Menlo Ventures’ Matt Murphy on why Anthropic is winning (and it's not the model) | Equity Podcast

Menlo Ventures Investment Strategy and AI Integration

• Menlo Ventures has established investment stakes in several prominent artificial intelligence companies, including Anthropic, Open Router, and Lovable. 10s
• In 2024, Menlo Ventures led a Series D investment round in Anthropic valued at more than $500 million. 1m0s
• Matt Murphy served as a board observer for Anthropic immediately following the Series D round, though he no longer holds that position following subsequent funding rounds. 1m15s
• Menlo Ventures utilizes Anthropic’s Claude as a foundational tool to improve efficiency, productivity, and collaboration within their investment team. 1m45s
• The firm integrates Claude with various platforms, including Affinity, Slack, and Harmonic, to automate tasks such as summarizing quarterly CEO updates and tracking signals regarding entrepreneurs and companies. 2m0s
• Beyond professional applications, Claude is used as a research tool to identify experts, gather information, and adjust the complexity of explanations. 2m35s
• While Matt Murphy uses Claude for research and productivity, he notes that other partners at Menlo Ventures, such as Tim Tello, DDOS, and Matt Cranney, are more advanced power users who build tools for the firm's portfolio companies. 3m15s

Founding and Early Development of Anthropic

• Dario Amodei is described as a special founder who possesses both a unique energy and an exceptional depth of knowledge regarding the development and construction of artificial intelligence. 3m45s
• Dario Amodei, a founder of ChatGPT at OpenAI, left his previous role because he identified a significant opportunity that was not being prioritized, leading him to start a new venture to pursue that vision. 0s
• The founding team of Anthropic, including Tom Brown and Daniela Amodei, was recognized as a special group with a proven track record. 0s
• Dario Amodei utilized his experience from OpenAI to build a competitive model at approximately one-tenth of the cost of previous efforts. 0s

Investment Thesis and Growth of Anthropic

• Menlo Ventures invested in Anthropic during its Series C round, despite the investment not fitting cleanly into the firm's existing early-stage or early-growth fund mandates. 0s
• At the time of the Series C investment, Anthropic was seeking to raise over $4 billion, which was an unconventional entry valuation for a pre-revenue and pre-launch company three years ago. 0s
• The investment decision was driven by the belief that the AI market would not be dominated by a single player and that Anthropic had the best potential to become the strong number two competitor to ChatGPT. 0s
• Menlo Ventures had pivoted its firm focus toward AI 6 to 12 months prior to the investment and had been committed to the sector for approximately four years. 0s
• While the firm initially viewed the possibility of Anthropic becoming the market leader as a 10-year goal, the company achieved significant growth within three years. 0s
• Anthropic has experienced rapid revenue growth, with reports indicating a leap from a $9 billion revenue run rate in 2025 to a $47 billion run rate as of May. 0s

Technical Architecture and Strategic Partnerships

• The technical architecture and compute multipliers developed by Tom Brown allowed Anthropic to build more efficient models with fewer resources. 0s
• Menlo Ventures led a Series D investment round of over $500 million in Anthropic when the company was still in the early stages of generating revenue 0s.
• Anthropic secured Google and Amazon as investors, technology partners, and distribution partners, which helped the company establish an organic revenue engine 0s.
• The investment decision was driven by Anthropic's clear focus on the enterprise market and the recognition that enterprises required the company's specific technology 0s.
• The investment thesis for Anthropic was built on three pillars: the quality of the team, their technical advantage, and the acquisition of massive strategic partners 35s.
• The significant amount of capital raised by Anthropic created a high barrier to entry, causing other early-stage foundation model competitors to fade away or fail to keep pace 53s.

Product Development and User Accessibility

• A major factor in Anthropic's success is the development of a "harness" that facilitates user interaction beyond a simple prompt box, allowing for easier integration and application 1m10s.
• Anthropic implemented tools such as Claude Code, Claude Code for life sciences, and Model Context Protocol (MCP) to enable users to connect their own data to the models 1m25s.
• The creation of Claude skills allows users to customize the technology once they have connected their data, which has significantly increased market accessibility and growth 1m35s.
• While a strong underlying model is necessary, the market growth Anthropic has experienced is largely attributed to the company's efforts to make its technology easy to use and accessible 1m50s.
• Founders can learn from Anthropic's approach by focusing on the "last mile" of product development, which involves understanding how to meet customers where they are and making the platform the easiest to use 2m15s.
• The importance of design-oriented development and ease of use is a recurring theme in successful technology companies, similar to the early growth of platforms like Airbnb 2m35s.

Market Dynamics and Multi-Model AI Adoption

• Anthropic currently maintains a strong position in the enterprise market, with data from Open Router and Perplexity indicating that the company commands a significant portion of token spending 0s.
• A trend is emerging where enterprises and growth-stage AI startups are increasingly adopting open-source models to manage and optimize their operational costs 25s.
• The adoption of AI technology typically follows a two-phase cycle: an initial phase of experimentation and learning, followed by a second phase focused on optimizing cost structures once an application becomes successful and generates significant revenue 1m15s.
• Tools like Open Router assist companies in optimizing their usage within the Claude model family by evaluating factors such as capacity, latency, and specific model capabilities like Sonnet or Opus 1m45s.
• The future of the AI landscape is expected to be multi-model, where companies utilize a combination of proprietary models like Claude, open-source models, and custom models trained on their own proprietary data 2m15s.
• Decisions regarding model selection will likely be driven by specific application requirements, such as the need for high-level reasoning, cost efficiency, or low latency 2m45s.
• Sophisticated companies are expected to split their API requests between proprietary and open-source or custom models, while the majority of customers without dedicated research resources will likely continue to rely almost exclusively on proprietary solutions like Claude 3m15s.
• The coexistence of proprietary and open-source models serves to drive competition within the industry 3m45s.
• A competitive market with multiple choices is beneficial for the industry because it strengthens incumbents and encourages continuous improvement 0s.

Trust, Safety, and Regulatory Challenges

• Anthropic benefited from its past interactions with the federal government during the Trump administration 7s.
• Some cybersecurity professionals have characterized the rollout of Anthropic’s cybersecurity-oriented model, Mythos, as being more focused on marketing than on providing actual protection, noting that vulnerability scanning tools have existed for a long time 14s.
• Despite criticisms regarding the Mythos rollout, available data indicates that Anthropic is performing well financially 35s.
• Anthropic’s commitment to trust and safety has been a core part of its investment thesis and has proven to be a differentiator in the enterprise market 53s.
• The leadership team at Anthropic, including Dario Amodei, is described as mission-driven, with a focus on the societal impacts of their technology 1m15s.
• Anthropic’s investment in life sciences is motivated by a desire to create beneficial impacts on the world, rather than solely by immediate revenue generation 1m21s.
• The cautious approach to releasing powerful models is intended to prevent potential catastrophes or security breaches that could damage the industry's reputation and slow down future innovation 1m40s.

Frameworks for AI Governance and Regulation

• The current regulatory landscape is complex, with various states attempting to implement their own AI regulations 2m6s.
• The CEO of DeepMind has proposed the creation of an independent governance board for the AI industry, modeled after the Financial Industry Regulatory Authority (FINRA) 2m25s.
• There is concern that federal intervention in AI regulation, as demonstrated by the Trump administration, could create challenges for young startups 2m35s.
• Current industry regulation processes, such as those involving NIST, are criticized for being ad hoc and unpredictable for model vendors 0s.
• Establishing a structured regulatory framework is necessary to prevent the release of potentially dangerous artificial intelligence into the wild, though there is a risk that such regulations could stifle innovation 0s.
• Utilizing industry experts rather than government officials for regulatory oversight is suggested as a more effective approach for understanding the impacts of AI technology 0s.
• A well-implemented regulatory framework could potentially streamline processes compared to the current case-by-case approach 0s.
• The current lack of a unified federal regulatory standard, combined with fragmented state-level regulations, creates a complicated environment that threatens to slow down innovation for startups 35s.
• While the idea of a group of AI industry experts collaborating on regulation is viewed as positive, it is noted that political challenges may make such cooperation difficult 35s.

Lovable Platform Growth and Market Impact

• Lovable, a platform designed to enable non-technical users to design and build, has experienced rapid growth, moving from zero to 300 in a year, with rumors suggesting even higher figures 1m15s.
• The growth trajectory of companies like Lovable is described as rare, though it has become more common in the industry over the last few years 1m15s.
• Lovable is categorized as a tool that empowers the "other 99%" of the population to innovate and create, similar to how previous investments like Uber or Rover unlocked value in shadow markets by providing platforms for previously untapped activity 1m15s.
• Lovable has expanded its platform to include hosting capabilities and payment processing, enabling users to build, run, and manage businesses rather than just creating individual projects 0s.
• Several businesses built on the Lovable platform have achieved revenue run rates exceeding one million dollars 0s.
• Lovable targets a broad market of "business creators" by utilizing a simple design aesthetic, pre-built skills, templates, and a persona-based interface that does not require technical expertise 0s.
• While there is minor overlap between Lovable and Cloud Code, the market is sufficiently large to accommodate both, with Lovable focusing on an approachable, non-technical user experience 0s.

Ubiquitous Adoption and Product-Led Growth

• The current AI wave is characterized by rapid adoption because the technology is accessible to anyone on any device through a familiar chat interface, unlike previous technological shifts like mobile or the internet that required hardware adoption or significant behavioral changes 1m25s.
• A key difference between current AI businesses and previous SaaS models is that AI is an "everybody phenomenon" rather than primarily an enterprise-focused one 2m25s.
• Modern AI companies are increasingly product-led rather than sales-led, allowing them to bypass the scaling constraints associated with building and ramping a traditional sales force 2m25s.
• Artificial intelligence technology has achieved ubiquitous adoption due to its low friction and high utility in creating new business opportunities 0s.

Market Trends and Future Outlook for AI

• LegalZoom has experienced significant growth, reaching $50 million in quarterly revenue and achieving $150 million in revenue rapidly by selling to the legal sector, a market previously considered unattractive by venture capitalists 12s.
• Large enterprises are demonstrating high demand for AI, evidenced by Salesforce spending $300 million on tokens to increase engineering productivity and Uber exhausting its budget for AI tools 35s.
• While there is current discourse regarding the return on investment for AI, the high level of user engagement with these capabilities is driving rapid growth across various categories 55s.
• Entrepreneurs face increased pressure as the benchmark for success has shifted; growth rates that previously placed a company in the top 1% now feel less significant compared to the exponential growth curves currently being observed 1m15s.
• The current landscape allows for a faster transition from an initial idea to a functional product due to a reduction in barriers to entry 1m45s.
• AI enables increased productivity by automating busy work, allowing teams to focus on high-impact activities and operate with smaller staff sizes 2m15s.
• Despite the utility of AI, current hype surrounding the technology still outpaces its practical reality, as the tools are not yet fully reliable 2m35s.
• The rapid pace of technological advancement makes it difficult to predict the state of the industry by 2027, given the unexpected developments that occurred between 2023 and 2026 2m55s.
• There is a significant market opportunity for a consumer-facing AI product that serves individual needs rather than being restricted to tools provided by employers 3m35s.
• The individual interviewed expresses a lack of interest in pursuing consumer-facing business models, specifically citing the challenges of competing on price or managing insurance claim disputes 0s.

Contact Information and Podcast Credits

• Matt Murphy provides his professional email address, matt@menlovc.com, for those wishing to reach him 25s.
• Matt Murphy invites interested parties to follow him on Twitter using the handle @MMurph 25s.
• The podcast Equity is hosted by senior reporters from TechCrunch 35s.
• Teresa Lo Consolo serves as the producer for the Equity podcast 35s.
• The editing for the podcast is performed by Kell 35s.
• Listeners are encouraged to subscribe to the podcast on YouTube or other podcast platforms and to visit techcrunch.com/events for information regarding future updates 35s.