ContestMachine, YC Founder Story

Company: ContestMachine Founders: Savraj Singh, Kalid Azad, Lee Linden YC Batch: Summer 2007 Industry: Marketing / SaaS / Promotions Founded: June 2007 | Closed: ~December 2008 Status: Inactive, No longer in business Outcome: Did not reach self-sustaining point. Savraj went on to found Wattvision (YC W09). Kalid joined another startup. Lee went to business school.

---

---

⚠️ Why This Story Is In This Collection

Most YC story collections only cover the unicorns. This one doesn't.

ContestMachine got into YC, ran a real product, collected tens of thousands of entries per day for real customers, and still didn't make it. That story is just as valuable as Stripe's. Because 95% of founders will experience a ContestMachine before they build their Wattvision. Understanding why a YC-backed company closes is one of the most useful things an aspiring founder can learn.

Read this one carefully. The lessons are worth more than the ones in the success stories.

---

The One-Line Summary

Three Microsoft engineers left their stable jobs, got into YC Summer 2007 alongside Dropbox and Arash Ferdowsi, built a real product that actual beauty bloggers loved, then ignored the most important warning Paul Graham ever gave them, and learned the hard way that being technical is not enough.

---

Lens 1, The Before State

Who Were They Before YC?

Savraj Singh grew up in suburban New Jersey, the son of an aerospace engineer father and a molecular biologist mother, both immigrants from the UK. He was the kid who won the science fair in fourth grade for research into aerodynamics, who eagerly read Popular Science every month, who taught himself to code on a family Commodore 64 before anyone around him took programming seriously.

He attended The Lawrenceville School, a prestigious New Jersey boarding school, where his hacker instincts showed up early: in his senior year, he wrote a network worm that messaged the entire school with an OS-level prompt reading "Class of 1999 rulez!", and somehow got away with it.

He graduated from Princeton with a Computer Science BSE in 2003 and walked straight into Microsoft, where he spent four years as a Program Manager on the Office 2007 User Experience Team, working directly under Julie Larson-Green on one of the most significant redesigns in Microsoft's history: the Ribbon interface. He was one of the youngest speakers at Microsoft's TechEd and PDC conferences. By any measure, Savraj had a glittering career ahead of him at one of the world's most powerful tech companies.

His co-founders had similar backgrounds. Kalid Azad was his Princeton roommate who had already left Microsoft a year earlier to work on side projects (instacalc.com and betterexplained.com). Lee Linden brought business perspective to the technical trio.

All three were adherents of Paul Graham's essays before YC was a household name. They were the kind of engineers who read Hackers & Painters and Do Things That Don't Scale and told each other: we could do this.

The Personal Pain They Were Living

Here is the uncomfortable truth about ContestMachine's origin: there wasn't a deep personal pain. There was an observation.

Savraj and his friends would meet regularly to pitch startup ideas to each other. They saw that running contests and giveaways online was clunky and manual, blog owners had no good tool to manage entries, pick winners, set time limits, or collect submissions. It seemed like a clear gap. So they decided to fill it.

The idea was reasonable. The execution would prove capable. But the foundation, personal frustration, lived experience with the problem, was thin.

Why They Almost Didn't Fit the "YC Founder" Stereotype

They fit it almost too well, on paper. Princeton graduates, Microsoft pedigree, strong technical skills. What they lacked was the thing YC actually prizes most: a founder who cannot stop working on a specific problem because they've lived it.

Savraj would later reflect that they left Microsoft to do YC driven by the dream of startups, the Paul Graham essays, the idea of building something of their own, more than by an undeniable pull toward this specific problem.

Key Insight for Aspiring Founders

The question isn't whether you can build the thing, it's whether you're the right person to discover what the thing should actually be. Savraj's technical skills were never in doubt. What ContestMachine lacked was a founder who had run a contest, lost sleep over a promotion, felt the pain in their gut. That gap between "observation" and "lived experience" mattered enormously.

---

Lens 2, The Idea Origin

How the Idea Was Actually Born

The idea grew from regular startup brainstorming sessions among friends. In 2007, the web was genuinely exciting: YouTube had just launched, Facebook was opening to the public, the first iPhone debuted near their Demo Day. Bloggers were exploding. And one of the clear pain points in the emerging creator economy was that there was no easy, embeddable tool to run a contest or giveaway from your website.

ContestMachine's thesis was straightforward: build a web service that handles everything, collecting entries, setting start and end dates, choosing entry requirements, contacting winners, so that a blogger or small business owner could run a promotion in minutes, not days.

It was a genuine product insight. The problem was real. The space was underserved. Three engineers from Princeton and Microsoft could absolutely build it.

The First Version

The first version was built in the summer of 2007, during the YC batch. Savraj recalls opening page 1 of the newly-printed Ruby on Rails documentation, they were early Rails adopters before it was mainstream. This was raw, first-principles building: learn the framework, ship the product, get to Demo Day.

What they shipped worked. Customers could embed a contest widget on their blog, set the rules, and let ContestMachine handle the rest. Within months, the software was processing tens of thousands of contest entries per day, real traction, real usage, real customers.

The Signal That Seemed to Validate It

Their first major customer segment was beauty bloggers, one of the most active communities on the early web, constantly running product giveaways for their audiences. Word spread through blogger networks and ContestMachine found real organic traction in this niche. At one point, Savraj noted, the software was collecting entries from some of the most popular lifestyle accounts on the internet.

For a moment, it looked like it was working.

The Pattern This Follows (And Where It Breaks)

ContestMachine follows the "underserved niche" idea archetype, spot a manual workflow in a growing market and automate it. This pattern works. The failure here wasn't the pattern; it was the execution of the crucial follow-up question: who will pay for this, and how much?

Beauty bloggers loved the product. But would they pay for it? Would the revenue per account be enough to build a business? These questions got less attention than they deserved.

---

Lens 3, The Application Anatomy

How They Got Into YC

ContestMachine was part of YC's Summer 2007 batch, one of the most legendary batches in YC history. Their batchmates included Arash Ferdowsi and Drew Houston, who were quietly building Dropbox in the same cohort.

Savraj and his co-founders applied as a team of three technical engineers with a clear, simple product concept and the credibility of Microsoft backgrounds. In 2007, YC was still early enough that a strong technical team with a plausible product idea had a real chance. The bar was different from today, fewer applicants, earlier-stage ideas accepted, more emphasis on founder quality than traction.

What Made Their Application Work

Their profile was strong on the dimensions YC valued most in 2007:

• Technical depth, three engineers, including a Princeton CS grad and Microsoft PM
• Simplicity of idea, easy to understand, clear problem, obvious market
• Team cohesion, co-founders who had known each other for years (Savraj and Kalid were Princeton roommates)

What the Application Missed

What they didn't have, and this is the crucial lesson, was any direct experience with the problem they were solving. None of them had ever run a contest or promotion. They were three engineers looking at a market from the outside.

Paul Graham noticed this immediately during the interview.

Application Scorecard

---

Lens 4, The Interview Moment

The Hardest Question, And The Warning They Ignored

This is the most important section of the ContestMachine story. And it comes directly from Savraj's own words.

During their YC interview, Paul Graham looked at the three engineers in front of him and said something that should have stopped them cold:

"You guys are making the classic startup mistake. You're three engineers, wading into an industry, contests and giveaways, and saying you can do it better. But has any one of you run a contest or giveaway?"

The answer was no.

Savraj's reflection, years later, is strikingly honest: "That was super prescient advice, and I remember we chose to not really consider it. Lesson learned."

Read that again. Paul Graham, in the actual YC interview, told them the exact reason the company would struggle. They heard it, understood it, and went ahead anyway.

They got in. They built. They launched. They got customers. And eventually, they ran out of momentum in exactly the way Paul predicted: three engineers who understood how to build a contest platform but didn't deeply understand the world of marketers, promoters, and community managers who would need to adopt and pay for it.

The Turning Point

The batch itself was exhilarating. Being in a cohort with the future builders of Dropbox, surrounded by Paul Graham's energy and the YC network, was transformative for how Savraj thought about startups. The YC experience itself, the peers, the weekly dinners, the partner meetings, worked exactly as designed.

The turning point wasn't a single moment. It was a slow reveal: after Demo Day, when it was time to convert interest into sustainable revenue, the limits of their domain knowledge started to show. They knew how to build. They weren't sure enough about whom to sell to, how to price, and how to grow the right customer base.

The Interview Archetype

ContestMachine's interview fits the "founder-market fit challenge" type. YC accepted the team because the founders were clearly strong. The warning Graham gave was real, but the team's technical credibility was enough to get them in anyway. The lesson for aspiring founders: getting into YC is not validation that your idea is right. The partners can invest in a team while still having serious doubts about the idea. Those doubts are worth listening to even when you've been accepted.

---

Lens 5, The Batch Experience

What the Summer 2007 Batch Felt Like

The Summer 2007 cohort is now famous in startup history. Dropbox was in the room. Arash Ferdowsi and Drew Houston were figuring out file sync. The energy of the batch, watching peers work on big ideas while you work on yours, was electric.

Savraj remembers the technical context vividly: this was the summer Ruby on Rails documentation was being printed for the first time. YouTube had just launched to the public. The original iPhone debuted around their Demo Day. The modern web was being built in real time, and they were inside it.

The YC batch compressed their learning about startups, how to talk to customers, how to demo, how to frame a pitch, in ways that four more years at Microsoft wouldn't have.

The Thing YC Gave Them That Mattered Most

For the ContestMachine team, the most valuable part of YC wasn't the $20k cheque. It was the cohort.

Savraj would go on to reference his YC network for the rest of his career. His second company, Wattvision, came directly from the mindset shifts of the YC experience, the permission to build, the courage to pursue contrarian ideas, the framework to think about problems differently. None of that existed before YC.

What Happened After Demo Day

Lee Linden left for business school at the end of the summer, a common pattern among tri-founder teams where one member has an easier exit path. Savraj and Kalid kept running ContestMachine into 2008, but momentum was fading. Kalid eventually joined another startup. Savraj was left alone to make a decision.

The company never reached "ramen profitable." They had real usage but couldn't convert it into a self-sustaining revenue base.

The Growth Number, What Real Traction Looked Like

At their peak, ContestMachine was processing tens of thousands of contest entries per day. This is not nothing. Many startups never see that level of organic usage. The failure wasn't lack of product-market fit in the narrow sense, people used the product and found it useful. The failure was in the business around the product: pricing, customer acquisition costs, revenue per account, and a path to scale.

---

Lens 6, The Mindset Shift

The Limiting Belief They Had to Confront

Savraj and his co-founders entered YC carrying a belief that many technically strong founders hold: if you build a good enough product, the business will follow.

This belief is not entirely wrong, bad products don't succeed. But it creates a blind spot: it leads founders to invest 80% of their energy in building and 20% in understanding the customer's world. For ContestMachine, the inverse would have been more useful.

The belief they needed to challenge, and eventually did, through the experience of ContestMachine shutting down, was: "being the best engineer in the room is sufficient."

What ContestMachine Actually Taught Savraj

Failure is rarely complete. ContestMachine gave Savraj something that no Microsoft salary could buy: the lived experience of building and launching a product, acquiring real customers, navigating the YC process, and deciding what to do when a company doesn't work.

When he found his next idea, watching his electricity meter with no way to see real-time usage on his phone, he already knew how to build, how to apply to YC, how to talk to early customers, and what questions to ask himself before committing. He had already paid the tuition.

Wattvision (YC W09), his second company, ran successfully for over seven years. He built it as a solo founder, applied with a working hardware prototype, and was accepted by Paul Graham himself after trekking to Cambridge with a demo. The ContestMachine experience made this possible.

The Identity Shift

The clearest articulation of Savraj's shift is in how he chose his companies after ContestMachine. Every role since, Wattvision (energy monitoring hardware), Daily Harvest (food-tech CTO), Bubble Goods, Nucleus Genomics, shares a common thread: small teams, hard problems, clear mission, tight loop to real customers. That is not a Microsoft PM's framework. That is a YC-trained founder's framework, shaped by building something that didn't work.

The Transferable Principle

Failure at YC is not failure, it's compressed tuition. ContestMachine ran for 18 months, served real customers, and taught three smart engineers things they could not have learned anywhere else. The measure of a YC experience is not whether your first company succeeds. It's whether you use what you learned to become the founder your second idea deserves.

---

Lens 7, The Replicable Playbook

Action 1, This Week: Run The "Have You Lived It?" Test

Before you go further with your current idea, answer Paul Graham's ContestMachine question honestly:

"Have you personally experienced the exact problem you're solving? Not observed it. Not inferred it. Lived it."

If the answer is yes, write down the three most specific, painful moments of that experience. Those moments are your product roadmap.

If the answer is no, that's not automatically a dealbreaker, but it means your first job is to become a deep domain insider before you build anything. Go work in the industry for 3 months. Interview 50 people who have the problem. Run the process manually yourself before you automate it.

Action 2, This Month: Do The Manual Version First

ContestMachine built software to automate contest management before deeply understanding how contests were actually run. The better path: run five contests manually for five bloggers, using spreadsheets and email, before writing a single line of code.

This month, find 5 potential customers for your idea. Offer to solve their problem for free, manually. You'll learn more in two weeks of doing this than in three months of building.

This is what Paul Graham calls "doing things that don't scale." ContestMachine could have done this. They didn't.

Action 3, Before Applying to YC: Separate Team Strength From Idea Strength

ContestMachine teaches a critical lesson about the YC application: a strong team can get you in even when the idea has a fundamental weakness. This is useful to know, but it means you cannot use your YC acceptance as proof that your idea is right.

Before applying, pressure-test your idea separately from your team quality. Ask yourself: if three people I'd never heard of applied with this exact idea, would YC fund it? If the honest answer is no, keep working on the idea before you apply. Don't rely on your credentials to carry a weak thesis.

Story Relevance Tags

---

The Batch They Were In, And Who Else Was There

YC Summer 2007 is one of the most studied batches in accelerator history. While ContestMachine was figuring out the contest market, the following founders were in the same room:

• Drew Houston & Arash Ferdowsi, building Dropbox
• A cohort that Paul Graham later described as the moment YC shifted from "interesting experiment" to "this is how great companies get started"

Being in a legendary batch does not guarantee your company succeeds. But it does guarantee you learn from people who are building legendary things. That learning compounds across careers, not just companies.

---

3 Things You Can Screenshot Right Now

"Paul Graham told them, in the actual interview, exactly why the company would struggle. They heard it, got in anyway, and built it anyway. Then it happened exactly as he said. Always listen to the specific doubts, not just the acceptance."

"ContestMachine had tens of thousands of entries per day and still didn't make it. Usage is not the same as revenue. Revenue is not the same as a business. Know which one you actually have."

"Failure at YC is not failure, it's compressed tuition. Savraj Singh left ContestMachine, applied to YC again as a solo founder, and built Wattvision for 7 years. Your first YC company is your MBA. Use it that way."

---

What Happened to the Founders After

All three built substantial careers. None of them stopped because ContestMachine didn't work. That is itself a lesson worth noting.

---