YC W23: Trends, Thoughts, Investments
YC’s demo day is one of the most critical events for pre-seed and seed investors worldwide — even for those who don’t want to invest in the current batch — because it acts as a thermometer. By following the demo day (especially, I should say, the pre-demo day), you get a sense of all the trends that are taking shape: the hot sectors, the size of the rounds, the valuation levels, who is active, who is hustling, who is really doing the work.
In this post, I will crunch some numbers, analyze trends, and comment on what happened.
I love writing this type of post!
During the Winter 2023 demo day (W23), 264 startups took the stage, following the traditional format of presenting for 1 minute with 1 slide each.
The event was held online on April 5 and 6.
The W23 batch lasted from January to early April 2023, and each startup received $500K in investment, divided into $125K at fixed terms (7%) + $375K on a SAFE MFN, totaling $132M in invested capital for Y Combinator.
W23 By the Numbers
2,983 investors attended.
264 startups pitched (out of 20,000 applications, only 1.32% were selected).
🧐 actually, the startups admitted to the batch were 282, so the percentage selected would be 1.41%. However, for the sake of simplicity, I’ll consider the number of those who presented on the demo day.
8% of the founders are women (down from 9% in S22, 10% in W22, and 12% in S21, still laughable numbers 😔), while 17% of the companies have a woman founder 👎
195 SaaS (73.8%, higher than ever!), 112 Machine Learning/AI (obvious!), 86 Developer Tools (more than doubled since last batch), 53 Fintech, 29 Consumer (cut in half since last batch).
238 (90.1%, up a whopping 11 % points since the last batch) with a focus on USA/World (always ironic that from YC’s point of view, the two geographies are overlapping 😂), which is the geography we are interested in.
31,000+ intro emails have been sent between investors and startups (at the time of writing, but this is not the final number because the demo day platform will close in a month, I assume it will grow by a good 10% at least). The YC investor dashboard allows you to ‘Like’ a startup, and automatically, an intro email is sent between the investor and founder.
On average, each investor in attendance expressed interest in 10 startups, and each startup received 117 intro emails.
👉 Note: this average is wildly distorted due to startups that have skyrocketed in popularity and others that have received fewer intros. Ideally, the median could provide a better understanding, but this is not possible since we only have aggregated data 🤷♀️.
As we have been collecting data for several years now, here is a comparison of the numbers from various batches to identify trends:
Some additional data about the batch was shared in the latest blog post by YC: Meet the YC Winter 2023 Batch.
Below is a table comparing percentages from previous batches to identify trends (Y Combinator has only been sharing this data for 3 demo days, so historical data is not very relevant, but we are starting to monitor it):
Meetings and Investments Made by Lombardstreet Ventures
This batch holds the record for most meetings, most of which occur before demo day.
At Lombardstreet Ventures, we selected 78 startups from this batch and contacted them to meet. Like last year, we started reaching out well ahead of demo day. 73 out of the 78 startups we connected with were contacted before demo day, some even several months in advance (we sent the first email in November).
Out of these 78 companies:
- 64 were met with;
- 5 were already oversubscribed: they quickly raised the capital they needed and stopped fundraising before YC;
- 7 never got back to us 🤷♀️;
- 3 were not actively raising capital, so we postponed the meeting until they did.
Of those we met with to explore further:
- 🎉 3 startups entered our portfolio, all of which we contacted pre-demo day. Welcome to Texel, Rubbrband, and Workpage!
- 😩 3 startups raised capital so quickly that they filled the entire round between the time we met them and when we proposed the investment (sometimes within a few hours);
- ✖️ we decided to pass on 58 deals after thorough due diligence.
The table below compares the numbers for Lombardstreet Ventures across various YC batches to analyze some trends:
We can observe the following trends:
The most notable change in our strategy is that we contacted a more significant number of startups (78!) and started much earlier (from November 2022).
Why did we contact more startups?
- First, we “widened the net” during the initial selection process, deciding to meet with more startups and then tightening the filter when choosing which ones to invest in
- Second, in this batch, many startups were closely aligned with our main investment interests (such as dev tools, open source, APIs, etc.), which caught our attention more than usual.
Most startups we contacted were pre-demo day (73 out of 78), making post-demo day meetings a slim minority.
- The number of meetings conducted has significantly increased, with 64 meetings in batch W23 compared to 27 in batch W21
- Consequently, the number of investments we didn’t pursue has also increased. It’s always regretful to pass on investments, and I really empathize each time with the disappointment on the other side of the email.
- The number of our investments remains constant at 3, despite missing out on another 3 rounds that we would have liked to enter — they got oversubscribed quickly even if we get to a decision within 48 hours from the first meeting.
Finding a solution to this challenge remains an open question for us.
Highlights and Observations from Demo Day YC W23 🤓
1) E v e r y t h i n g Happens Pre-Demo Day 🕵️♀️
During previous batches, things were happening both before and after demo day and now, without a shadow of a doubt, we can say that the real action happens pre-demo day.
Many startups arrive at demo day with their round full or their valuation increases by $5–10M before a new raise.
Starting in November, every early-stage VC started delving into their network and the Internet to identify which startups were selected for this batch and to contact and evaluate them as quickly as possible. It’s a race against time, an actual intelligence operation, as the list of startups is not made public until demo day, and as it approaches, rounds fill up quickly.
2) Valuations Didn’t Really Lower, Despite the Market 🚀
We thought (and hoped!) valuations would drop a bit more.
However, we have not seen any company raise below $13M post (all with SAFEs, of course), even though many had pivoted during YC and had actually started working on their idea only a couple of weeks earlier.
Most startups were raising at valuations from $15M to $20M, with some at $25M or $30M.
The standard raise was $1.5M at a $15M post-money valuation.
So, once again, the general trend of pre-seed and seed rounds appearing unaffected by the market situation is confirmed.
In short, valuations look precisely the same as in the last two demo days.
3) Focusing on the Earliest Stage
YC’s focus is increasingly on the pre-seed stage, as evident in the type of selections they make:
77% with 0 revenue, 52% with just an idea
On top of that, they took a strategic decision: shut down the growth fund and create the co-founder matching platform.
The perception is that YC aims to become the world’s leading point of reference for pre-seed investments.
The reason for this shift is apparent:
+$500M funds now dominate the subsequent stages, and the only spot left relatively unstructured is the very early stage.
BTW, that’s great for us because it’s precisely where we want to be.
4) SF is Back! 🌉
As Garry Tan said during the demo day: a staggering 86% of the batch's companies were present in San Francisco for the duration of the batch.
This batch focused more on the in-person side of things, with weekly meetings and a 3-day conference in Sonoma to kick it off.
Let’s see what the future brings, but it looks suitable for the supporters of IRL:
But let’s not only focus on the past 3 months.
Where are the companies based? Of the 238 companies focused on the US market, 158 are based in the Bay Area (with the majority, 140, in San Francisco), 49 in New York, 10 in Los Angeles, and 11 in Seattle (more than doubling since the last batch).
Many companies are fully or partially remote, while some are focused on the US market but not physically present here (Tel Aviv, Toronto, and Zurich are the main “hot spots” outside the US).
In the table below, I have compared the locations for the last 2 batches to see the trend.
Disclaimer to best read these numbers: I calculated them by hand from the demo day portal they may have inaccuracies. Also, many startups have more than one location listed.
5) Team Composition
There were very few solo founders, with most teams comprising 2 or 3 members. We did not come across any teams with 4 members.
We saw less diversity than usual, with women founders at an all-time low of 8%. We only met 3 teams with exclusively women founders, which I understand can be harder to find, but it is discouraging.
This is how PG explains why there are so few startups founded by women-only teams:
I read on the Internet (so it must be true) that only 1.7% of VC-backed startups are founded by women. The percentage of female hackers is small, but not that small. So why the discrepancy?
A possible explanation emerges when you realize that successful startups tend to have multiple founders who were already friends. People’s best friends are likely to be of the same sex, and if one group is a minority in some population, pairs of them will be a minority squared.
We saw many ex-Big Tech founders, likely due to the many layoffs we saw in 2022.
The most common structure had initial co-founders/team in the same room (SF/NYC) and then a distributed team to scale up.
6) AI and Pivots
AI is eating the world so fast that startups cannot keep up (which is no news).
While I don’t have exact numbers, based on my personal observations, I can say that at least 30% of the batch pivoted in some way at some point (some even made a 360-degree turn). This can be attributed to the numerous advancements and innovations in AI launched in recent months.
However, it’s important to remember that:
- Pivoting, usually within the same space or general vision, is entirely normal, if not expected, in the pre-seed stage.
- YC invests in teams—and so do we—not just ideas: if these are high-quality teams with expertise in AI, they will find a way to build something meaningful in the space
7) Additional Thoughts
- Many tools were inspired by internal tools built within big companies, developed by teams from those companies, and supported by angel investors from the same companies.
- The explicit goal for everyone was to close the round by demo day.
- Many intros came in from other founders, even if we passed on thier company. This trend became more evident in the post-Demo Day days: maybe YC suggested the companies help each other fundraise because winter is coming.
- Is the “Palantir Pack” the new “Paypal Mafia”? Who knows, but for sure, it’s a good one.
- Few biotech, space, health, climate, food, and ag-tech startups.
- Fewer hardware and deep-tech companies than last batches.
- This was the most attended YC demo day of all time, with 2,983 investors accessing the demo day website (the highest number in the last 3 years, so I’m confident no other demo day has surpassed this count).
- Despite that, W23 did not have the record number of intros requested. The demo day website currently shows around 31,000 intros made, whereas the number of intros in the S21 batch surpassed 50,000.
While it is true that investors may pay a premium to invest in YC-backed startups, the high standard of selection maintained by the Paul Graham and Jessica Livingstone program often makes it worthwhile.
🖖
