Jeff Parker

Data Science & Design

[Jeff Parker @JeffreyP99]

Anthropic locking in that much compute spending is wild, those commitments are insane 😭

[Pinegap AI @PineGapHQ]

Anthropic IPO Deep Dive: Five Critical Takeaways From the S-1

1. Hypergrowth is now profitable. Anthropic's revenue surged from $1.2B in 1H25 to $16.3B in 1H26. Q2 revenue hit $11.6B, marking its first GAAP operating profit of $974M (8% margin; 19% adjusted).

Net revenue retention reached ~900%, while customers generating $100K+ annually jumped from 950+ to 11,900+ in just one year.

2. The massive net loss doesn't tell the full story. The reported $80.7B loss in 1H26 was largely due to non-cash accounting adjustments on convertible notes and partner financing as the company's valuation increased.

The underlying business generated $2.7B in adjusted net profit and $3.9B in operating cash flow.

3. AI computing costs are staggering. Major compute agreements with Google, Amazon, Microsoft, Broadcom, SpaceX's AI unit, Nscale and Lambda total roughly $637B by our estimates. That excludes AMD and another estimated $125B–$150B in TPU capacity. Many contracts run through 2033–2036 with fixed payment commitments.

The company has $131.7B in cash and investments, a $15B credit line and $77B in unused partner financing. Even management warns that margins could come under pressure as more computing capacity is added.

4. Revenue growth comes with risks. Nearly 48% of revenue comes through cloud marketplaces, while the five largest customers account for around 25%.

An estimated 5–7% of revenue in the first nine months of 2026 came from customers later found to have broken usage rules.

More than 80% of Q3 revenue came from models launched in the previous six months, showing how much growth depends on constantly releasing better AI models.

5. Public investors will have limited control. A Long-Term Benefit Trust chooses most of the board, while a founders' entity controls 50.1% of voting power, regardless of economic ownership.

Shareholder dilution is another concern, with $26.6B in future RSU expenses, Amazon notes convertible into 1.84B shares, partner financing that can be settled in stock, and additional share issuances tied to public benefit commitments.

Want the full detailed memo on Anthropic IPO?

DM us or comment your email below, and we'll send it over.

[Jeff Parker @JeffreyP99]

Give Me Labels Under Icons Over Having To Tap Around And Guess What Everything Does

[𝓐𝓺𝓮𝓮𝓵𝓪𝓱🧚🏼 @aqee_sofine]

Modern dating is just women collecting stories for their group chats…

[Jeff Parker @JeffreyP99]

Nah, meet somewhere public first. If he can't respect that, that's an easy block

[Genevieve 🌼✨ @beaumont_valois]

WE’VE BEEN TEXTING FOR LESS THAN AN HOUR, AND HE ALREADY WANTS ME AT HIS HOUSE. 🚩

When I suggest meeting somewhere public first, he says location doesn’t matter if my intentions are genuine. SIR, I DON’T EVEN KNOW YOU! 🤷🏻‍♀️

Is he being reasonable, or is this EXACTLY why you don’t ignore red flags when dating? 👀🔥

[Jeff Parker @JeffreyP99]

@habibk79 I'd wanna see who's actually paying for this stuff

[Jeff Parker @JeffreyP99]

Loneliness will have you thinking about texting somebody you know you need to leave alone

[MP | Masculine Peak 🧠 @MasculinePeak]

No matter how lonely and desperate you feel, NEVER go back to an ex-girlfriend.

She's gone, forever.

[Jeff Parker @JeffreyP99]

Yall think texting every day when you first start dating somebody is too much?? 🤔

[Jeff Parker @JeffreyP99]

@Tojizeninhc Q3 earnings gonna be interesting with all that crypto activity

[Jeff Parker @JeffreyP99]

No dating apps but this pic definitely about to get some attention

[Fitmompaula @fitmompaula]

Minding my own business and getting my protein and workouts in. No dating apps, no drama

[Jeff Parker @JeffreyP99]

Gmail really had to switch everything up right before election day

[Doug Landry @douglandry]

i need the makers of the gmail app to understand that i was not looking for a redesign 9 days out from election day

[Jeff Parker @JeffreyP99]

A few thousand impressions and zero sales is probably a positioning problem

[David @David_cbolt]

My product launch got zero sales.

I launched last week, a few thousand impressions on @X but no sales.

The gut reaction could be to lower the price or create a free tier.

I'm doing the opposite.

On Monday the price doubles. From $6/mo to $12/mo.

Why?

Because I believe in the product.

It started as an dashboard that you connect all your MCPs/APIs to so all of your stats are in one place.

It's now a truly an AGENTIC WORKSPACE, a place for you and your agents to collaborate.

Here's what I run through it.

- Product plans/scopes
- Client Pipeline
- Social posts
- Agent visibilty (current sessions, usage limits etc)
- All my stats, analytics and search console data
- My calendar

My agent sees it all. Not buried in .md files, in a workspace they can interact with.

[Jeff Parker @JeffreyP99]

These Are Clutch When You Need To See How Real Apps Handle A Flow 🙌

[Abraham John 🦄🦓 @Abmankendrick]

UI/UX Designers, here are my go-to best mobile design inspiration sites on the internet that you should bookmark:

Mobbin → https://t.co/YxsQM2IoaD
Refero → https://t.co/2SW5aLZh4q
Appshots → https://t.co/lsHtKQIm5A
60fps → https://t.co/N4N9EUHc5B
UI Sources → https://t.co/eaqsvuqPGQ
Chamjo → https://t.co/kqA5s4V8RY
Page Flow → https://t.co/LA3laDjZoZ
Appllama → https://t.co/COYNoIlIq6
Asoinspo → https://t.co/cPXF1D3vv9
Collect UI → https://t.co/dPgg9O1qFB
Spottedinprod → https://t.co/jOi3B5C5U2
Pttrns → https://t.co/IoKaYLMkop

[Jeff Parker @JeffreyP99]

I'd rather hear we're done straight up than sit there guessing why the texts stopped

[Jeff Parker @JeffreyP99]

That man built different I'd be spending the whole day trying not to stare

[あさひ @MuscleEcstasy]

Sun, sea, and ASAHI. ☀️🌊

A calm smile,
a strong body,
and the kind of presence that makes it hard to look away.

Some days just feel better at the beach.

Would you spend the day with him?

[Jeff Parker @JeffreyP99]

Getting strangers to spend money on something new is tough as hell, that's a big milestone

[Andrew Gazdecki @agazdecki]

Getting your first 100 customers is a pretty incredible accomplishment for any startup.

[Jeff Parker @JeffreyP99]

This Looks Clean As Hell, Those Cards Make The Whole Page Feel Way More Polished

[Basit A. Khan @basit_designs]

Sleek UI cards & subtle animation can lift the experience, can't they?

[Jeff Parker @JeffreyP99]

That Locked Copy Step Is Easy To Forget, Then The Monthly Comparison Gets Messy Fast 🤦🏾‍♂

[Josh (Your CFO Guy) @YourCFOGuy]

From budget to actual to forecast, here are the 5 steps that connect them.

If you skip saving a locked copy of your budget, your actuals end up being compared against a forecast that moves every month.
So let's start with how the budget gets built.

Step 1, build the first pass.
When you're the one in the spreadsheet, it's easy to think you have to put the whole thing together yourself.
But most of the time you don't have the details on every contract and every customer.
So build it WITH the people who run the business, like your CEO and your department heads.

And I'm a big believer that every budget should be built bottoms up.
That way your inputs show exactly how you get to your outputs.
Sometimes I'll work with a startup that tells me their market is worth $30 billion.
And if they capture just 1% of it... well, that's their budget.
But that tells you nothing about how the business gets there.

Step 2, get sign off.
Once your first pass is ready, it needs a sign off.
That could be management, the board of directors or anyone else in the company who has to approve it.

Step 3, save down a separate version.
This is that locked copy I mentioned up top.
You save a separate copy of the file and THAT copy becomes your budget.
Lock it and don't touch it.
The original keeps going as your rolling forecast, the one you'll update every month.

Step 4, actualize the rolling forecast.
Each month you roll that forecast forward with a new set of actuals from your accounting software.

Step 5, compare those actuals against the budget.
Now you put the actuals in your rolling forecast right next to the budget... the numbers that got signed off.
If a variance is small, you most likely don't need to dig into every account behind it.
But if it's big, you drill down and see exactly which accounts make it up.

So steps 1 to 3 happen once for each budget.
Then steps 4 and 5 repeat EVERY month.

Now hold your own process up against these 5... is one of them missing?

[Jeff Parker @JeffreyP99]

Man Getting A Client Just From Somebody Liking Your Writing Is A Pretty Solid Start

[Graig @marvel24926]

I started my side hustle in 2017 by accident.

I tried to start a business in 2015/2016, and it failed. But I had written some content marketing for it that got a lot of views. So one day at an event in 2017 someone cornered me and said “I love your writing, can I pay you to write for me?”

I had just gotten a job at a tech startup and was not looking to start a business at the time, but I wasn’t going to turn down the market literally asking me to start a business. So I made it a side-hustle (I also made it 100% remote because I had a job and couldn’t go to a client office).

From there, I started introducing myself as a freelance writer when I met people. When they asked what kind of writing I did, I asked what they had budget for. It was a bit of a joke, but helped me get a couple clients.

That’s how I got started — I grew the business and eventually took it full-time in 2019, which is the business I run now.

[Jeff Parker @JeffreyP99]

Double texting somebody you like is fine man

[Gabriel ~ Silver.dev @Conanbatt]

Coding agents can build software faster than ever. But building something people actually want is a different problem.

Finding product-market fit still takes real users, market feedback, and time.

Talked about this with @kentcdodds on my podcast.

More on this: https://t.co/mkgsHkIREV