Who bears the risk?

Bond Street was later acquired by Goldman Sachs and integrated into its digital financial business development system.

For Haber, this experience illustrates one thing:

Technology can improve financial service processes, but it cannot eliminate credit risk, customer acquisition costs, and capital constraints in financial businesses.

This is why fintech entrepreneurs need to understand software but also truly understand the economic structure of financial products.

Ten, why do top founders spend 95% of their time on sales?

Haber mentioned in an interview that entrepreneurs spend a significant amount of time actually engaged in sales.

Here, sales is not just about selling products to customers.

Entrepreneurs are conveying their vision to different groups every day.

They need to persuade talented individuals to join.

Persuade investors to provide capital.

Persuade customers to try the product.

Persuade partners to invest resources.

Sometimes they also need to persuade the media, regulators, and other stakeholders to understand the value of the business.

Therefore, entrepreneurs are actually facing four important groups over the long term:

Talent, customers, capital, and partnership networks.

1. Why must founders be good at attracting talent?

Startups typically lack the stability, brand, and compensation resources that large companies have.

Talented individuals joining a startup means taking on career risks.

They need to believe that the company has prospects and also trust the founder.

Thus, entrepreneurs must be able to explain the company's mission, business opportunities, competitive advantages, and future development paths.

But vision cannot replace real operating conditions.

Talented individuals will also evaluate compensation, equity, team culture, and the company's financial status.

2. Why must founders be good at attracting capital?

Venture capital firms face a large number of startup projects.

Entrepreneurs need to explain market demand, product advantages, growth data, and the use of funds within a limited time.

But fundraising is not just about telling a grand story.

Excellent entrepreneurs must be clear about:

How much money is needed?

Why is this funding needed?

What milestones will the funds create?

If the market environment deteriorates, how long can the company sustain itself?

Investors ultimately buy the potential for future value growth of the business, not the entrepreneur's presentation skills.

3. Why must founders continuously sell to customers?

This is the most important and also the easiest aspect to overlook.

Tech entrepreneurs often believe that as long as the product is advanced enough, customers will buy it proactively.

Reality is often not the case.

Corporate procurement requires budgeting, internal approval, security audits, and organizational coordination.

Even if the product can improve efficiency, it may not be adopted due to high migration costs.

Therefore, entrepreneurs must understand how purchasing decisions occur.

Who has the budget?

Who is the final decision-maker?

Who actually uses the product?

Who bears the risk of deployment failure?

Why should customers buy now rather than a year later?

These questions are often closer to the key to commercial success than product demonstrations.

The essence of sales ability is to help different stakeholders understand and believe in a future that has not yet been fully realized, while continuously proving this trust with actual results.

Eleven, mission-driven founders vs. mercenary founders: Why do top VCs value motivation?

Haber distinguishes entrepreneurs into two typical types:

Missionary, mission-driven entrepreneurs.

Mercenary, mercenary entrepreneurs.

This is an investment judgment framework rather than a strict scientific classification.

Mission-driven entrepreneurs usually have a long-term interest in a particular problem and are willing to continuously research and solve it.

Mercenary entrepreneurs may be more focused on short-term arbitrage, financing valuations, and quick exits.

Venture capital typically has a long cycle.

From early investment to final exit, it may take years.

Startups may also experience technological failures, market changes, financing difficulties, and team conflicts.

Therefore, investors have reason to pay attention to the founder's ability to remain committed during long-term difficulties.

But it cannot be simply assumed that those with a strong sense of mission will definitely succeed.

Some entrepreneurs may be very passionate about their products but neglect market demand.

Others may start with pragmatic business goals and ultimately build excellent companies.

Thus, the truly reasonable judgment criterion is not whether the entrepreneur is good at articulating a mission.

But rather:

Do they truly understand customer problems?

Are they willing to face unfavorable facts?

Can they continue to execute through long-term difficulties?

Do they have the ability to change erroneous judgments?

Can they establish a reasonable balance between commercial returns and long-term value?

A sense of mission can help a business navigate difficulties, but operational discipline determines whether the mission can exist long-term.

Twelve, why do interpersonal relationships have a compounding effect? Understanding Karmic Boomerang

Haber discussed the "Karmic Boomerang" in the interview, which can be understood as a kind of long-term reciprocal relationship's cyclical effect.

It emphasizes creating real value for others first, rather than immediately asking for a return every time a connection is made.

This concept is significant in the business world.

Because long-term cooperation highly relies on trust.

And trust cannot be established merely through a self-introduction or a social event.

1. Real value is more important than contact information.

Suppose an entrepreneur wants to meet a particular investor.

The usual approach is to send a message:

"Hello, I would like to connect with you for potential future collaboration."

But the other party may receive a large number of similar messages every day.

If the entrepreneur can provide industry research, customer information, or professional talent referrals that the other party truly needs, it may create a different foundation for the relationship.

The distinction here is:

The former primarily seeks attention from the other party.

The latter first provides a valuable contribution.

However, providing help must be based on legality, authenticity, and respect for the other party's wishes, and cannot be coercive sales under the guise of altruism.

2. Why is helping others solve their fourth or fifth most important problems more valuable?

Haber made a very practical observation.

Many excellent entrepreneurs' most important issues often occupy a large amount of their energy.

For example, financing, product development, and core customer delivery.

But they may still have many important matters that cannot be prioritized.

For example, finding a specific professional talent, entering new markets, contacting potential partners, or researching an unfamiliar field.

If you happen to be able to help solve these problems, you may become someone they trust for the long term.

This help does not necessarily require a large amount of funding.

Sometimes, an accurate talent recommendation, a high-quality business introduction, or a research report that truly solves a problem can create value.

3. Why might the value of interpersonal relationships compound over time?

Suppose an entrepreneur establishes genuine cooperative relationships with 20 outstanding professionals over the long term.

These individuals may come from finance, technology, law, media, or business management fields.

Over time, each person's experience, resources, and social networks may expand.

If both parties maintain trust over the long term, new cooperation opportunities may continuously arise.

However, the value of interpersonal networks does not simply grow with the number of contacts.

What truly matters is:

the quality of relationships, professional complementarity, level of trust, and ability to cooperate.

Having the contact information of ten thousand strangers is not necessarily more valuable than having twenty partners who truly trust you.

From a financial perspective, interpersonal relationships cannot be simply viewed as freely tradable assets.

But from a business strategy perspective, a high-quality relationship network can reduce information search costs, cooperation friction, and transaction costs.

This capability is particularly important in venture capital, corporate mergers and acquisitions, entrepreneurial cooperation, and talent recruitment.

Thirteen, why do truly excellent business opportunities often arise between different fields?

Haber's career experience illustrates that a person does not have to be defined by a single industry forever.

He has studied biochemistry, been involved in pharmaceutical revenue investment, engaged in venture capital, created a fintech company, and worked at Goldman Sachs in strategic roles.

These experiences enable him to understand business opportunities from different fields.

For example, the development of fintech largely comes from the combination of traditional financial services and internet technology.

Stripe transforms payment infrastructure into software interfaces that developers can use.

Plaid helps financial applications connect users' financial accounts under authorization conditions.

The value of these companies does not only come from inventing new financial concepts.

Rather, it comes from using software technology to improve the existing connection methods and user experience of financial services.

Today, similar opportunities may also arise in fields such as AI and healthcare, law, insurance, industrial manufacturing, and education.

However, the difficulty of cross-industry entrepreneurship is often underestimated.

For example, an excellent AI engineer may not understand insurance claims rules.

An experienced insurance practitioner may not necessarily have the ability to train and deploy AI systems.

Truly competitive teams often need to combine expertise from both fields.

Cross-industry opportunities do not come from knowing many terms at the same time, but from discovering real problems that have long existed between two industries but have not yet been effectively solved.

Fourteen, what exactly is the $1.7 billion Apps fund looking for in the AI era?

In 2026, a16z announced the raising of $1.7 billion for the Apps direction.

This scale of funding reflects institutional ongoing attention to AI application layers and software development.

But the scale of funding itself does not guarantee future investment success.

From the investment themes publicly discussed by a16z, the important opportunities in the AI application layer mainly come from three directions.

First, the transformation of traditional software categories into AI-native products.

Second, software begins to undertake tasks that previously required human labor.

Third, companies establish irreplaceable AI applications around proprietary data and closed-loop workflows.

The economic logic behind these three directions is not the same.

1. Software tools are evolving into work execution systems

In the past, enterprise software mainly helped employees manage work.

For example, customer relationship management software is responsible for storing customer information.

Employees still need to personally contact customers, schedule meetings, and follow up on sales.

In the future, AI systems may not only record customer information but also help complete customer classification, draft emails, query information, and remind follow-up tasks within authorized limits.

Software is gradually evolving from recording work to participating in executing work.

This may expand the value space of the software market.

But it also means that software vendors need to take on more quality, reliability, and accountability issues.

2. AI applications may shift from software budgets to labor budgets

Traditional SaaS companies mainly compete for enterprise software spending.

AI applications have the opportunity to change some service and labor costs.

For example, a company originally pays $1 million annually to handle a repetitive business.

If an AI system can complete part of that work and is rigorously validated, the company may be willing to pay significantly more than traditional software tools.

Because the reference point for purchasing decisions is no longer just the subscription fees of other software.

It also includes the overall cost of existing business processes.

But this does not mean that AI vendors can directly convert all the labor costs saved by customers into their own revenue.

Customers may demand to share efficiency gains, and competitors may also lower prices.

3. The real barriers to AI applications may come from business processes

Suppose two startups both use the same large language model.

The first company simply provides a generic chat interface.

The second company integrates AI with the insurance company's claims database, approval processes, internal permissions, and review mechanisms.

When the second company is already deeply integrated into the client's business, switching suppliers may require the client to re-integrate systems, migrate data, train employees, and bear operational risks.

This may create switching costs.

Therefore, AI application entrepreneurs should not only focus on model capabilities.

They should also pay attention to whether they have truly mastered the client's important workflows.

The most valuable AI applications in the long term may not be the products that showcase technology the most, but the systems that are the hardest to replace in the client's business.

Fifteen, can the a16z model be replicated by ordinary entrepreneurs?

It cannot be simply replicated in scale, but its organizational principles can be learned.

a16z has vast capital, a professional team, and a global business network.

Ordinary entrepreneurs do not need to establish platform departments with hundreds of people.

What is truly worth emulating is:

From day one of entrepreneurship, do not let all capabilities exist only in the founder's mind.

First, establish a sustainable and updatable customer knowledge base

Record customer problems, changes in needs, historical collaborations, and solutions.

Make customer relationships not just private chat records, but gradually become business knowledge that the enterprise can legally manage.

Second, establish standardized workflows

For example, customer inquiries, quotations, deliveries, quality checks, and follow-up services.

These processes do not need to be fully automated from the start.

But the steps, responsibilities, and evaluation criteria should be clearly defined.

Third, establish a partner network

Enterprises do not need to hire all professionals.

They can establish long-term collaborations with designers, lawyers, accountants, technical consultants, and industry experts.

This network can help small businesses call upon external capabilities when needed.

Fourth, use AI to consolidate organizational memory

Organize authorized meeting minutes, project experiences, customer feedback, and internal processes into a searchable knowledge system.

But data permissions and confidentiality boundaries must be clearly defined.

Fifth, continuously measure economic outcomes

Do not assume that productivity will necessarily improve just because AI is used.

It is necessary to accurately record time costs, tool expenses, rework rates, customer satisfaction, and final business results.

Only systems that can continuously improve economic outcomes may become true competitive advantages.

Sixteen, what can ordinary investors learn from top VC methodologies?

Although ordinary investors cannot directly replicate a16z's investment strategy, they can still learn its analytical methods.

Especially the distinction between a company's short-term performance and long-term competitive advantages.

Suppose there are two software companies.

The first company has rapid revenue growth but relies on expensive advertising to continuously acquire customers.

The second company grows at a slightly slower pace but has a higher customer retention rate, stable product usage habits, and lower customer acquisition costs.

Which company has more long-term value?

You cannot make a judgment solely based on revenue growth rates.

You need to compare customer lifetime value, customer acquisition costs, gross margins, cash flow, and competitive barriers.

For example, suppose a company spends $1,000 to acquire a customer.

The customer generates $600 in gross profit annually and stays for an average of five years.

Ignoring the time value of money, retention distribution, and other costs in this simplified scenario, the cumulative gross profit from the customer would be $3,000.

This business model may have further research value.

But if the customer only stays for an average of one year, the company may not be able to recover enough customer acquisition costs.

This is why top investment institutions value unit economic models.

Furthermore, investors should also analyze whether a company's competitive advantages can be sustained.

A company that has leading technology today does not mean it will still be leading five years from now.

A company that has a large customer base today does not mean customers will not leave for lower prices or better products.

Excellent investment research must analyze:

Why a company is successful and what circumstances might cause it to no longer be successful.

Seventeen, the four practices most worth executing for entrepreneurs

The first: test the founder's advantages on one page

Try to clearly answer five questions:

What specific problem are we solving?

Why would customers be willing to pay?

Why has no one succeeded in solving this in the past?

Why are the conditions for commercialization present now?

Why is our team more advantageous than competitors?

If you cannot clearly answer, it indicates that the business model still needs further research.

The second: study the failures of the past decade

Choose the industry you are preparing to enter and look for relevant companies that failed in the past five to ten years.

Focus on analyzing:

Were customer acquisition costs too high?

Was market demand overestimated?

Did the product fail to generate repeat purchases?

Did regulations or supply chains create obstacles?

Did the company consume too much capital before the business model matured?

Do not just learn how successful people succeeded.

Also study why failures occurred.

This often helps entrepreneurs avoid costly mistakes.

The third: establish your own business knowledge system

Start with the most easily lost information from daily operations.

For example, meeting minutes, customer feedback, partner information, and project reviews.

Use legal and secure tools to organize this content and establish a searchable knowledge base.

The focus is not on saving all information but on continuously refining experiences that can be used for the next decision.

The fourth: actively create value for partners every month

Choose a few people worth long-term cooperation and understand the real problems they need to solve.

If you can provide useful professional information, talent recommendations, or business cooperation opportunities, offer help within appropriate boundaries.

But do not simplify interpersonal relationships into a transaction that guarantees future returns.

Long-term trust comes from real abilities, reliable behavior, and mutual respect.

Conclusion: The strongest competitors in the future may not necessarily have the most capital, but rather those who best understand how to build compounding systems.

David Haber's experience reveals a broader business principle than venture capital.

Organizations that can truly succeed in the long term cannot rely solely on one correct decision.

They need to continuously create new competitive advantages.

Royalty Pharma establishes its financial business through professional asset investment and capital allocation.

Goldman Sachs has become a long-standing financial institution through complex customer relationships, financial products, and organizational capabilities.

a16z attempts to evolve venture capital from relying on the personal skills of a few partners to having independent investment teams, professional platforms, and internal technology systems.

Today, AI further provides new possibilities.

A large amount of knowledge that was previously scattered across personal experiences, meeting records, and corporate documents is becoming easier to organize, search, and utilize.

This may reduce organizations' reliance on individual employees' memories and improve information sharing and decision-making efficiency.

But AI itself cannot guarantee business success.

If a business lacks real customer demand, AI will not automatically create demand.

If the business model cannot generate profits, AI cannot improve the economic structure out of thin air.

If an organization does not have the correct incentive and accountability mechanisms, automation may even accelerate the spread of poor decision-making.

Therefore, what truly matters in the future is not just having the most advanced technology.

It is whether one can organize technology, capital, talent, knowledge, and business relationships into a continuously improving system.

For entrepreneurs, this means not just pursuing the next round of financing, but building a business that can genuinely serve customers.

For investors, it means not just focusing on popular trends, but studying whether a company has the ability to deliver long-term capital returns.

For financial institutions, it means not just relying on a few investment stars, but establishing sustainable organizational advantages.

For ordinary people, it means not just accumulating non-transferable work experience, but gradually forming a system of professional knowledge, trustworthy relationships, and personal capabilities that can be continuously utilized.

A successful fund can provide investors with huge returns in a certain cycle; a truly excellent institution must learn to accumulate competitive advantages over time.

This is precisely the most thought-provoking significance of David Haber's "Firm over Fund."

It applies not only to the venture capital industry but also to startups, financial institutions, technology companies, and anyone who hopes to build a long-term career.

One of the biggest business opportunities in the future may not rely on quickly profiting from a temporary technological advantage, but rather on transforming continuously evolving technological capabilities into an organizational system that can create long-term value.

Sources and Further Reading

1. The Luba Show, October 5, 2026, How a16z's Most Connected VC Picks Founders podparadise.com
2. David Haber, January 12, 2026, Firm > Fund a16z.news
3. Ben Horowitz, January 9, 2026, Why Are We Here? Why Did We Raise $15B? a16z.com
4. Andreessen Horowitz, David Haber's official profile a16z.com
5. David Haber, August 9, 2022, Leading With Software When Building a Lending Business a16z.com
6. Andreessen Horowitz, January 19, 2026, The AI Opportunity That Goes Beyond Models a16z.com
7. Royalty Pharma, January 2026 SEC proxy statement, Business and Portfolio Information sec.gov

ABAB NEWS｜Opinion

This article is based on public interviews, corporate materials, and industry research. The business projections and hypothetical figures in the text are for explanatory purposes regarding economic mechanisms and do not constitute investment advice.