Summary - Generic
Do Things that Don't Scale
This summary focuses on Paul Graham's essay "Do Things that Don't Scale," which offers crucial advice for early-stage startups. The core message is that startups don't succeed automatically; founders have to actively make them take off.
Manual User Acquisition
The most common "unscalable" thing founders need to do is recruit users manually. This applies to almost all startups, even those with seemingly urgent solutions like Stripe.
The Collison Installation
Graham highlights Stripe's aggressive user acquisition, using the term "Collison installation" to describe their tactic of setting up users on the spot instead of passively sending links.
Why Founders Resist
Founders resist manual recruitment due to:
- Shyness and Laziness: They prefer coding to talking to strangers and facing potential rejection.
- Small Initial Numbers: They underestimate the power of compound growth.
Compound Growth
Graham emphasizes measuring progress by weekly growth rate. Even small weekly growth, compounded over time, leads to significant user bases. He cites the example of growing from 100 to 14,000 users in a year with a 10% weekly growth rate.
Airbnb: A Classic Example
Airbnb exemplifies this approach.
They tackled the inherent difficulty of launching a marketplace by going door-to-door in New York, recruiting users and helping them improve their listings.
Initial Fragility
Early-stage startups are incredibly fragile. Inexperienced founders and investors often misjudge them by the standards of established companies. Graham compares this to judging a newborn baby's potential.
Don't Dismiss Your Startup
The biggest danger is founders dismissing their own startups. Even Bill Gates underestimated Microsoft's potential, returning to Harvard after starting the company.
Focus on Potential
The key question is not "Is this company taking over the world?" but "How big could this company get if the founders did the right things?"
Finding Users
Graham suggests finding users by:
- Solving your own problems: This makes it easier to find your peers.
- Observing early users: Identify the most enthusiastic users and find more like them.
He cites Ben Silbermann targeting design bloggers for Pinterest.
Delighting Customers
Founders should take extraordinary measures to make users happy. Wufoo, for example, sent handwritten thank you notes to each new user.
Why Founders Don't Delight
- Engineering Background: Customer service isn't part of traditional engineering training.
- Scaling Concerns: They worry about unscalable efforts.
- Lack of Personal Experience: They haven't experienced exceptional customer service themselves.
Scaling Delight
Graham argues that delighting customers often scales better than expected, partly because you can find ways to scale anything, and partly because it becomes ingrained in the company culture.
The Advantage of Being Small
Small startups can provide a level of service no big company can.
The Contained Fire Strategy
Focusing on a deliberately narrow market can be an effective strategy.
It's like containing a fire to make it hotter before adding more logs.
Facebook's Approach
Facebook initially targeted only Harvard students, creating a sense of community and achieving a critical mass of users before expanding to other colleges.
Startups as Early Adopters
Other startups are often the best early adopters, being more open to new things and growing quickly.
Pulling a Meraki
For hardware startups, Graham suggests "pulling a Meraki," which means assembling products yourself to overcome the high cost of factory production runs.
Pebble's Example
Pebble assembled their first few hundred watches themselves, which helped them gain traction and ultimately raise $10 million on Kickstarter.
Benefits of DIY
- Faster design iteration
- Valuable learning experiences
Consulting and Manual Work
For B2B startups, Graham suggests acting as consultants for a single user, perfecting the product for their needs before expanding to others.
Viaweb's
Experience
Viaweb used their own software to build online stores for merchants who were initially reluctant to use it themselves.
This provided valuable feedback and helped them understand their users.
Manual First, Automate Later
Startups can do things manually at first, automating later as they grow. This allows for faster launches and better understanding of what to automate.
Stripe's "Instant" Accounts
Stripe initially provided "instant" merchant accounts by manually signing up users for traditional accounts behind the scenes.
Avoiding the Big Launch
Graham advises against big launches, arguing that they don't matter as much as founders think.
Focus on User Happiness
What matters is making initial users happy, not the number of users at launch.
Partnerships Don't Work
Partnerships, especially with big companies, rarely work for startups, especially in the early stages.
Effort Matters
Any strategy that omits effort, whether relying on a big launch or a big partner, is suspect.
Startups as Vectors
Graham suggests thinking of startup ideas as vectors, with two components:
What you're going to build 2. The unscalable thing(s) you'll do to get started
The Importance of Both Components
Founders need to work hard in both dimensions.
Unscalable Actions Shape the Company
The unscalable actions taken early on can shape the company's DNA, leading to lasting benefits like a focus on user acquisition and a culture of delighting customers.
This summary provides a comprehensive overview of the key points in Paul Graham's essay, highlighting the importance of unscalable actions, customer delight, and a relentless focus on growth for early-stage startups.