The Unscalable Path to Scale
Do things that don't scale to build the foundation for things that do
Bill Gurley articulates a paradox that every marketplace and platform founder must embrace: you must do tons of unscalable things in order to build something that scales. The scalable version of your business — the automated, self-reinforcing platform — cannot exist until you manually create the initial conditions that make it possible. For a marketplace, this means personally recruiting the first suppliers, handholding the first buyers, and creating the initial content or inventory that makes the platform worth visiting. For a user-generated content platform, this means seeding content yourself, hand-curating contributions, and personally engaging with early contributors. This manual, labor-intensive work is the opposite of what the eventual scaled business looks like, which is why many founders resist it. They want to build the scalable system from day one. But without the unscalable groundwork, the scalable system has nothing to scale. The insight resolves the chicken-and-egg problem that plagues every platform business: buyers will not come without sellers, and sellers will not come without buyers. The answer is that you personally are both the first buyer and the first seller until organic participation takes over.
- You must do tons of unscalable things to build something that scales.
- The scalable system cannot exist until you manually create the initial conditions that make it possible.
- Every successful marketplace went through an intensely manual phase before achieving self-reinforcing growth.
- Resistance to unscalable work is the number one reason marketplace startups fail at launch.
- Identify Your Critical Mass RequirementDetermine what minimum level of content, inventory, users, or liquidity your platform needs to deliver enough value that organic growth can begin. For a marketplace, this might be a specific number of sellers in your launch geography. For a UGC platform, it might be a specific volume of high-quality content. For a social network, it might be a specific density of connections within a target community. This critical mass number is your target for the unscalable phase — you need to reach it through manual effort before scalable growth kicks in.Pro tipStart in one geography, one vertical, or one community. Achieving critical mass in a narrow space is far easier than trying to achieve it broadly. Uber started in San Francisco. Airbnb started with air mattresses at conferences.WarningDo not underestimate the critical mass requirement. Most founders assume a few dozen suppliers or a few hundred pieces of content will be enough. Usually, you need significantly more than you think.
- Do the Manual Work YourselfPersonally recruit the first suppliers, create the first content, serve the first customers, and solve the first problems. Do not delegate this to junior employees or contractors — the founders themselves should be doing the unscalable work because it provides irreplaceable learning about the business. When Airbnb founders photographed apartments, they learned what made listings attractive. When Glassdoor founders interviewed employees, they learned what information job seekers valued most. This learning is as valuable as the content it produces because it shapes every subsequent product decision.Pro tipDocument every interaction during the unscalable phase. The patterns you observe in manual work become the design specifications for the eventual automated system.WarningThis phase is exhausting and feels counterproductive because you are doing work that will eventually be automated away. Embrace it — the manual phase is where you build the deepest understanding of your business.
- Transition to Scalable Systems Once Critical Mass Is ReachedOnce you have achieved critical mass through manual effort and organic growth begins to appear, systematically replace manual processes with scalable systems. Automate what you have been doing by hand. Build tools that enable users to do what you have been doing for them. Create incentive structures that motivate the behavior you have been manually seeding. The key is timing the transition correctly: too early and you lose the personal touch that makes the early product special; too late and you burn out or cannot keep up with demand.Pro tipThe transition from unscalable to scalable should be gradual, not binary. Automate one process at a time and monitor whether the quality holds. Some unscalable elements (like personal customer engagement) may be worth maintaining longer than expected.
In Airbnb's earliest days, founders Brian Chesky and Joe Gebbia went door-to-door in New York City photographing apartments listed on their platform. Professional photos dramatically increased booking rates but the process was completely unscalable — two founders with cameras visiting individual apartments. However, this unscalable effort solved the cold-start problem by making listings attractive enough to generate bookings, which attracted more hosts, which attracted more guests.
Gurley developed this principle through decades of investing in marketplace businesses at Benchmark Capital. He observed that every successful marketplace in Benchmark's portfolio — eBay, Uber, Airbnb, OpenTable — went through an intensely manual phase before achieving the scalable, self-reinforcing growth that made them valuable. The concept was influenced by Paul Graham's famous essay 'Do Things That Don't Scale' (2013) but Gurley brought specific marketplace experience and examples that demonstrated the principle with empirical precision. His examples include Glassdoor founders interviewing employees at Starbucks, Yelp founders wearing branded t-shirts in nightclubs to recruit reviewers, and Airbnb founders personally photographing apartments to create attractive listings — all deeply unscalable activities that were necessary to create the conditions for scaled growth.