It’s been a decade since writing “Welcome to the Unicorn Club,” so it seems a good time to reflect on the wild ride we’ve been on the past ten years.

So much happened, this became a long post! Below we share context and takeaways, then more in depth explanations. For a visual journey, there’s a companion slide deck with lots of pictures. So grab a beverage and settle in for what we hope is a fun, eye opening read.

‍

A refresher on 2013

Cowboy Ventures had just gotten started. To inform our investment strategy, we assembled and studied a dataset of US-based, VC-backed startups that had grown to be worth $1B+ within ten years, using ‘unicorn’ as a shortener to capture their magic.¹ 

‍

‍

Our original analysis found just 39 unicorns out of thousands of startups funded. Some highlights: 

• The majority (62%) had “exited” - gone public or been acquired. 
• Most were consumer-oriented: ~80% of the value, ~60% of companies.
• Enterprise-oriented companies had great capital efficiency (26x), 2.4x better than consumer companies.² 
• One company in a decade became a “superunicorn” (worth >$100B): Facebook.
• Contrary to the prevailing stereotype, average age at founding was 34; 38 for enterprise software companies. 
• Co-founding teams of 3 with common work, school and tech experience were the majority. 
• The Bay Area was unicorn HQ (70%). NY, home to 3, was the second largest hub.
• There was very little diversity - no female CEOs, just 5% had a female co-founder.

‍

2013 to 2021: a rising tide for VC funds, startups, and valuations

Historical returns, growing markets (social, mobile, cloud, commerce, security, crypto, AI…), COVID-era effects and low interest rates attracted 3x more capital ($580B more!) to VC funds between 2013-2021 (btw, that’s also $11B+ more in VC fee income). 

‍

This enabled established VC firms to raise record-breaking funds; public “crossover” funds joined the party too. Over a thousand newer firms also raised funds. The industry gained thousands of new investors with fresh ‘checkbooks’ and limited mentorship or oversight, given the pace. 

‍

2021 was a perfect storm of near zero interest rates while much of the world spent their days behind screens (zooming, shopping, streaming, trading, therapizing…), immersed in tech. Due diligence processes were rushed, round sizes and valuations broke records, and a huge herd of unicorns was crowned.

‍

2022 to the present: the tide has turned

In March 2022, the Fed raised rates, triggering a multi-year downward effect on public multiples and enterprise budgets. Although big VCs continued to raise big money (64% of 2022 VC raised went to >$1B funds), investors mostly froze (~40% of VCs stopped dealmaking in 2023). Companies refocused on margins and profitability, cutting costs in several waves. Unicorns started to fall via down rounds, public delistings and shut downs. 

‍

Sadly, we’re far from done. It’s a dynamic time to snapshot and learn from unicorns. 

‍

‍To note: last-round valuations are an imperfect gauge, and likely inflate current herd size and value (Cerebral, Clubhouse and OpenSea may be examples). And, point-in-time methodology misses some companies. For example, Stripe was founded in 2010, Zoom in 2011, Snowflake and Coinbase in 2012. None was a unicorn in 2013, and are now more than 10 years old, so excluded from this dataset.
‍‍

Welcome to the 532 companies in the 2023 Unicorn Club!‍
‍
‍Takeaways, with deeper explanation following³:

1. Unicorns ballooned 14x in the past decade, from 39 to 532! They now serve a wider array of sectors (we’re tracking 19) from climate and crypto to vertical SaaS. 
2. The pendulum swung hard to enterprise (78%), the inverse of 2013.
3. It’s a bloated herd that will thin in coming years (likely to about 350) because…
   • A whopping 93% are ‘Papercorns’ - privately valued companies. 
   • 60% are ‘ZIRPicorns’ - their last valuation is from ‘20-22, when interest rates were near zero. Many are running out of runway.
   • Many are on the cusp: 21% are valued just at $1B (sorry?!)
   • ~40% are trading below $1B in secondary markets.⁴
   • BUT. There is lots of substance in this herd.
   • And we see evidence of a Software Unicorn Power Law - the US will be home to more than 1,000 unicorns by 2033.
4. There’ve been very few exits - only 7% (35 companies) vs. 66% a decade ago.‍
5. Capital efficiency declined significantly. This will be bad for exits, venture returns, founders and employees.‍
6. OpenAI looks to be the 1st superunicorn of the decade, and AI likely the megatrend.‍
7. The Bay Area gained in #s but lost ground as Unicorn HQ while other hubs grew. ‍
8. More unicorns and geos, more co-founders with diverse backgrounds - but some things didn’t change at all.
9. Diversity still wanted - still lots of opportunity to improve founding team composition.‍
10. If past is prologue, expect a mixed outlook for the current herd, and many more unicorns in the future. ‍

‍

Read on dear reader, for deeper insights, explanations and predictions.

‍

1. Unicorns ballooned 14x in the past decade, from 39 to 532

_‍This herd is worth $1.5T in aggregate value (vs. $260B in 2013).

‍

• It’s still really rare and improbable - less than 1% of VC-backed startups became unicorns. It’s 5x more likely for an exceptional candidate to get into Stanford, Harvard or MIT than to found a unicorn.

• Unicorns now span a wider array of sectors (we’re tracking 19!). Most sectors weren’t on our 2013 list (social, commerce, general enterprise dominated). Click below for more about the most valuable companies in each sector.
  ‍

{{slidedeck}}

‍

• The most valuable unicorns in 2023 have innovated across a wide spectrum that stretches from general AI and healthcare, to food delivery, HR software and stock trading.

• There has been a big broadening from the sectors on the 2013 list (and as previously noted, last-round valuation is an imperfect gauge: 75% of the “unexited” companies on this top 20 list were last valued in 2022 or earlier).

• Unlike in our 2013 analysis, there were no best years for unicorn founding. 2020 & 2021 accelerated the ‘crowning’ of all ages and stages of companies. Today’s unicorns are 7 years old on average, same as a decade ago, which seems a good thing.

• We don’t analyze fallen unicorns here. But anecdotally, we see being crowned a unicorn too quickly could be a curse. Fallen unicorns like Hopin and Bird were crowned within one year of founding; car leaser Fair in two years; Convoy and Knotel just three years after founding.

‍

2. The pendulum swung hard to enterprise (78%), almost the inverse of 2013

• 10 years ago, 15/39 (38%) of unicorns were enterprise, comprising $55B in total value (20%). Workday, ServiceNow, Splunk, and Palantir were the most valuable enterprise unicorns at the time.

• Today, there are 416 enterprise unicorns — making up 78% of the list — worth $1.2 trillion and driving 80% of aggregate value (versus 20% in 2013).

‍

• Consumer companies make up 20% of the herd’s aggregate value, versus 80% in 2013. Remember “SoMoCo”? Social (Facebook, Twitter, Pinterest), mobile (Uber, Square) and e-commerce (Groupon, Gilt, Fab).

• The most valuable consumer companies today are in areas like last-mile delivery (DoorDash, GoPuff), health (Devoted, Ro, Cityblock), and gaming-driven platforms (Discord, RecRoom) - companies that drove new habits during COVID.

• What caused so much capital to flock to enterprise companies in the past decade? The attraction of historical capital efficiency, the growth in adoption and predictability of SaaS business models (high gross margin & customer retention), and a growing number of highly valued potential acquirers were likely a big draw. Global adoption of cloud made it easier for customers to adopt new software and opened a big window for a whole new ecosystem of application layer, infrastructure, data & analytics and security companies to be built.

• The cyclical pendulum does swing. Given the hard shift to enterprise, we hope and expect more exciting consumer unicorns will be born in coming years. For inspiration -  many of today’s leading consumer internet experiences are ~2 decades old (Ebay, Expedia, Opentable, Tripadvisor, Stubhub, Yelp), possibly fertile territory?

‍

3. It’s a bloated herd that will thin in coming years (to ~350) because…

• A whopping 93% of unicorns are ‘Papercorns’ - they’re privately valued on paper, not yet ‘exited’ or ‘liquid’. It’s a big change vs just 36% private in 2013.

• 60% are what we call ‘ZIRPicorns’ - they were last valued between Jan 2020 and March 2022, when interest rates were near zero and multiples at peak.

• When money was flowing, unprofitable private companies were often able to raise enough to fund 2-5 yrs of operations. This means operating runways are getting short for many unicorns. Many are working to get profitable on their existing cash, which is challenging given the current economy - and even harder when starting with a low gross margin business.

• Given the chilly current M&A environment, founder resistance to recaps, and investor fear of ‘catching a falling knife’ investing in a down round - we expect more abrupt shutdowns in 2024 (e.g. Convoy, Olive Health, Zume). Zombiecorns to Unicorpses, anyone (yeah, groan)?

• Coining the term also seems to have affected the market (sorry?!). 21% are valued at ~$1B (vs 10% in 2013), on the cusp of “unicorn” - and 46% at <$2B.

• Around 40% are trading below a $1 billion valuation in the secondary markets, according to real order data for 290 unicorns on our list from Hiive. It’s possible more might trade at less than $1 billion in the secondary markets.

• On the bright side, we see ~350 healthy businesses of substance in this herd - almost 10x in a decade! Many ZIRPicorns and papercorns raised enough capital and are navigating the new climate to grow into and beyond their last valuation. Many will become public companies in the coming years with rock solid financials, tested by the downturn. And a few of these future public companies will grow to become superunicorns.

• And, we also see evidence of a Unicorn Software Power Law.  

• Unicorns grew an amazing ~30% on average YoY in the decade, spurred by enterprise software spend, consumer tech adoption, VC funding and interest rates. 

• We see an echo to Moore’s law here. As compute capacity, capability and usage increase, unicorns grow as well. The current momentum in AI should add fuel to innovation and demand.

• Adjusting the current herd to 350, future unicorn growth to a less bubbly 15% y/y, and improving current capital efficiency a bit - we’ll have 4x more, or ~1,400 US unicorns in 2033. 

• Counter arguments include scarcer VC funding, higher interest rates, and software consolidation. But, previous downturns have been fertile for unicorn founding. This will be exciting for the future of innovation, jobs and the tech economy, despite current conditions.‍

‍

4. There have been few exits - only 7%, vs. 66% in the prior decade

• A paltry 35 of 532 are public or were acquired for >$1B, a consequence of increased private capital, investors willing to invest at higher-than-public multiples, and a more challenging regulatory and M&A environment. 

• Founders took more ‘money off the table’ than ever ($700M at WeWork, $200m at Hopin). This practice can misalign founder incentives with investors and employees, and reduce urgency for broader liquidity.

• Average time from founding to an IPO or acquisition was a speedy 6 years. 

• An impressive 75% of founding CEOs led from founding through an exit.

• The elite public unicorn club: just 3% (vs. 41% in 2013) are public, a huge change (also an impact of private capital, and investors willing to invest in richer-than-public valuations in the last decade).

• These companies are split between enterprise and consumer, and span a wide array of sectors.

• Around 70% of current public CEOs were founding CEOs. It’s so impressive to go from startup founder and individual contributor to lead and manage a multi-$B public company - big, big kudos.

• Reflecting the times, there are more fallen public unicorns than healthy public unicorns in this herd. At least 20 unicorn companies went public in the decade then fell below <$1B in value (hello SPACs), vs. just 14 current public unicorns. These, and companies previously valued >$1B but subsequently acquired or recapped at a lower valuation, are not included in our analysis.

• And, just 21 companies were acquired, spanning a diversity of sectors (and 2/3 enterprise, 1/3 consumer). Average acquisition price was $2.4B, almost 2x 2013. Notably, 33% involve hardware (e.g. Cruise, Raxium, Ring, Zoox), likely a consequence of low interest rates for both unicorns and acquirers.

‍

5. Unicorn capital efficiency dropped significantly - precipitously for enterprise companies

‍This will be bad for exits, venture returns, founders and employees. 

‍

• The most successful VC-backed tech companies have historically delivered ‘outsized returns.’ Delivering a 26x return in 10 years is outstanding (~40% IRR) - but investors and talent have to risk years of illiquidity versus other asset classes or industries. 

• In the past decade, tech lost its capital efficiency edge. Formerly impressive enterprise company capital efficiency of 26x plummeted to 7x, putting it in-line with consumer co efficiency (despite typically higher margins and customer retention), which also dropped from 11x to 7x. Given many unicorns are currently overvalued, even 7x is likely inflated.

• In other words, investors would have been better off investing in public superunicorns like Salesforce, Amazon,and Microsoft, up 8x, 9x, and 9x respectively in the decade, than in many companies in our current unicorn herd.

‍

An example to illustrate capital efficiency:

• Let’s consider a company that raised $600M in preferred stock rounds plus $100M in debt in the past decade.

• With strong customer traction and a magnetic CEO, in 2021 the company hits $100M ARR and quickly attracts multiple term sheets, raising at a $3B valuation, mirroring 30x high-growth public comps. 

• Employees get common options at 50% the preferred price — a $1.5B valuation.  

• Then comes 2022. Churn skyrockets and new sales dry up. The company reduces headcount from 1,000 to 500 over several waves of layoffs. The team fights like hell and gets back to $100M ARR by 2025, and almost reaches break even.

• In 2025, the company is offered a $500M acquisition offer by a PE firm, or 5x revenue, the comparable public market multiple.

• Management takes the deal, agreeing to a 8% ‘carveout’ for themselves ($20M) and employees ($20M) as stock options are “underwater.” Debtors get paid in full. Investors get only 60% of their investment back, and employees who stay with the company get about $30k at exit after 4-10 years of work. 

• The company’s capital efficiency: .83x ($500m/$600m equity raised).

‍

Ok, back to capital efficiency.

• Interest rates again had a big negative impact here. As rates lowered, investors relied on historical venture returns, while competition grew for allocation in ‘hot deals’. This caused many to overlook valuations, business model margins, payback periods and burn rates.

• In our 2013 analysis, Workday and ServiceNow were efficiency standouts, each at 60x. Their market caps are also 5x and 19x bigger now, respectively.

• If you’re looking for more evidence companies raised too much in the past decade: about 20% of companies on our list are worth less than 4x their capital raised. Given how many are likely overvalued, reality will likely be worse than this. 

• Categories with lowest average capital efficiency: climate/energy, real estate, and healthcare.

‍

6. OpenAI is likely the decade’s superunicorn, and AI the megatrend

• Prior major waves of tech innovation have each crowned a “superunicorn”, a venture-backed co that grows to be worth >$100B+ over time - e.g. Microsoft, Cisco, Amazon, Meta (Meta the only to be ‘crowned’ in <10 yrs). 
• OpenAI is close to being the first AI superunicorn (rumored to be raising at >$100B) at just 8 years old.
• At times, crypto looked like it would be the megatrend of the decade. Coinbase hit $76B in value in Nov ‘21 after going public months earlier, but is now worth ~$32B.⁷ 

• There are now 15 VC-backed tech superunicorns, and they got a lot more valuable in the past decade. Meta was worth $122B in 2013 and is now worth ~$950B, or 8x.

• Superunicorns have superpowers that help create and/or disrupt entire categories - like Netflix (worth more than Comcast, Paramount, Warner Bros combined) and Tesla (> next 5 largest public automakers combined).

• Of our original dataset, three more became superunicorns in recent years: ServiceNow, Uber, and Palo Alto Networks. And Airbnb is on the horizon at $88B. They’ve grown to be worth more than Hyatt and Marriott combined, and have the benefit of network effects. 

• Superunicorn power may compound further in the coming years with software consolidation and tighter capital constraints for smaller scale players.  

‍

7. The Bay Area gained in #s, but lost ground as unicorn HQ, while other hubs grew 

• The significance of geography continues to evolve, given COVID-era effects. Many unicorns are now spread across multiple hubs and offer hybrid work (and notably, at least 22 have no physical HQ office, according to FlexIndex). 

• But there are clear economic and local network effects to geos as unicorn hubs. And many cite productivity and creativity gains from in-person work, relevant for hopeful future unicorns.

• The Bay Area is still the largest unicorn pasture, but lost a lot of ground  - from 69% to 45%. On the bright side, it’s home to 238 unicorns (incl. the 4 most valuable: OpenAI, Databricks, DoorDash, Samsara), 9x more than 2013.

• Can the Bay regain its stature as a unicorn central? A lot will depend on return to work effects, the concentration of AI talent, quality and cost of living vs. other hubs - and the next generation of unicorns.

• New York grew a lot (11% to 19%!) as the second largest hub, now home to 100 unicorns (about 40% are crypto/web3 or fintech; e.g, OpenSea, Chainalysis).‍

• Many geos grew to be home to >10 unicorns: LA (CloudKitchens, Blockdaemon), Boston (Devoted Health, Circle), Seattle (Auth0, Outreach), Austin (Everlywell, WorkRise), Chicago (VillageMD, Tegus), Denver (Guild, Crusoe).

‍

8. More unicorns and geos, more co-founders with diverse backgrounds - but some things didn’t change at all

• 14x more unicorns, ~14x more co-founders! Our list now includes 1,300+ founders vs. ~100 in 2013.⁸ What didn’t change much:

• 83% of unicorns had co-founders (90% last time), 3 co-founders on avg (same as last time).

• Average age at founding was 35, one year older vs. 2013. 20 somethings and college dropouts are still outliers.

• ~70% of founders worked previously in tech, very similar to last time. 

• ~65% of founders went to school or worked together (a decrease from 90%.

• 67% of teams have a co-founder with founder experience of some kind (vs. 80%). Prior pursuits ranged from founding a small tutoring biz, to Jet.com and Twitch, to a fallen unicorn like WeWork. 

• BUT. Co-founder school & work backgrounds are much more diverse now.

• Only ~20% of founders went to a “top 10 school” (as defined by US News & World Report), vs. two-thirds in 2013, a big change.

• No school has >5% “market share”. Stanford still leads as an alma mater, but for just 5% of co-founders vs. 33% in ‘13 - also a big change.

• About 40% of co-founders are ‘non-technical.’ 60% majored in a STEAM field; ~25% business, 15% humanities, a big change. In 2013, 90% of CEOs had technical degrees. This bucks the “founders must be technical” stereotype.

• And according to People Data Labs, just a third of founders previously held software engineering roles.

• That said, a surprisingly high 20% of co-founders previously worked at a superunicorn - likely tied to 70% of co-founders having worked previously in tech. Google leads as breeding ground for ~6% of co-founders.⁹ 

‍

9. Diversity wanted - LOTS of opportunity to add diversity to founding teams

• As unicorns proliferated, co-founder geographies, educational and work backgrounds did too. This is exciting news for future founders.

• There’s been slow improvement in co-founder gender. 14% now have a female co-founder (vs. 5%), 5% have a female founding CEO (0% in 2013!). 

• But, these numbers are still quite pathetic. There are more founders named Michael, David and Andrew than there are women unicorn CEOs. At this rate, we won’t reach equal gender representation until 2063.

‍

• It’s challenging to track other aspects of identity like race, orientation, disability. Myriad studies show diverse teams deliver better results, including in downturns - so improving diversity in tougher times seems a no brainer.

• Case in point: the elite public unicorn club has higher gender diversity at the top - 14% female CEOs (2) and 21% female co-founders (3).

‍

10. If the past is prologue, expect change ahead for this unicorn herd and a very different, much larger list in 2033

Numbers, sectors, founder backgrounds changed so much this past decade.

‍‍

Our original 39 have had mixed fates: 

• About half are currently public companies and ~80% of them are worth more today. 

• Network effect companies (love network effects!) also grew stronger  (e.g. Uber+Airbnb+LinkedIn = value of whole orig list).

• Enterprise companies fared more reliably, 6x in value since 2013. All but one (Rocket Fuel) maintained unicorn status through a second decade.

• Some consumer companies had bumpier journeys. 33% are smaller today (e.g. Gilt, Lending Club, Yelp); some have had fire sales or shut down (Tumblr, Zulily).

‍

VC firm changes will continue to have a big impact

• In 2013, there were ~850 active venture funds. Today there are ~2,500. 

• As unicorns fall out of the dataset, there will be more change at VC firms. Some VCs will retire, some will raise smaller funds (potentially right-sized for better returns) and/or downsize teams, some firms will fold. We’ve seen “VC musical chairs” in previous cycles, and it’s happening again.

• As we noted in our original post, multi-$B VC funds need multi-$B outcomes to deliver acceptable returns. As funds got bigger, their incentives changed, which drove round sizes and valuations higher, with some negative consequences. We don’t expect VC firms to go back to 2013 in staff or AUM. But lessons here should give founders, VCs and LPs lots of caution about too much money, too early and too fast - and from whom.

‍

That said, as noted in our  Software Unicorn Power Law section - technology continues to change our world.

• Founders have access to more inspiration, capital and talent than ever. And unicorns are relatively few and small vs the number of much larger, slower moving companies with outdated technology.

• We see miles of green fields for more unicorns in years to come.

‍

In closing, what does this all mean?

• The march of technology created so many more unicorns, serving more sectors of society, in just ten years. One can’t help but reflect on the enormous societal and financial impact of VC-backed tech. 

• It feels akin to living through a new Industrial Revolution, powered by software. We’re excited for more history to unfold.

• Less than 10% of today’s Fortune 500 companies are considered tech companies. In coming decades, we expect the share to be much bigger - and many to be from the 2023 herd.

• We also learned macroeconomic factors and capital efficiency matter. The massive influx in private capital funded lots of companies. But capital efficiency declined, which will erode financial outcomes, given how many are still ‘papercorns.’ 

• The cycle is not over. The venture ecosystem will feel the impact, and benefit from these lessons, in years to come as more layoffs, down rounds and shutdowns impact founders, employees and investors. 

• Valuation is a convenient but imperfect, impermanent measure of success. Becoming a unicorn too early might even be a curse. Future founders can see how chasing valuation without a focus on underlying fundamentals can lead to unintended poor outcomes.

• There were times in the decade when many believed building a unicorn was easy and common. Elder unicorns know - it’s not. It requires the sustained magic of product and velocity, customer love, business model economics, capital efficiency, relentless execution and more, over many years.

‍

We tip our hats once again to the hundreds of companies (no longer tens!) who achieved and maintained unicorn milestones. Building and sustaining $1B+ in value in the past ten years was statistically improbable and a major, special team effort.

‍

Finally - if you made it this far, thank you for taking the time to dive into our labor of love. We hope you enjoyed it. We’d love comments on what you appreciated most, disagree with, what we missed, or what you’d like more of. For more like this (although we promise other posts are shorter), give us a follow on LinkedIn, Medium, and X (@CowboyVC, @aileenlee, @allegra_v2).

‍

Notes

_{Cowboy Ventures is a seed-focused VC firm backing teams building customer-loved tech companies that scale to sustained success. We're based in Palo Alto, CA, and invest across the US. Big thank yous to Cow-lleagues Amanda Robson, Jillian Williams, Matt Lu, and Ted Wang for help with this post, and to Crunchbase, Pitchbook, People Data Labs, FlexIndex, and Hiive who provided so much of the underlying data.}

_{1 .Our original analysis was based on a hand-curated dataset of US based, VC-backed companies founded between 2003-2013 that grew to be valued at $1B+ in private or public markets. To be explicit - we did not include companies based outside the US, biotech, medical device-focused companies, or companies more than 10 years old.} 

_{2. We define capital efficiency as current valuation divided by private capital raised.}

_{3. Like in our original analysis, our 2023 dataset is based on US based, VC-backed tech (sw, hw, internet) companies most recently valued at $1B+ in public or private markets and founded in 2013 or later. We use PitchBook, Crunchbase, People Data Labs, and news articles as sources but private market data is challenging. If you spot something inaccurate or we missed, please let us know.}

_{4. Thank you to Hiive, a secondary marketplace for private stock, for data on orders for 290 current unicorns.} 

_{5. This chart illustrates all VC-backed companies valued at more than $1 billion for the first time in a given year, including those founded before 2013, which we do not include in our analysis.}

_{6. The vast majority are papercorns, so this list likely does not reflect realized capital efficiency. Cruise’s efficiency is based on their $1B acquisition by GM in 2016 after being founded in 2014. They have raised $Bs more since as a GM subsidiary, which paints current efficiency very differently.}

_{7. Coinbase was founded in 2012, so not included in this dataset.}

_{8. We tracked data for ~1,300 founders using publicly available information and People Data Labs.}

_{9. We found 87 with work experience at Google = ~6%.} 

‍

‍

‍

‍

‍

‍

‍

‍

‍

‍