Cofounder and CEO of Pathpoint, fmr @hvflabs and PM.
Joined March 2009
- Tweets734
- Following1.7K
- Followers556
- Likes3.5K
This is backwards.
Leaving aside Waymo’s own data, Swiss Re reported 92% less bodily injury claims and 93% less property damage claims from AVs.
AV companies must report every incident to the NHTSA (human drivers do not).
Waymo’s own data shows 94% fewer serious injuries or crashes.
34 people have died so far this year in DC from traffic crashes (up from 21 YTD last year).
And no one rushed this. We are literally entering year 4 of safety studies by the District.
The attached Lead Edge memo from Mitchell Green is iconic. I forget when I first saw it, definitely pre-COVID? I think it's called "The Hierarchy of Bullshit." A little under a year ago, Pathpoint got to the top of that pile...
And I think it's a great framework -- I see founders every day throwing around these vanity metrics (most, these days, are around bullet point 4; in the 2010s a lot never got there, getting stuck at bullet 5-8; fortunately it's been almost two decades since bullets 9-10 ensare people).
From a fundraising perspective, it is absolutely a Hierarchy of Bullshit -- go in with a vanity metric and not much else and you won't get far.
However, from a building distribution perspective, especially when able to finance that build with venture capital...it's actually not a dumb roadmap. So long as you know where you're going. Which is easier said than done.
Point being, when Pathpoint started building a distribution network of clients that wanted an alternative way to place E&S business, the best way to measure it was registrations ("appointments" in our world). It wasn't a measure of a moat or anything real in terms of profit, but it told us if people were opting in and if we were on track to have the scale we needed as a small business wholesaler. As soon as we could get registrations, we focused on keeping them active. Then we focused on aligning that activity to top line sales. Then we ensured that those top line sales were sticky and converted to revenue with efficiency. Then we made sure that that revenue could easily cover not just the salaries of the underwriters that placed the business but all the supporting staff, vendors, compute, and other costs that were necessary. Then we made sure that what was left of that generated operating income (which, in our world, is what's used to value an enterprise).
Oddly enough, for the first time, we're going back a bullet point (with a year of cashflow, it's time to go on offense and invest in bringing in new talent) -- and given the soft market, we need to continue to validate that the "core" business (the business today) remains profitable enough to finance various waves of growth.
Anyway, the point is is many: (1) this memo has stuck with me over the years, and directly informed how I constructed strategy; (2) it's actually a ladder of strategy (building foundations upon foundations) for the right type of businesses if you don't fall victim to vanity.
A nice essay for Sunday reflection, especially as someone coaching one of their kids (girls soccer) for the first time this year wanting to strike the balance of helping them find the excitement and benefits of competitive team sports without boxing them in or making the decisions for them: newyorker.com/magazine/2026/…
Signs the Eng team you work with is AI pilled: one Eng ships dark mode in a day when no one other than the Eng team asked for it and they also shipped a hundred other features this month not disrupting the roadmap at all.
A feature gap that is actually a feature: poor spreadsheet formatting by AI models (as simple as not starting in B2, manual cell highlighting vs. standard banker style formatting) forces me to reformat each manually which ensures I un sloppify / validate it before I distribute.
I might be wrong but I also think the startup salary discount has washed out of the ecosystem for anything funded by top tier funds. Sure if you’re not in the prime time I guess but anything in SF or Valley ecosystem is paying a very competitive wage. Not beating FAANG but not worse than the norm tech benchmark. You can think you’re paying a discount but likely you’re just procuring talent from another pool. And we are pro long exercise windows at PP FWIW but it has some downsides (early employees lasted 2 years have more equity than late employees that will stay for 8).
New founder hack: got your insta algo to think you’re your own ICP. Hoo boy the AI slop and AI software being shilled to insurance brokers and agencies right now is something.
This is interesting. We're seeing a corollary at Pathpoint: we can't tell if in insurance a vertical player (like an insurance version of Legora/Harvey) or the model labs (like Cowork) will be the UI for work in the space. Both scenarios are possible and likely "work" (1 of 2)
In reflecting on the Claude Tag launch, I had a thought: the labs might be freezing parts of the enterprise AI market without shipping a thing because of the Osborne Effect.
The Osborne Effect is where consumers delay buying the current product because they think a better one is coming. Tim Cook blamed it for a "pause in purchases" on Apple's 2017 call after iPhone X rumors leaked. The name comes from Osborne Computer, the fastest-growing computer company in the world in 1981. Execs started teasing the successor, customers stopped buying, the new product wasn't ready. Bankrupt by 1983.
Now look at AI agents. Claude Tag just shipped. Slack has its own. OpenAI and Google are pushing "apps" faster than anyone can track. And countless well-funded startups are chasing the same wedge.
Some enterprise buyers might be cautious of piloting a seed-funded agent that sits in the strike zone of the labs. Not because the startup is worse (I'm sure they are way better) but because their teams have bandwidth for maybe 2 experiments this quarter, and betting on something the labs might ship in 6 months is a tradeoff.
The counter is real: waiting for the platform means lock-in. But integrating 10 point solutions is its own tax, and enterprises hate complexity more than lock-in. See Microsoft Teams eating messaging + video + files. Almost nobody loves Teams. Almost every large company uses Teams.
The answer for startups might be that they need to move from point product to solution faster than ever (e.g., what Ramp and Rippling are doing).
Attention is all you need. But attention from corporate buyers is the scarcest resource in enterprise AI.
But we don't know if we should *wait* to pilot more vertical stuff (we've piloted a lot) while we keep making improvements with cowork (skill libraries, staff training, custom connectors, evolving our own MCP, etc. and see how much better Cowork gets or the vertical players get.
Linear for relationship management is a very smart idea.
We're launching Lightfield today.
It's a CRM designed for founders going zero to one—shaped in the past year by hundreds of founders who took a bet on us and now use it daily, and a waitlist of 20,000 more.
When starting out, it takes countless hours of talking to customers to figure out what works.
From day zero, Lightfield builds and updates itself from your unstructured conversations with customers.
It becomes your customer memory: helping you execute deals, communicate without missing details, and understand patterns across your business.
It's free to try. Would love to hear what you think: lightfield.app
Real time financing estimates, collaboration features, GL and Property coverage for convenience, gas, and liquor stores and hair care services, and we finished converting every product and carrier to our new appetite and underwriting engine all live in the same 2 weeks. Plus another 10+ small UX improvements. The Pathpoint team has set a new pace, they’re an incredible group!
🎇Introducing your new digital team workspace…Agency Collaboration! Work on your agency accounts faster and easier as a team on one user-friendly website. Learn more about the new Agency Collaboration feature before it goes live at bit.ly/3UbRKiP
#insurancenews #team
Same reason most of the actual applied design and product principles from Lean Startup / MVPs are no longer relevant. Most everything needs to be way more features these days. Even more so in tech enabled services.
WSIA Marketplace kicks off on Sunday -- the annual conference for the E&S/wholesale market (which now represents over $120B of premium a year).
We've got more planned -- Pathpointers attending will be relaying live reactions back we'll publish out to share our sense of the pulse of the market (with an emphasis on how people from our own team attending for the first time are managing the day to day). Stay tuned!