Getting comp right at @SymphonyAI means bringing data together from a lot of different places.
The more data you can layer into @PaveComp, the more the Pave Agent has to work with. Thank you to Leanne for sharing how SymphonyAI does it, both here and on stage at TRL.
Matt Schulman retweeted
Update on the Ramp equity grants for early hires. Got an anonymous tip that the first product designer at Ramp (within first 5 hires), received 0.25% over 4 years for their initial grant.
That’s ~24% under the low end I originally modeled for hire #5, so my estimates ran high. That 0.25% would be worth ~$28.5M at Ramp today.
I reached out to @Matthewschulman and @rithikaneti at @pavecomp, and they pulled their data on equity grants across US VC-backed tech companies (chart below).
Among the first 5 employees, 0.25% sits above the median (0.16%) but below the 75th percentile (0.42%). For a product designer in the first 5, it lines up almost exactly with the 75th percentile.
And more importantly, benchmarks for “founding” roles are a bit more of an art rather than a science this early in hiring, since grants can be ultra-specific to individuals and their worth to the organization.
It can change with role, seniority, stage, growth etc, so I wouldn’t treat it as prescriptive.
It’s also just a bet on the company from the candidate’s perspective, and sometimes employers will need to flex to close a stellar candidate.
Hope my post helped shed some light on what outcomes can look like at a high growth startup, post-dilution.
Some more exciting updates coming soon between @pavecomp + @Levelsfyi. This post sparked some awesome discussion, and there’s a lot of neat data between the two of our platforms 📊
Matt Schulman retweeted
Update. Got an anonymous tip that the first product designer at Ramp (within first 5 hires), received 0.25% over 4 years for their initial grant.
That’s ~24% under the low end I modeled for hire #5, so my estimates ran high. That 0.25% would be worth ~$28.5M at Ramp today.
I reached out to @Matthewschulman and @rithikaneti at @pavecomp, and they pulled their data on equity grants across US VC-backed tech companies (chart below).
Among the first 5 employees, 0.25% sits above the median (0.16%) but below the 75th percentile (0.42%). For a product designer in the first 5, it lines up almost exactly with the 75th percentile.
And more importantly, benchmarks for “founding” roles are a bit more of an art rather than a science this early in hiring, since grants can be ultra-specific to individuals and their worth to the organization.
It can change with role, seniority, stage, growth etc, so I wouldn’t treat it as prescriptive.
It’s also just a bet on the company from the candidate’s perspective, and sometimes employers will need to flex to close a stellar candidate.
Hope my post helped shed some light on what outcomes can look like at a high growth startup, post-dilution.
Some more exciting updates coming soon between @pavecomp + @Levelsfyi. This post sparked some awesome discussion, and there’s a lot of neat data between the two of our platforms 📊
Today on generational wealth. Ramp's first 10 employees may well be sitting on $1B + in stock. Employee #1's initial stake alone could be worth over $300m
In 2019, Ramp was at $0.21 / share. Today it raised at $160 / share.
Startups typically give their first 10 hires ~5% of the company. @sama says it should be 10% minimum, and Ramp is the type of company to be on the generous side.
Especially given the talent density at the company. 1/3rd of its first 60 hires were ex-founders. By 2021, @tryramp had set aside ~1/4th of the company for employees. Most startups set aside closer to 15%.
If the first 10 got ~10% between them and never sold (via secondary / tender), their original grants alone would be worth over $1B now.
Hire #1, ~$340M
Hire #10, ~$41M
That's post-dilution (~80%). Share price still up ~760x from 2019. Not yet fully liquid ofc.
Once again, only from original grants (total could be higher)
We don’t know the true grant values, so these are estimates. Chart and assumptions below.
Matt Schulman retweeted
Anthropic just published its own workforce data showing the decline in junior share across its technical teams
At 10K+ companies in @pavecomp data, this trend is consistent
If entry-level roles keep shrinking, who’s being trained to replace today’s senior talent?
How do you win the top AI talent at the frontier?
A LOT more equity compensation.
AI Hypergrowth companies in Pave’s dataset are paying an average of 2.4x more New Hire Equity than the rest of the market.
And yes, these AI Hypergrowers are also paying about 21% more on average in base salary.
How does this compare with what you are seeing in the market?
Do Forbes AI 50 companies like OpenAI and Cursor have more managers or more ICs?
The answer is clear: more ICs.
The average AI Hypergrowth company in Pave’s dataset has an average of 83% ICs versus 75% at the “rest of tech.
Similarly, the average AI Hypergrower in this analysis has an average of only 3% execs vs 8% at the “rest of tech”.
Do you expect this “bottom-heavy” org chart trend top spread to the rest of the market?
Methodology note: AI Hypergrowth companies include 80% of the Forbes AI 50 in @PaveComp’s dataset including OpenAI, Cursor, Cognition, Databricks, Perplexity, Harvey, Legora, Decagon, Sierra, Fin, and 30 more.
Does adding “engineer” to your title give you staying power in the new AI labor market?
Naturally, it depends on the department.
But for some roles like Customer Support Engineer, there is certainly some material optimism.
What other new “engineers” are seeing pop up on your org chart?
Join Pave Data Lab for more insights. pave.com/pavedatalab?utm_sou…
What is the ideal ratio of SWEs per PM in this new world of agentic coding?
For decades, Jeff Bezos has preached a “pizza team” role that limits meetings to ~8 people. At Amazon, this meant that a PM typically was responsible for up to ~8 SWEs.
But if we look at the data, the top AI Hypergrowth companies in the Forbes AI 50 are now operating with ~2x historical averages. For instance, companies in the Forbest AI 50 with more than 1,000 employees now have an average of ~11-17 SWEs per PM.
What are your predictions?
Methodology note: AI Hypergrowth companies include 80% of the Forbes AI 50 in @PaveComp’s dataset including OpenAI, Cursor, Cognition, Databricks, Perplexity, Harvey, Legora, Decagon, Sierra, Fin, and 30 more.