Ten Years of Advice from a16z Partners to Web3 Founders: In the New Cycle, Just Focus on Three Things

In the Web3 world, cycles are not an accident, but the norm. The alternation between bull and bear markets is like the tides of capital and the four seasons of nature. For founders, the biggest challenge has never been predicting the next reversal, but how to survive amidst the ups and downs, and even build long-term value against the current.

Recently, a16z crypto partner Arianna Simpson shared her over a decade of experience investing in the crypto industry on a podcast. From the shock of the Bitcoin white paper to the product-market fit of stablecoins, to the overlap between Crypto and AI, and advice for founders.

These observations and experiences are not only applicable to Silicon Valley. In the view of Portal Labs, they also provide valuable ideas for Web3 founders and high-net-worth investors in China to reference and learn from.

The Nature of Cycles

Arianna entered the world of cryptocurrency over ten years ago, when she was shocked for the first time by reading the Bitcoin whitepaper. However, what truly made her stay was not the moment of excitement, but the ups and downs she experienced over the following decade. She witnessed the birth of Bitcoin, the prosperity of DeFi, the frenzy of NFTs, and also experienced the ensuing bubbles and cooling. It was through this long-term observation that she gradually formed a clear understanding: the crypto industry has never grown linearly, but rather advances in dramatic waves, with emotions and capital ebbing and flowing alternately.

Therefore, she shifted her focus from "predicting the next wave of opportunities" to "identifying who is building against the wind." Her investment approach resembles a kind of following: following what the best founders are doing. When the strongest founders flock to stablecoins, funds should converge there; when cutting-edge teams continue to invest in Crypto × AI or DePIN, new value vacuums often take shape. It's not about having grand conclusions first and then finding projects to validate them; rather, it's about calibrating one's worldview and capital allocation based on the direction of frontline builders.

For Chinese Web3 founders and high-net-worth investors, this methodology is more actionable than "cycle prediction." For founders, the cooling-off period is not an excuse, but a form of selection: being able to push products and stacks forward in years without applause indicates that both the direction and the people are right; for allocators, what really needs to be evaluated is not the popularity of the theme, but whether the team can maintain speed, discipline, and mission density during difficult years. This order of "identifying people—looking at long-term execution—then discussing valuation" is more capable of transcending bull and bear markets than any short-term narrative.

stablecoin

Narrowing the lens to stablecoins. Arianna's judgment is quite simple: the reason it has become the focus now is not because of new speculative stories, but because both ends are really using it - consumers use it for cross-border transfers and to hedge against local currency fluctuations; businesses use it for settlement, allocation, and as a bridge for accounts receivable and payable. More importantly, in the past year and a half, the two infrastructure "valves" of speed and cost have finally been opened, allowing stablecoins to transform from an imagined payment network into a real settlement layer.

This point directly hits the Web3 founders and high-net-worth individuals in China. For overseas teams, what often really constrains them is not the product, but the cash flow: how to send money to Southeast Asia's labeling teams, Africa's node maintainers, and Latin America's channel partners in a stable, low-cost, and traceable manner; how to allow overseas clients to make payments without complex corporate processes; how to manage periodic receivables in a dollar environment and control exchange rate risks in a local currency environment. The value of stablecoins lies not in the "coin," but in the "track." When you standardize the entry and exit of funds, identity verification, reconciliation receipts, and tax traces onto an auditable track, the complexity of cross-border business will significantly decrease.

Of course, there will be more and more issuers, but users will not pay for every new token. Arianna's intuition is that in the short term, there will be a flourishing of many options, but in the long term, it will certainly converge to a few stablecoins that have "scale, credibility, and ecological niche"; further on, the front-end experience will become abstracted, and users will hardly perceive specific tokens, while the back-end will automatically complete clearing and settlement through "rail transit interconnection."

This means that in the future construction of stablecoins, the team should not waste energy on the impulse of "I want to issue one too," but should focus on more pragmatic designs, such as how to thoroughly "native" your business processes, risk control, and financial systems with stablecoins. When your product can naturally operate on a path priced in dollars, settled with stablecoins, and reconciled on-chain, your cross-border efficiency and credibility will stand out directly among peers.

For high net worth individuals, stablecoins are a new cash management tool and a "low-friction channel" for global liquidity. However, this does not mean there is no risk. At the portfolio level, reserving an on-chain track for "liquidity turnover" and "hedging against local currency fluctuations" is a more future-oriented portfolio hygiene. In simple terms, two principles: carefully choose counterparties, diversify custody and wallets; make "compliance and interpretability" the first constraint, not a last-minute fix.

Crypto × AI × DePIN

Arianna emphasized that supercycles are often not driven by a single technology, but rather by several curves resonating together within the same time window. The clearest combination today is the decentralized incentives of cryptocurrency, combined with the centralized computing power and data hunger of AI, layered with the orchestration of real-world resources through DePIN.

Translate it into the "deployable" language of the Chinese founders: We have rare long-term accumulation in hardware supply chain, manufacturing and deployment, and engineering organization of edge nodes. If you can use stablecoins to connect the "contribution-metric-payment" chain to incentivize real-world data and resources on-chain, then package these resources into standardized products consumable by AI (datasets, annotations, bandwidth, storage, inference time slices), you will have the opportunity to create a "supply-side platform". This is not PPT-style token economics; it is serious operations science: defining metrics, anti-cheating, settlement frequency, dispute resolution, reputation systems—all need to be engineered.

Another important context is "authenticity." The existence of deepfake content is not terrifying; what is terrifying is an unverifiable environment. Verifiable timestamps, generation paths, device signatures, and traceability of the operating entities are the "new water, electricity, and coal" for the future of content and commodity internet. This presents a near-term opportunity for Chinese teams engaged in brand internationalization, second-hand transactions, and luxury goods circulation. Doing difficult yet correct things: making "authenticity verifiable" the default rather than an optional paid feature.

Let's take a look at AI Agents. It is irresponsible to hand over a credit card to a semi-mature agent for "self-service online shopping"; however, providing it with a wallet that has limits, can be revoked, and is auditable, allows it to complete a series of tasks (subscription, purchasing APIs, paying commissions) within a clear strategy, which is practical. In other words, "the wallet is the permission system." The real application is not the inflated "universal agent," but the vertically deep "bounded rational agent"—binding permissions, budgets, logs, and counterparties together using an on-chain wallet within a strongly constrained business domain.

Financing and Governance

The financing environment from 2020 to 2021 may have left many in Web3 with the illusion that there is no need for a deck, no need for a model, and investors would offer outrageous terms in direct messages on Twitter.

Arianna said bluntly: that is a "twilight mirage," not the norm, and today we should return to the basics. Prepare usable materials, honestly present indicators, and set financing targets at a conservative but achievable level; it is better to close a round of reasonable funding first and then snowball, rather than to aim for 50 million at the outset and end up with nothing.

For Chinese founders, a more realistic order is: first run the base, then talk about money. First, the engineering resilience of technology and products, performance, risk control, observability, and operability; second, compliance and policy paths, KYC/AML, data cross-border segmentation, auditability of fund and data flows, and tax and invoice closure; third, verifiable business closure, real payments, positive unit economics, and stable payment rhythms. In publicly available narratives, talk less about "coins" and focus more on supply-side infrastructure: for example, standardizing computing power / bandwidth / sensor data into billable APIs using DePIN, or digitizing existing assets with RWA and embedding them into compliance issuance and settlement processes. Once there is a chain of evidence for these three things, use milestones to segment capital supplementation, rather than letting financing dictate business direction.

Governance must also return to common sense. A 50-50 split is not fair; it is inaction. Equity, board of directors, reserved matters, vesting periods, cliff periods, founder departure clauses, and intellectual property ownership—none of these are sexy, but each one determines whether you can weather the first major storm. Arianna even does not shy away from the merits of "solo founding"—at least it won't lead to a falling out with oneself. Portal Labs suggests that rather than getting tangled in the "number of partners," it is better to properly outline the "list of rights and responsibilities" and the "conflict resolution mechanism"; by clearly rehearsing the worst-case scenario, you can run faster when the best time comes.

Competition and Expansion

Being plagiarized is not news; being obsessed with competing is. Arianna's approach is to reclaim the narrative: define the topic with product cadence, key metrics, and customer stories, instead of directing traffic to competitors. For Chinese Web3 teams, it is especially important to address the "infrastructure" of PR and communication: professional branding teams, media whitelists, KOL advocates, product education in user communities, and transparency in technical documentation. The narrative is not a PR speech, but the evidence of your continuous delivery.

At the same time, uncontrolled growth is both a good thing and a crisis. When the service water level is breached, it should be handled in a graded manner like firefighting: first protect the safety of funds and user assets, then ensure availability, and finally optimize the experience. If necessary, throttling, opening temporary whitelists, outsourcing customer service and risk control, and even quickly bridging to supplement computing power are all acceptable trade-offs. Write the "disaster recovery plan" when things are calm, not when learning from trending topics.

Mergers and acquisitions represent another signal. Traditional giants are beginning to act as buyers in the crypto space, and "puzzle-like acquisitions" within the industry are also emerging. Ideally, you would be the acquirer, but an excellent target for acquisition could also be the optimal solution for the team, users, and early shareholders. The evaluation criteria are straightforward: strategic fit, user value, team continuity, and respect for the technological roadmap. Leave emotions to social circles and terms to lawyers.

Do difficult but correct things for a little longer.

The market will not provide standard answers for founders, and the cycle certainly won’t. Therefore, don’t rush to predict the waves; focus on those who can still push the system forward in headwinds, and then allocate time and resources to them. In the context of China, the answers are simpler yet more challenging: slogans should not just be shouted, but the accounts, systems, and compliance documents must be made concrete; growth is not trending topics, but stable and reusable supply and returns; competition is not about confrontation, but about holding the narrative power and reclaiming topics through continuous delivery.

If we were to leave a message for China's Web3, Portal Labs believes it should be: first, do the difficult but right things for a longer time, chase fewer trends, and see who is still around in ten years, and whose system is still running. The cycles will continue to rise and fall, but what truly determines victory or defeat has never been the weather; it is the foundation on which you build your house.