Instacart doing things that don't scale

In the world of startups, there’s a well-known piece of advice from Y Combinator’s Paul Graham: “Do things that don’t scale.” This counterintuitive wisdom suggests that in the early days of building a company, founders should embrace labor-intensive, manual processes that could never work at scale—but might be exactly what’s needed to get initial traction. Few companies embody this principle more vividly than Instacart, the grocery delivery service that grew from a scrappy startup to a multi-billion-dollar company by starting with some decidedly unscalable methods.

The Founder’s Journey: 20 Failed Startups Before Instacart

Apoorva Mehta’s path to founding Instacart was anything but direct. After working as a logistics engineer at Amazon, Mehta spent two years building approximately 20 different “minimum viable products” across various industries—enterprise software, consumer apps, advertising platforms—all of which ultimately failed. Rather than becoming discouraged, Mehta treated each failure as a learning opportunity, developing skills in rapid prototyping, business strategy, and managing the psychological roller coaster of entrepreneurship.

In 2012, Mehta had his grocery delivery epiphany. Growing up in a small town near Toronto, Canada, he remembered waiting at bus stops in freezing weather, laden with grocery bags. He realized that while technology had transformed many aspects of daily life, grocery shopping remained largely unchanged. People still had to physically visit stores, navigate aisles, wait in checkout lines, and carry heavy bags home—a time-consuming process that had remained essentially the same for decades.

Mehta envisioned a service that would allow customers to order groceries from their favorite local stores and have them delivered within an hour. This wasn’t a novel concept—companies like Webvan had famously tried and failed in the early 2000s—but Mehta’s approach would be fundamentally different. Instead of building warehouses and managing inventory, Instacart would leverage existing grocery infrastructure and use technology to connect customers, stores, and a network of shoppers who would pick and deliver orders.

Y Combinator and the Missed Deadline

Mehta applied to Y Combinator, the prestigious startup accelerator, to help turn his vision into reality. There was just one problem: he missed the application deadline—not by a day or even a week, but by two whole months. As a single founder without a team, his chances seemed slim.

In a bold move that foreshadowed Instacart’s scrappy approach to problem-solving, Mehta used his own product to get noticed. He ordered a six-pack of beer through his rudimentary Instacart app and personally delivered it to Y Combinator partner Garry Tan. The unorthodox application worked—Instacart was accepted into Y Combinator’s summer 2012 batch, and Mehta soon connected with his co-founders, Max Mullen and Brandon Leonardo.

The Trader Joe’s Catalog: Buying and Photographing Every Item

With Y Combinator’s backing, the Instacart team faced their first major challenge: how to create an online catalog of grocery products when most stores didn’t have public APIs or comprehensive digital inventories. This was particularly true for Trader Joe’s, a popular grocery chain known for its unique, private-label products but lacking a robust online presence.

The solution was as labor-intensive as it was ingenious. The Instacart team took their Y Combinator funding and did something that would be impossible to sustain at scale: they went to their local Trader Joe’s store, purchased one of every single item on the shelves, and brought the entire haul back to their office.

They then set up an impromptu photography studio and methodically photographed each item—hundreds of products ranging from produce to frozen foods, snacks to household items. These photos, along with manually entered product information like names, prices, and descriptions, became the foundation of Instacart’s first digital catalog.

This painstaking process allowed Instacart to offer something no other service could at the time: the ability to browse and order from Trader Joe’s complete inventory online. For customers who loved Trader Joe’s unique offerings but dreaded the notoriously crowded stores and limited parking, this was a compelling value proposition.

The manual catalog creation wasn’t just about having product images—it was about creating a seamless customer experience from the very beginning. By photographing actual items from the store, Instacart ensured that what customers saw online accurately represented what they would receive. This attention to detail helped build trust with early users, a crucial factor for a service asking people to outsource something as personal as grocery shopping.

Fulfilling Orders Without Shoppers

Another unscalable practice in Instacart’s early days involved order fulfillment. Ideally, the platform would have a network of shoppers ready to pick and deliver orders as they came in. But building that network takes time, and Instacart needed to serve customers immediately to gain traction.

Mehta’s solution? When orders came in and no shoppers were available, he and his team would personally fulfill them. The founders would drop whatever they were doing—coding, business development, investor meetings—to shop for and deliver groceries themselves.

This hands-on approach served multiple purposes. First, it ensured that early customers had a positive experience, even if it meant the CEO was delivering their groceries. Second, it gave the founding team direct insight into the shopping and delivery process, helping them identify pain points and opportunities for improvement. And third, it demonstrated their commitment to making the service work, regardless of the personal sacrifice required.

As Mehta explained in interviews, this willingness to do things that don’t scale was crucial for understanding their business from all angles. By personally shopping for orders, the team learned firsthand about store layouts, product availability, checkout processes, and delivery logistics—insights that would have been difficult to gain any other way.

The Personal Touch: Building Relationships with Store Managers

In addition to creating their own product catalog and fulfilling orders themselves, the Instacart team took another unscalable approach: building personal relationships with store managers at their partner locations.

Rather than pursuing corporate partnerships from the top down—which would have been difficult for an unproven startup—they started at the store level. Team members would visit individual grocery stores, introduce themselves to managers, and explain the Instacart concept. They would address concerns, answer questions, and emphasize how Instacart could drive additional sales without requiring any integration or effort from the stores themselves.

This store-by-store, manager-by-manager approach was time-consuming but effective. It allowed Instacart to start operating in new locations quickly, without waiting for formal corporate agreements. It also helped the team build goodwill with the people they would interact with daily as they shopped for customer orders.

The personal relationships proved valuable when issues arose, as they inevitably did. If an Instacart shopper encountered a problem or needed assistance, having a friendly relationship with the store manager could make all the difference in resolving the situation quickly and maintaining service quality.

Testing and Iterating in a Single Market

Another aspect of Instacart’s unscalable beginnings was their intense focus on a single market—San Francisco. Rather than attempting to launch in multiple cities simultaneously, they concentrated all their efforts on making the service work perfectly in one location.

This geographic constraint allowed the team to be physically present in the stores where they operated, to personally deliver orders when needed, and to quickly address any issues that arose. It also enabled them to iterate rapidly on their product based on real customer feedback.

For example, when they discovered that customers wanted more precise control over produce selection (like specifying ripeness of bananas or avocados), they could immediately update their app to include these options. When they realized that certain items were frequently out of stock, they could develop substitution workflows to maintain customer satisfaction.

This intense focus on a single market might have seemed limiting, but it allowed Instacart to refine their service to the point where it truly delighted customers—creating vocal advocates who would spread the word about the service to friends and family. Only after establishing this strong foundation did they begin to expand to additional cities.

The Pivot from Unscalable to Scalable

As Instacart gained traction and began to grow, they gradually transitioned from these manual, unscalable processes to more sustainable systems. They developed relationships with data providers to access standardized product information, built tools to help shoppers efficiently fulfill orders, and created training programs to ensure consistent service quality as they expanded their shopper network.

The photography studio gave way to professional product images, the founders stopped personally delivering groceries, and corporate partnerships replaced individual store relationships. But the insights gained during those early, unscalable days informed every aspect of how Instacart built their scaled operations.

For instance, having personally shopped for hundreds of orders, the team understood exactly what information shoppers needed to efficiently navigate stores and find products. Having manually photographed thousands of items, they knew which product details were most important to customers making purchasing decisions. And having built relationships with store managers, they understood the concerns and priorities of retail partners.

Lessons from Instacart’s Unscalable Beginnings

Instacart’s journey from photographing every item in Trader Joe’s to becoming a grocery delivery giant offers several valuable lessons for entrepreneurs:

1. Start with what works, not what scales. Instacart’s willingness to do things manually—from creating product catalogs to fulfilling orders themselves—allowed them to launch quickly and learn directly from customer experiences.
2. Use constraints as advantages. By focusing intensely on a single market, Instacart could provide exceptional service and iterate rapidly, building a loyal customer base before expanding.
3. Get your hands dirty. The founders’ direct involvement in shopping and delivery gave them insights that would have been impossible to gain otherwise, informing product decisions and operational improvements.
4. Build relationships at all levels. Instacart’s store-by-store approach to building retail relationships allowed them to operate and expand without waiting for corporate partnerships.
5. Learn from failure. Mehta’s experience with 20 failed startups before Instacart taught him valuable lessons about rapid prototyping, business strategy, and resilience—all of which contributed to Instacart’s eventual success.

The Legacy of Doing Things That Don’t Scale

Today, Instacart is a household name, valued at billions of dollars and operating in thousands of cities across North America. The company has sophisticated technology for inventory management, shopper dispatching, and delivery optimization. They have formal partnerships with major grocery chains and a vast network of shoppers fulfilling orders.

But none of this would have been possible without those early days of doing things that didn’t scale—buying and photographing every item at Trader Joe’s, personally delivering orders, and building relationships one store manager at a time.

Instacart’s story reminds us that behind many successful technology companies are founders who weren’t afraid to embrace manual, labor-intensive processes to get their businesses off the ground. It’s a testament to the power of Paul Graham’s advice to do things that don’t scale—not as a permanent strategy, but as a way to start, learn, and build something that can eventually grow into something much bigger.

As Mehta himself has said, “Great founders are learning machines.” Instacart’s unscalable beginnings provided the perfect environment for learning—about customers, operations, retail partners, and the grocery industry as a whole. Those lessons formed the foundation upon which they built their scaled, technology-driven approach to grocery delivery.

In a world obsessed with automation and efficiency from day one, Instacart’s story is a powerful reminder that sometimes the best way to build something innovative is to start with the most human, hands-on approach possible. By embracing the unscalable, they ultimately created something that could scale to serve millions of customers across North America—one photographed Trader Joe’s item at a time.