20VC Newsletter - May 12th 2024
Here are the transcripts and top takeaways from 20VC episodes released this last week!
Monday’s episode with Sarah Tavel, General Partner at Benchmark:
Here is the full transcript:
My 7 key takeaways:
Will Foundation Models Commoditise?
There is a tremendous amount of investment to progress these models.
Each successive model will get more & more expensive to train.
This suggests a world where we will get an oligopoly
Frontier Models Will Be Closed Source
The frontier models today are all closed-sourced.
What Meta is doing with LLaMA may change the game.
Open-sourcing isn’t rational in the short term but will be beneficial in the long run.
Why It Is Rational to Pay More for AI Startups on Valuation
Selling work as an AI startup could have a 50x market vs selling software.
You're selling against the cost of the headcount as opposed to a productivity improvement for that headcount.
This makes a higher entry price more rational.
Value Is in the Application Layer Not the Infra Layer
The application layer will be where most of the value will get captured and created.
You have to imagine who owns the user over time.
Whoever owns the end user will be able to provide more value.
What I Know Now That I Wish I had Known: The Importance of Why Now In Venture
A great “why now” story becomes a strong current that pushes the company forward.
This builds crazy resilience & great founders will maximize that opportunity.
Without one, it feels like you’re paddling a boat really hard with no progress.
Why Benchmark Does Not Do Reserves
We are very oriented towards our initial investment.
But once we invest, we are 100% aligned with the founders.
We don’t believe in pro-rata.
The Best VCs are Recruiters and the Best Firms are Headhunters
Executive recruiters are great at what they do because they do it a lot.
I’ve recruited & closed so many candidates it’s now a superpower for me.
We aspire to be the best recruiting firm compared to all others.
Wednesday’s episode with Tom Hulme, Managing Partner & Head of Europe at Google Ventures:
Here is the full transcript:
My 9 key takeaways:
Three Types of Investors
Those who are smart & add value.
Those who are passive & stay out the way.
Those who are passive, think they’re smart & try to interfere.
You NEED to avoid #3.
Biggest Advice to Angels Starting Out
People think they need to write $200k checks to become an angel investor.
The amount you invest is almost irrelevant.
Your job is to learn & demonstrate you add value.
Writing small, $5k checks is the best way to learn.
What Questions Should Every Investor Ask?
Unfair advantage: What is their unique insight & why are they uniquely placed to solve it?
Timing: Why now? Why do they need this difficult thing to happen now?
Concerns: What might go wrong? What’s keeping them up at night?
Foundation Models are Like Power Stations
The training time is the power station & the inference is the power.
Everyone is building similar power stations with little edge.
They have to depreciate those assets in foundational models over a few months & I can’t see it happening.
We have not invested in foundational models.
The Three Core Pillars of Venture
Sourcing deals
Selecting deals
Supporting deals
VCs commonly move into more supporting & less sourcing through their career.
The best VCs are good at all of those.
Being a VC is Being a Founder on Anti-Depressants
You have all the highs (just not as high) & all the lows (just not as low)
You are feeling the ups and downs all the time & every day is a roller coaster.
Roelof Botha: “Great board members are like shock absorbers. They reduce the highs and they will cushion the blows and let founders know.”
Why Complete Conviction is BS
I can’t understand how anyone can have complete conviction on any investment.
I was skeptical of physics concepts in uni even with solid proof.
I can describe a bear & bull case for every one of my portfolio.
Why You Have to Have a Fund Returner Advice is BS
A lot of VC strategy is a lagging indicator of what worked in the past.
There’s a lot more ways to be successful at the growth stage.
Have a strategy & stick to it.
Why I Do Not Trust Ambitious Young People Who Insist on Remote
Being in office is a ridiculously important learning opportunity.
You can’t shadow & learn from colleagues on a Zoom call.
I can’t understand why people are more efficient working from home.
Friday’s episode with Larry Shurtz, Chief Sales Officer at Genesys:
Download the full transcript:
My 6 key takeaways:
Why Verticalization Requires More Than an AE Industry Nerd
You can’t truly verticalize & deliver value to customers with only an AE.
There cannot be a disconnect between them and the technical team.
Every single person that touches the account has to be involved.
Those That Say They Do Not Have CS Is BS
A lot of companies say they don’t need CS.
Someone is playing that role under a different title.
They are responsible for correcting architecture & aligning use cases.
What it Means If You Cannot Forecast Well
You don’t truly understand your business.
You’re not doing the right amount of diligence with customers & opportunities.
We might not be asking the right questions & talking to the right people.
Why I Hate Panel Interviews When Hiring.
I prefer one-on-one interviews.
The relationship between the leader and employee matters
It’s hard to get that out of a 5-person panel.
Why I Don’t Do Case Studies
The real value is to understand their mindset & thought process.
It gets to the what & the how.
I could easily get that out of a 45 minute chat.
One Question I Ask Every Interview
I would ask which culture did you prefer out of all their previous companies.
The range of answers I’ve gotten have been shocking.
Understanding their culture is important to value alignment.
Let us know what your big takeaways from this week’s shows were in the comments below!
Thank you for reading, and don’t miss the great guests we have next week:
Tom Blomfield, GP at YC.
Dan Siroker, CEO & Co-Founder at Limitless.