Why was scalability so important in early internet companies?

Early web servers and bandwidth were slow, making technical scaling a major challenge. Investors also expected business models to have the potential for massive growth, equating scalability with venture capital fundability.

How did Airbnb's founders do 'things that don't scale'?

When struggling for users, they took professional-quality photos of the listings themselves to make them more attractive, directly addressing a key customer need that wasn't scalable.

What unscalable task did Fleek perform to learn?

Fleek's founders manually sourced and sold secondhand clothes from wholesalers to shops in London for four months to learn about what sells, pricing, and demand elasticity.

How did Instacart launch without grocery store partnerships?

They bought one of every item at Trader Joe's, photographed them, listed prices online, and launched their service over a weekend, acting as a manual proxy until they could secure formal deals.

What was DoorDash's initial MVP strategy?

They built their initial product in an afternoon using Google Drive for menus, basic HTML/CSS, a founder's phone for dispatch simulation, and 'Find My Friends' app to test customer demand without complex engineering.

When should a startup stop doing unscalable things?

You should transition to scaling when you've answered key questions about product-market fit and customer demand. Advisors and investors can help signal when it's time to invest in scalable infrastructure.

Key Moments

Startup Experts Discuss Doing Things That Don't Scale

Y Combinator

Science & Technology3 min read26 min video

May 30, 2024|214,186 views|4,651|113

YC Y Combinator

Save to Pod

Key Moments

TL;DR

Early startups should prioritize learning by doing manual, unscalable tasks to delight customers and find product-market fit.

Key Insights

Founding belief: Early startups should do things that don't scale to learn and delight early customers.

Scalability obsession, influenced by Google, was a problem; PG's essay offered a solution.

Unscalable actions like manual customer service, product placement, and direct founder involvement are crucial for early traction.

Examples like Airbnb, Algolia, Instacart, and DoorDash showcase successful manual efforts.

The goal is to optimize for learning and customer understanding, not immediate scalability.

Knowing when to transition from unscalable tasks to actual scaling is key for long-term success.

THE CONTEXT: THE PROBLEM WITH SCALABILITY OBSESSION

In the early 2000s, the tech industry, heavily influenced by companies like Google, was obsessed with scalability. This meant not only technical scalability of servers but also scalable business models capable of generating billions. This created immense pressure on founders, making it difficult to secure funding without a clear, scalable path. This environment fostered a focus on theoretical future problems rather than immediate customer needs, setting the stage for a counter-intuitive but essential piece of advice.

PAUL GRAHAM'S ESSAY: INVERTING THE CONVENTIONAL WISDOM

Paul Graham's 2013 essay, "Do Things That Don't Scale," directly addressed the prevailing scalability obsession. He argued that a startup's primary challenge is not technical infrastructure but acquiring and delighting early customers – getting from zero to one. The essay encouraged founders to ignore theoretical scaling issues and focus on fixing immediate problems, even if it meant manual, one-off efforts. This contrarian approach transformed Silicon Valley's culture, shifting focus to practicality and customer acquisition.

THE ESSENCE: OPTIMIZING FOR LEARNING AND CUSTOMER DELIGHT

The core principle behind "doing things that don't scale" is deep learning and customer understanding. This often involves manual, albeit time-consuming, tasks that provide invaluable insights. Examples include Airbnb founders manually taking professional photos of listings to improve quality, or Algolia's founders directly implementing their search solution for early customers to understand usage deeply. The goal is to intimately grasp customer pain points and needs, which is difficult to achieve through generic processes.

PRACTICAL EXAMPLES: MANUAL EFFORTS THAT DROVE TRACTION

Numerous successful startups exemplify this philosophy. Instacart, lacking grocery store partnerships, manually photographed and listed inventory to launch. DoorDash built its first version in a day using manual tools like Google Forms and 'Find My Friends' to simulate real-time dispatch. Fleek manually transported and sold clothes to understand the secondhand market dynamics. These founders, often coming from well-established tech backgrounds, embraced tasks that felt beneath them, recognizing their critical role in achieving product-market fit.

FOUNDER FACE TIME AND THE ADVANTAGE OF BEING SMALL

Founder FaceTime, or direct founder involvement, is a powerful, unscalable advantage. Early-stage founders can offer a level of personal commitment and accessibility that larger companies cannot match. This includes being readily available to customers, directly addressing their needs, and selling themselves as much as the product. This personal connection builds trust and loyalty, proving crucial when competing against well-funded, established players, and reinforces the idea that personal investment is paramount in the startup journey.

THE TRANSITION: WHEN TO STOP AND WHEN TO SCALE

While "doing things that don't scale" is vital for early traction, it's not a permanent strategy. The critical juncture is knowing when to transition to scalable operations. This often requires external guidance from advisors or investors who can assess readiness. Founders might get stuck in the consulting revenue trap, mistaking short-term gains for sustainable growth. Ambitious growth targets help differentiate scalable startups from consultancies, signaling the potential for massive expansion once product-market fit is validated.

EMBRACING CHAOS: THE STARTUP WAY OF ITERATION

Startups operating with unscalable methods embrace a certain level of chaos, viewing errors not as failures but as learning opportunities. Unlike large companies that meticulously plan to avoid errors, startups often 'turn on all the water' to identify and fix problems quickly. This iterative, error-prone approach, driven by extreme motivation to fix issues when usage is high, is the startup's inherent advantage. It allows for rapid experimentation, quick pivots, and the ability to build robust solutions from real-world usage, not theoretical models.

Mentioned in This Episode

●Software & Apps

●Companies

●Organizations

●People Referenced

Doing Things That Don't Scale: Startup Edition

Practical takeaways from this episode

Do This

Focus on learning and solving your biggest problem right now.

Manually perform tasks to understand customer needs deeply.

Take professional photos of listings (like Airbnb).

Go to wholesalers and sell clothes by hand (like Fleek).

Implement software directly in customer's systems (like Stripe/Algolia).

Offer direct founder access and personal cell phone numbers.

Buy inventory and photograph it yourself to launch (like Instacart).

Build MVPs using basic tools (Google Drive, forms) to test demand (like DoorDash).

Embrace errors and chaos as part of the startup journey.

Use unscalable methods to answer key questions quickly.

Avoid This

Worry about theoretical scaling problems in the early days.

Wait to build perfect software before getting early customers.

Be afraid of tasks that feel beneath you (cold calls, admin).

Replicate big company playbooks that prioritize error-proofing over speed.

Get addicted to non-scalable consulting revenue.

Avoid getting in the trenches with your customers.

Delay launching because the product isn't 'scalable' yet.

Common Questions

It's an approach, popularized by Paul Graham's essay, that encourages early-stage startups to focus on manual, unscalable tasks to learn quickly, gain customers, and delight them, rather than worrying about future scalability.