Startup Experts Discuss Doing Things That Don't Scale

A little over ten years ago Paul Graham published the essay "Do Things That Don't Scale." At the time, it was highly controversial advice that spoke to the drastically different needs of an early startup versus the needs of a much larger, more established company.

YC Partners discuss PG's essay, its influence on Silicon Valley, and some prime examples of YC founders that embraced the mantra "Do Things That Don't Scale."

00:00 Intro
02:09 Paul Graham's Essay
04:17 Prioritizing Scalability
05:38 Solving Immediate Problems
08:53 Fleek's Manual Connections
10:32 Algolia and Stripe
12:25 Learning Over Scalability
15:20 Embrace Unscalable Tasks
17:41 Experiment and Adapt
19:06 DoorDash's Pragmatic Approach
21:26 Swift Problem Solving
22:33 Transition to Scalability
23:30 Consulting Services
25:05 Outro

[ycombinator]

What's something you've done at your own startup that is a good example of "doing things that don't scale"?

[jdconley]

PG essays as YouTube with modern examples? I'm here for it! Great stuff!

[kivatinos]

Do things that don't scale is so true, I used to install PCs and run ethernet in the walls of medical clinics just to get to know the medical staff and doctors for my first startup.

[jdphdfv]

I did this exact thing: started a Trader Joe's delivery service. I didn't have outside capital, so instead of buying one of each item like Instacart, I bought people's receipts and took pictures of the items in the store. The service went viral in Manhattan, but Trader Joe's corporate and their lawyers came after me extremely aggressively. Nonetheless, this is good advice.

[GlobalVentures-TSkillsLLC]

Absolutely loved your video discussing Paul Graham's "Do Things That Don't Scale" philosophy! It's such a timeless piece of advice that resonates deeply, especially in the dynamic world of startups and entrepreneurship. At Global Ventures HQ, we're big believers in this approach too. It's all about focusing on personalized, hands-on support to empower entrepreneurs and businesses at every stage of their journey. Looking forward to more insightful content from you!

[winkletter]

Interesting! This is an example of *structural coupling* where organisms and environments change in a complementary way through recursive interactions. We think the organism just adapted to external objects, but the organism and environment have to go through a process of co-creating the object. In other words, founders need an environment (users and other stakeholders) to operate within so they can go through this process of object construction (finding the product.) The users are as much a part of this process as the founders. And you shouldn't/can't scale until this structural coupling occurs, which is product-market fit. But again, it's not the product fitting into an existing market. The founder and users interact to create the product-market coupling.

[YourFractionalCMO]

Such a great video. I just sent it over to one of my clients who needs this advice so badly. Thanks for helping me communicate it well!

[jeroenjansen]

Michael Seibel trying really hard to pretend he has not heard the story 1000 times before

[sandygermsmanson]

Personally, I think one of the best "things that don't scale" things a founder can do is spending a lot of time speaking with customers, even the annoying ones

[mmjretv]

This is journeyman advice and not destination lover business advice.

What I heard:

1. Set the tone even if you have to do it yourself to get the customers you want.
2. Manage and record delight.
3. Be multi-layered in your approach to energy expenditure to secure profits.
4. Don't miss your moment by thinking about scaleable perfection.
5. Realize non-scalable businesses are secure places for VCs to learn how to/ where to store cash.
6. Humility will maximilize your learnings in case someone scales on your idea before you can.

[EcomCarl]

Absolutely agree on the power of non-scalable actions in the early stages of a startup! It’s those initial, hands-on experiences that really teach founders what their customers truly need. 💡

[AbassAdebayo]

This video is for me.

Thank you YC Partners.

[peterbloomingdale]

I manually verify users on my social platform by calling them individually to ensure they are real. Only reason it worked so far is because I have no users 😊

[chapterme]

Chapters (Powered by ChapterMe) -
00:00 - Intro: Early-stage founders need time for learning
00:14 - YC partners discuss do things that don't scale culture
01:07 - Silicon Valley's obsession with scalable business models
05:41 - Airbnb founders inspired by Paul Grahams urgency
07:52 - Unscalable examples Fleeks clothing marketplace
11:23 - Algolia's support for Product Hunt paid off
12:24 - Strong, exciting founders focus on learning, self-service
13:14 - YC works with best founders for learning, trenches with customers
13:29 - Founder FaceTime with customers advantages, sales tips from Ryan
14:12 - Tips for successful sales and marketing
16:12 - Instacart bought Trader Joes for Instacart deal
19:00 - DoorDash founders describe smart one-day tech build
23:05 - Scale up early, avoid consulting revenue addiction
25:13 - Outro: Experiment quickly and fail fast

[matthewmolinar]

This came at the right time. I got a working mvp together in 3 days, and made a demo too. Got some interest online, but then i started noticing it didn't always work earlier this week. Began re-writing the backend. I'm rethinking that. there's a freaking product, let's go find people

[JL-qfhq]

That’s a very fancy way of saying running experiments with the least amount of set up to get directional information, sacrificing accuracy, precision, repeatability for quick validation. It’s akin to running screening experiments in science when we’re not exactly sure what we’re looking for but want a readout on input/output or stimulus/response to start building a model.

[neugey]

14:00 is the most important part, at least in my sector. Building relationships in your user community is so important. Show them you're worthy of trust before getting them to use your service. 9 times out of 10, brand and relationship is more important than selling.

[RevenueMagnet]

Question regarding the Marketplace with Buyers and Sellers: How would you start getting the first user's of a marketplace? Without Seller's, it's not interesting for buyers to browse the marketplace.

[JohnSmith-opls]

This essay didn’t transform anything, it’s basically just a rehash of part of lean startup methodology which was already trendy all over the startup scene at the time and before, this essay was written.

[george_davituri]

Recruit users manually and give them an overwhelmingly good experience, provide a level of service no big company can. It's 2024 don't worry about technical scaling, instead care about scalable business model, product market fit, to satisfy customers and things will auto scale.