Skin in the Game: Y Combinator versus Creative Destruction Lab

David Ferris
7 min readFeb 22, 2023

I have had the unique opportunity to participate in three different startup accelerators: Y Combinator, Creative Destruction Lab, and The Hult Prize. To my knowledge, Phonic (the startup I co-founded) is one of only two companies to ever participate in both YC and CDL, giving us a unique perspective to compare the two programs from the inside. This is not an exhaustive list of pros and cons- I’m simply sharing my story and will let the reader draw their own conclusions.

A rough timeline:

  • From January to August 2019 some university classmates and I (Mitchell Catoen, Devon Copeland, Rares Gosman) won a series of pitch competitions. We spent 5 weeks at the “Hult Prize Accelerator” and went on to pitch for $1M in funding from (among others) Bertil Hult and the Clinton Foundation. We did not win.
  • In January 2020 I participated in the Y Combinator W20 cohort (the so-called “COVID batch”). Our demo day was cancelled and then hastily moved online, so we deferred fundraising until S20. We eventually closed $2.3M USD.
  • In September 2021 we opened a Toronto office and decided to apply to CDL. We were accepted into the Vancouver “Prime” cohort. We participated in two sessions and were eliminated from the program November 2021.

Program Differences

Collaboration vs. Competition

YC’s application has a 2% acceptance rate (compared to CDL’s ~20%). YC also accepts startups from all over the world while CDL focuses on developing Canadian companies. After acceptance, all Y Combinator startups participate in an intense three month program that culminates in a “demo day”, where thousands of investors gather to watch 60 second pitches. About two thirds of YC startups choose to pitch on demo day. During the program, YC companies meet in weekly sessions with groups of ~10 other companies. These sessions resemble group therapy more than Shark Tank, with founders sharing progress updates and the mentors ask probing questions about markets and validation. The final few meetings are geared towards fundraising in preparation for demo day. Because every startup can pitch at demo day, the YC cohorts are intensely collaborative, with founders going out of the way to help one another generate early traction. We learned as much from our batch-mates as we did from our group partners.

By contrast, CDL is elimination style- not every startup makes it to the “super session” (CDL’s version of demo day). At each session startups must be selected by mentors in order to quality for the subsequent session. Sometimes mentors will support a startup through multiple sessions, and other times the startup will work with new mentors. Occasionally strong companies are eliminated simply because mentors don’t feel like they can assist. This elimination approach has the benefit of increased accountability, since startups that fail to achieve their goals risk being disqualified from the program. However, I felt that founders tended to “perform”, selecting less-ambitious goals and spending more time impressing mentors rather than seeking their advice. The result is that YC companies move faster and accomplish more (in spite of CDL being 6 months long and YC only three).

Ultimately, it’s hard to discriminate between early stage companies, and a funnel approach lets CDL progressively invest more time in the startups that perform well in their eyes. It also indicates that CDL may have a lack of confidence in their ability to pick winners and directly compromises the prestige associated with a CDL acceptance.

Skin in the Game: Equity vs. Non-Equity Accelerators

Y Combinator is the most well-known and successful startup accelerator in the world, graduating companies like Stripe, AirBnb and Coinbase. It’s also an equity accelerator. The standard deal for our cohort (W20) was a 7% stake for $150,000 in a SAFE note (although theses terms occasionally change). This immediately valued the business at over $2 million.

By contrast, CDL is an equity-free accelerator that don’t take any stake in the startups they mentor (in fact their non-profit structure prohibits it). What follows is a case study in incentives: The YC partners are bought into YC, and by extension the startups.

YC being “bought in” doesn’t manifest as micromanagement but rather a soberness about a startup’s metrics and reality. YC mentors have a familiar attitude: love for the individual founder, but a dispassionate and calculating eye for the founder’s actions. It can be jarring to be told, “This idea is bad. Delete your code.” but it’s usually delivered with complete optimism. During one pivot I remember being told “Think about all the cool stuff you can build instead of this bad idea!”

Many of the CDL mentors shared this positivity, but I experienced more cynicism from the Canadian mentors and startups alike. Most companies in YC are constantly pivoting between different ideas, while most of our CDL cohort stayed committed to their original visions. I suspect this is because pivoting means temporarily regressing in order to move out of a local optima, a risky play with elimination at stake.

Ironically, the CDL framework implicitly seems to encourage startups to be risk-averse. As one CDL alumni told me, “You gotta play the game.” By contrast, on the first day of YC Dalton Caldwell stated: “Don’t try to hack YC. Build a real business.”

Company Stage: Incubator vs. Accelerator

We participated in CDL after completing YC (and raising ~$2.3 million). The CDL mentors regularly commented that Phonic was quite late stage for the program. Despite this, I felt that CDL and YC are both directed at the same stage of company: the sweet spot where a few founders have a little bit of traction but haven’t yet raised institutional money. In fact, lots join YC in a deliberate discount/down round because the value-add is so substantial. I didn’t see any later stage companies in CDL, although there were some mature biotechnologies and hard science projects.

Focus: Science vs. SaaS

CDL has an explicit science/hard-tech focus. From their webpage, “Creative Destruction Lab (CDL) is a nonprofit organization that delivers an objectives-based program for massively scalable, seed-stage, science- and technology-based companies.”

YC has a bio track (~25% of the companies), as well as a variety of hardware, software, science, and marketplace startups. Most of YCs major wins have come from SaaS (Stripe, Brex, Gusto) and marketplaces (AirBnB, Instacart, Coinbase).

Mentors: Partners vs. Volunteers

I recently asked one of the more senior YC partners why the “YC recipe” hadn’t been successfully copied anywhere. Their answer was that the quality of mentors made the program vastly more costly than it might seem. It’s really hard to cultivate a network of people as smart and successful as the YC partners, much less get them to spend time mentoring startups. The mentors are all successful former founders (although they frequently joke that they couldn’t be that successful).

By contrast the CDL mentors are more numerous, and are typically volunteering their time away from their positions in industry and academia. Some of the mentors are angel investors, although most volunteer for other reasons. We were fortunate to have amazing, supportive CDL mentors, several of whom were currently working on startups themselves (ex. Kim Kaplan). We did however speak with a few startups who complained about their mentors, and there was a general consensus that the network quality was spotty. CDL is definitely aware of this, actively advising companies to seek out relevant mentors. I spoke with a CDL staffer about this, and they candidly shared that they are aware of the mentor quality issues but lack the governance or incentives to cull their network. Of course there are still incredible mentors in the CDL network.

Anecdotally, I also felt like the YC mentors had a little more “crazy like a fox” in them. During one group session we debated whether it was worth building a new version of the NBA. This might have been slightly overdone, as we did not think building a new NBA was a viable idea.

Demo Day vs. Super Session

Y Combinator’s demo day is an exclusive event with a few thousand of the best, most active angels and VCs in the world. They have successfully created an environment where the investor game theory (usually asymmetrically advantaging VCs) is leveled for founders. This is a result of demo day dynamics, historical performance, and the initial SAFE/MFN deal structure.

Since we were eliminated from CDL prior to the super session, I am limited in what I can say on CDL’s demo day. I expect that most entrepreneurs can expect a higher degree of price sensitivity and less competitive rounds. CDL alumni have warned about predatory behavior from investors at the super session (mostly in the form of aggressive deal terms) but I can’t verify these claims. It appears that CDL may need a more robust investor-vetting mechanism.

CDL makes some intros that could lead to further opportunity. YC drops the opportunity of a lifetime in your lap and asks “What are you gonna do with that?”

A Microcosm of Canadian and US Venture Capital

I’m incredibly passionate about the Canadian tech scene. I owe many of my opportunities to the publicly funded University of Waterloo. When raising our first ever seed round, we deliberately saved a portion of our round for Canadian capital, and we incorporated in Canada so that we could retain a Toronto headquarters. We hired engineers almost exclusively out of Waterloo.

Given this, I’m conscious that the above comparison will no doubt seem like CDL bashing. Truthfully, in a head-to-head most accelerators pale when compared against Y Combinator’s 2-decade pipeline of billion dollar successes.

However, the Canadian ecosystem is uniquely frustrating because of the lies we tell ourselves. For years I have heard “Toronto is the next Silicon Valley”, “Waterloo is the new Stanford” and now “CDL is the Canadian YC”. While I can appreciate the energy, these narratives tend to obscure real opportunities for improvement. Initiatives like CDL, Velocity and Communitech should be relentlessly copying and learning from the best. We should be less afraid to say the quiet part out loud.

Let me conclude on a positive note and share a few things that CDL does well.

  • It helps Canadian startups access funding and mentorship
  • It is a watering hole for Canadian investors
  • It lowers the barrier for Canadian scientists to commercialize their research
  • It acts as an index and lets us track early stage Canadian tech performance
  • It is a wonderful network full of friendly, well-intentioned individuals

I hope that CDL is able to continue to build on what they have, and that this resource is helpful for startups participating in either program.