Does "annualized" mean the trick where they take the highest month and multiply by 12? It seems to me they'd simply say "annual" if it was the revenue for the year.
Does "annualized" mean the trick where they take the highest month and multiply by 12? It seems to me they'd simply say "annual" if it was the revenue for the year.
Maybe I'm dense, but what even is the point of a metric like that, other than tricking people who think you're talking about annual? I'm having a hard time coming up with any legitimate uses for...
Maybe I'm dense, but what even is the point of a metric like that, other than tricking people who think you're talking about annual?
I'm having a hard time coming up with any legitimate uses for such a thing, other than wild adhoc extrapolations.
I think the idea is that it's a projection of the company's annual revenue. Like if a brand new company started up and earned $10,000 in its first month, you wouldn't say "this company earned...
I think the idea is that it's a projection of the company's annual revenue. Like if a brand new company started up and earned $10,000 in its first month, you wouldn't say "this company earned $10,000 in the last year." Instead you'd say "this company's projected to earn $120,000 per year." The logic holds the same for a company that's been ramping up (or losing) revenue if you want to look at how much it's actually worth right now, without factoring the fact that it was making less (or more) earlier in the year.
In theory it’s not that wild for a rapidly growing company to be judged on recent performance - especially if they’re working with longer B2B contracts, in which case the whole bunch of deals...
In theory it’s not that wild for a rapidly growing company to be judged on recent performance - especially if they’re working with longer B2B contracts, in which case the whole bunch of deals you’ve just signed last month are already agreed to continue for the next 12+ months anyway, and it’d be more misleading to average over the 11 prior months where those revenue streams didn’t exist than to project them forward into the year you already know they will exist.
But in practice, anything remotely reasonable in business has been stretched far past its breaking point, the pieces have been sold off for scrap, and then 10x their actual value has been recorded as a tax loss. The concept can be sound, but I wouldn’t trust the people using it as far as I could throw them.
You wouldn't use it in a situation where you have annual data, but when people use annualized data they can't use annual data because it would require knowledge from the future. It's inheritely an...
You wouldn't use it in a situation where you have annual data, but when people use annualized data they can't use annual data because it would require knowledge from the future.
Yes, that’s how it works. However, we have data points for March, May, August, and “this month,” so it’s not just taking the highest month, and the rate of increase is quite impressive. (I guess...
Yes, that’s how it works. However, we have data points for March, May, August, and “this month,” so it’s not just taking the highest month, and the rate of increase is quite impressive. (I guess it is true that the latest month is always the highest one.)
It’s a private company, so we unfortunately don’t have real financial statements, just unofficial numbers in the news.
From the article: When I last checked at the end of May, they were at about $3 billion. ... ... Also, they just signed a deal with Google to rent up to a million TPU's.
From the article:
Artificial intelligence startup Anthropic is projecting to more than double and potentially nearly triple its annualized revenue run rate next year, fueled by the rapid adoption of its enterprise products, according to two people familiar with the matter.
The company is on track to meet an internal goal of $9 billion in annual revenue run rate - a calculation of annual revenue extrapolated from the current sales pace - by the end of 2025, the people said. For 2026, Anthropic has set even more aggressive targets: a base case of more than doubling to $20 billion in annualized revenue and a best case of as much as $26 billion, the people said, requesting anonymity to discuss private figures.
Anthropic told Reuters its annual revenue run rate is approaching $7 billion this month, but declined to comment on future projections. The company has said its annual revenue run rate was more than $5 billion in August.
When I last checked at the end of May, they were at about $3 billion.
...
Fueling the expansion is the uptake of enterprise products, which are built for organizations. Anthropic has more than 300,000 business and enterprise customers, which account for about 80% of its revenue.
Binge watching old seasons of Dragons Den (OG Canadian Sharks Tank) : revenue is not the same as profit. I'm not convinced. Revenue != Profits Annualized Revenue != Revenue Front Load Sales +...
Binge watching old seasons of Dragons Den (OG Canadian Sharks Tank) : revenue is not the same as profit. I'm not convinced.
Revenue != Profits
Annualized Revenue != Revenue
Front Load Sales + Delay Expense + Circular business
Anthropic "sells" $20 of services in the first quarter. Their annualized revenue becomes $80. The agreement could be that Anthropic spends $80 to buy back capacities or whatever from these same companies later in the year. Everyone has an inflated annualized revenue, no one has actually turned any profit. Is something like this expressly illegal and enforced?
True that revenue isn’t profit. Strong revenue growth is a sign of a popular product, though maybe not if the only thing making it popular is that it’s sold at a loss. Perhaps operating expenses...
True that revenue isn’t profit. Strong revenue growth is a sign of a popular product, though maybe not if the only thing making it popular is that it’s sold at a loss. Perhaps operating expenses will go down with some effort? It’s a private company so we only get glimpses.
Anthropic claims that they have over 300,000 business customers. It seems like that would be a lot of shady deals to try to negotiate?
But it is true that there is worry about circular deals:
Nvidia plans to invest in OpenAI, which is buying cloud computing from Oracle, which is buying chips from Nvidia, which has a stake in CoreWeave, which is providing artificial intelligence infrastructure to OpenAI.
The AI boom that is revolutionizing how people live and work has become increasingly fueled by just a handful of companies turning to one another for the vast amounts of capital and computing power needed to drive their breakneck growth.
…
The latest example of that network of investments came Monday, when OpenAI — the maker of ChatGPT — announced a deal with artificial intelligence chipmaker Advanced Micro Devices, or AMD.
Under the terms of the OpenAI-AMD partnership, OpenAI will purchase AMD’s chips for an undisclosed sum in exchange for the right to take a stake of as much as 10% in the semiconductor giant.
The announcement came just weeks after Nvidia unveiled a deal under which it pledged to invest up to $100 billion in OpenAI.
These particular examples seem like more of a worry for Nvidia, though? They’re funding their customers to buy product from them, in return for stock. With fewer steps, it would be like giving these companies expensive hardware in exchange for stock.
Investing in related companies isn’t illegal if it’s properly accounted for. It’s risky if the investments go bad.
Question: Why in their right mind would anyone ever consider working with Anthropic? Because Anthropics biggest customer was Cursor, that provided an LLM code assistant supported by Claude and it...
Question: Why in their right mind would anyone ever consider working with Anthropic?
Because Anthropics biggest customer was Cursor, that provided an LLM code assistant supported by Claude and it was the AI Success Story. I had people swearing to me that it made them more efficient and it was picking up mistakes and it reduced ticket times and I'm not one to argue with lived experience.
And then Anthropic jacked up their prices to an unsustainable level and released Claude Code.
Now I'm not making any accusations, but an LLM company with possibly one somewhat successful customer conveniently begins offering a comparable service at a lower cost.
What happens if my company starts using Claude and I discover some novel revenue generators. Can't exactly copyright this sort of discovery. And am I going to trust the company that is currently paying out a $1.5bil to authors for copyright infringement. In an industry that is insanely over-leveraged and has a very loose respect for ownership rights.
Also, who makes a decision on"triple annualized revenue". The meme is that IT managers do nothing but theres a bit more to forecasting.
You can go to Microsoft or Adobe or AWS or Google or almost any other vendor and pretty easily quantify a product and feature set against a cost per user per year. Worst case it's an email or two. I need to have every product I'm paying for maped out in my architecture or a workflow with how its interacting and integrating with each other. I can set up skills profiles to streamline new hires and plan training to enable current users to leverage underutilized tools that we have at our disposal. There is often vendor development plans that allows me to consider how the software will be changing and if there are new opportunities or identify areas that are lacking. Service level agreents to insure service quality. Doesn't matter if I'm planning a boardroom deal for a full cloud migration initiative that will completely transform the organization or getting a single 365 license for a workstation, the number that will be in a forcast is based on evidence and is an actual number. Maybe it's different for a mega corporation but you need to work on some facts
I work for an enterprise company that uses Claude models, and the idea is that we use them in a way that we could swap models/providers later on. I imagine most companies are building...
I work for an enterprise company that uses Claude models, and the idea is that we use them in a way that we could swap models/providers later on.
I imagine most companies are building infrastructure like this, model agnostic.
My understanding is that the AI companies and the vendors that depend on them have had trouble with selling monthly subscriptions and heavy users. Apparently some users would buy a subscription...
My understanding is that the AI companies and the vendors that depend on them have had trouble with selling monthly subscriptions and heavy users. Apparently some users would buy a subscription and use it all day, every day, or even buy multiple subscriptions and run them all 24-7 (multilple agent workflow) at a substantial discount to the API price. Either the vendor or the AI lab is losing money on that.
It’s not wrong to take advantage of a loophole / introductory offer, but also not surprising that the loophole was closed after a few months.
Also, I guess people really like the product, though the price was too low.
Does "annualized" mean the trick where they take the highest month and multiply by 12? It seems to me they'd simply say "annual" if it was the revenue for the year.
I don't even know if they take the highest month, but yeah, that is the general idea. Multiply monthly income by 12.
Maybe I'm dense, but what even is the point of a metric like that, other than tricking people who think you're talking about annual?
I'm having a hard time coming up with any legitimate uses for such a thing, other than wild adhoc extrapolations.
I think the idea is that it's a projection of the company's annual revenue. Like if a brand new company started up and earned $10,000 in its first month, you wouldn't say "this company earned $10,000 in the last year." Instead you'd say "this company's projected to earn $120,000 per year." The logic holds the same for a company that's been ramping up (or losing) revenue if you want to look at how much it's actually worth right now, without factoring the fact that it was making less (or more) earlier in the year.
In theory it’s not that wild for a rapidly growing company to be judged on recent performance - especially if they’re working with longer B2B contracts, in which case the whole bunch of deals you’ve just signed last month are already agreed to continue for the next 12+ months anyway, and it’d be more misleading to average over the 11 prior months where those revenue streams didn’t exist than to project them forward into the year you already know they will exist.
But in practice, anything remotely reasonable in business has been stretched far past its breaking point, the pieces have been sold off for scrap, and then 10x their actual value has been recorded as a tax loss. The concept can be sound, but I wouldn’t trust the people using it as far as I could throw them.
You wouldn't use it in a situation where you have annual data, but when people use annualized data they can't use annual data because it would require knowledge from the future.
It's inheritely an extrapolation.
Most recent month (or quarter), not highest.
When revenue is growing, they happen to be the same. That's usually a good sign, though.
Yes, that’s how it works. However, we have data points for March, May, August, and “this month,” so it’s not just taking the highest month, and the rate of increase is quite impressive. (I guess it is true that the latest month is always the highest one.)
It’s a private company, so we unfortunately don’t have real financial statements, just unofficial numbers in the news.
From the article:
When I last checked at the end of May, they were at about $3 billion.
...
...
Also, they just signed a deal with Google to rent up to a million TPU's.
Binge watching old seasons of Dragons Den (OG Canadian Sharks Tank) : revenue is not the same as profit. I'm not convinced.
Revenue != Profits
Annualized Revenue != Revenue
Front Load Sales + Delay Expense + Circular business
Anthropic "sells" $20 of services in the first quarter. Their annualized revenue becomes $80. The agreement could be that Anthropic spends $80 to buy back capacities or whatever from these same companies later in the year. Everyone has an inflated annualized revenue, no one has actually turned any profit. Is something like this expressly illegal and enforced?
True that revenue isn’t profit. Strong revenue growth is a sign of a popular product, though maybe not if the only thing making it popular is that it’s sold at a loss. Perhaps operating expenses will go down with some effort? It’s a private company so we only get glimpses.
Anthropic claims that they have over 300,000 business customers. It seems like that would be a lot of shady deals to try to negotiate?
But it is true that there is worry about circular deals:
The AI boom's reliance on circular deals is raising fears of a bubble
…
These particular examples seem like more of a worry for Nvidia, though? They’re funding their customers to buy product from them, in return for stock. With fewer steps, it would be like giving these companies expensive hardware in exchange for stock.
Investing in related companies isn’t illegal if it’s properly accounted for. It’s risky if the investments go bad.
Question: Why in their right mind would anyone ever consider working with Anthropic?
Because Anthropics biggest customer was Cursor, that provided an LLM code assistant supported by Claude and it was the AI Success Story. I had people swearing to me that it made them more efficient and it was picking up mistakes and it reduced ticket times and I'm not one to argue with lived experience.
And then Anthropic jacked up their prices to an unsustainable level and released Claude Code.
Now I'm not making any accusations, but an LLM company with possibly one somewhat successful customer conveniently begins offering a comparable service at a lower cost.
What happens if my company starts using Claude and I discover some novel revenue generators. Can't exactly copyright this sort of discovery. And am I going to trust the company that is currently paying out a $1.5bil to authors for copyright infringement. In an industry that is insanely over-leveraged and has a very loose respect for ownership rights.
Also, who makes a decision on"triple annualized revenue". The meme is that IT managers do nothing but theres a bit more to forecasting.
You can go to Microsoft or Adobe or AWS or Google or almost any other vendor and pretty easily quantify a product and feature set against a cost per user per year. Worst case it's an email or two. I need to have every product I'm paying for maped out in my architecture or a workflow with how its interacting and integrating with each other. I can set up skills profiles to streamline new hires and plan training to enable current users to leverage underutilized tools that we have at our disposal. There is often vendor development plans that allows me to consider how the software will be changing and if there are new opportunities or identify areas that are lacking. Service level agreents to insure service quality. Doesn't matter if I'm planning a boardroom deal for a full cloud migration initiative that will completely transform the organization or getting a single 365 license for a workstation, the number that will be in a forcast is based on evidence and is an actual number. Maybe it's different for a mega corporation but you need to work on some facts
I work for an enterprise company that uses Claude models, and the idea is that we use them in a way that we could swap models/providers later on.
I imagine most companies are building infrastructure like this, model agnostic.
My understanding is that the AI companies and the vendors that depend on them have had trouble with selling monthly subscriptions and heavy users. Apparently some users would buy a subscription and use it all day, every day, or even buy multiple subscriptions and run them all 24-7 (multilple agent workflow) at a substantial discount to the API price. Either the vendor or the AI lab is losing money on that.
It’s not wrong to take advantage of a loophole / introductory offer, but also not surprising that the loophole was closed after a few months.
Also, I guess people really like the product, though the price was too low.
I also aim to triple my revenue by 2026.
Overachievers over here. I'm just tying to beat inflation (and failing miserably).
Mirror: https://archive.is/EjLeA