News
Entertainment
Science & Technology
Life
Culture & Art
Hobbies
News
Entertainment
Science & Technology
Culture & Art
Hobbies
During the era of big data, data gravity was the core strategic imperative. Wherever the biggest dataset resided, customers ran their compute workloads that generated all of the profit and revenue growth for the last generation of data companies. Today, the battle is for AI gravity. Why? AI requires orders of magnitude more compute than other workloads, so there’s much more money & profit to be made serving customers running them.
As software startups begin to sell agentic systems, the procurement process will change. Unlike classical software, where the application either meets the criteria (price, integration into other software, particular features) or doesn’t, agentic systems operate on a performance continuum. Here’s a recent evaluation table for Codestral, Mistral’s open-source code generation AI. All of these benchmarks are machine-generated : HumanEval & HumanEvalFIM are not human testers - but open-source projects that evaluate AI code.
Large scale ETL (extract, transform, load) processes are a critical part of any data pipeline. They are responsible for moving data from one place to another, transforming it into a usable format, and loading it into a destination system. In the world of blockchain, these processes are even more complex. In web2, the engineering team building a payment processing system will convey to the analytics team the data schema. In web3, any programmer can create transactions & inject meaning into fields.
Machine learning advances tend to evolve in bursts. Researchers publish a new paper with a newly discovered technique. It launches the industry forward & more researchers rapidly iterate to improve it further. Progress looks like this - a series of aS curves one after another. No one knows the time period between the rapid progress or the slope of the curves or how much progress we’ll make during one of these curves.
In a world where AI agents are 2.5-3x as productive as humans, which would parallel mechanical robots, how does a software company price? Building on yesterday’s post, pricing in software companies may change significantly when AI agents become the norm. The SaaS business model of the last 20 years for SaaS is a beautiful one. Annual prepaid contracts are free loans to software companies ; seat-based pricing is a tangible metric for pricing ; as a client grows so does this account, producing good net dollar retention.
Pricing an AI product will be a defining question in software for the next few years. AI products offer productivity gains. But greater productivity may reduce the demand for seats over time, ultimately decreasing the size of software markets. We can observe the market trends today across some of the larger SaaS companies who offer AI pricing. Company Product Base Price AI Price Ratio Github Github Enterprise 21 10 0.67 Gitlab GitLab Duo 19 20 1.
The SQL statement above is a quote from our recent Office Hours with Benn Stancil. It’s not a SQL statement that would work today in a cloud data warehouse. But an LLM would understand it : summarize the book Moby Dick in two sentences. Sure enough, ChatGPT answers the question : This pseudocode blends the structured queries of data analysis with the unstructured data contained in a classic novel. This is how Benn views the future of BI
“This is the first of your newsletters that doesn’t align well with what I’ve been seeing in the field.” After publishing The Four Barriers to AI Adoption, Dave Morse, a reader & a friend who was most recently CRO at Hebbia & VP Sales at Scale AI sent me this email. Dave continued : “The biggest blocker to adoption at AI application companies is user education and limitations of frontier models.
AI adoption is slower than expected in many spaces. Some of the reasons are straightforward, but others are more subtle. Most leaders wants to inject AI into their business to develop a competitive advantage. There are four challenges. The first challenge is understanding the technology’s ability. Because the capabilities evolve so quickly, it’s hard to keep up. If PhDs in the domain are rushing to understand the capabilities reading papers every week, how are business leaders meant to grok the state of the art?
It’s time for the 2024 Annual Theory Go-to-Market Survey. This is a brief 28-question survey. Our goal is to understand how startups have evolved their sales, marketing, customer success, and cash management over the last four years by comparing these results to those through the go-go years of 2020 and beyond. We will publish these results and answer questions about them at upcoming Office Hours. With this data, we should be able to draw some broader conclusions about the shift from growth to efficiency & determine if the buyer behavior changes in the private market parallel those in the public market.
What does it take to go public? Has it changed over the last 15 years? We gathered data on the US venture-backed software companies that went public between 2010 & today. We corrected the trailing 12 months’ revenue at the time of IPO for inflation & plotted the data. Before 2018, only one company IPOed with more than $200m in revenue. In fact, the median revenue at IPO at $90m. Today, the median revenue at IPO is $189m (corrected for inflation), more than double.
2012 was the year of the Seedpocalypse. Also called the Series A Crunch, a fear gripped Startupland : raising a Series A. Two years later, this indigestible excessive bolus of fundraising rounds hit the Series B market & Series Bs became the most challenging round to raise. Whenever there are “too many” of fundraises of one type, the next round becomes the hardest to raise. In 2024, the Series A Crunch has returned.
I was chatting with a friend of mine about the advent of robotic surgery and he was lamenting the challenges associated with training younger doctors. Before robotic surgery, medical surgeons stood shoulder to shoulder alongside seasoned surgeons operating. Today, the head surgeon manipulates a robot independently while students watch through a window or video. A lot has been written about training AI. But what about training humans? Shouldn’t the same pattern reverberate through the work that we expect the next generation of AI to automate, including paralegal functions, accounting, computer programming, and sales development?
Databricks revealed some sensational growth this week, as they did last year. Exiting this quarter to $2.4 billion annual run rate, the company’s revenue growth is accelerated year-over-year by 10 percentage points. Quarter Q2 2023 Q2 2024 Quarterly Revenue, $m 375 600 Revenue Growth 50% 60% Customers 10,000 Gross Margin 85% 80% Net Dollar Retention 140% Data Warehouse Revenue, $m 100 400 Average Annual Customer Value 37,500 Net dollar retention is a major driver of growth at 140%, which is top decile.
What stood out to me in yesterday’s Apple announcement wasn’t the headline, but the subtitle. “Setting a new standard in privacy.” For privacy to become one of the leading selling points of software, competitive dynamics & user preferences have evolved. The mantra repeated over the last 20 years on the internet has been privacy is dead. Users simply don’t care. People are willing to trade their privacy for free & targeted experiences.
One line of software can impact a billion rows of data. A short SQL statement can delete a table, reformat date into the US format, or compute the average quota attainment by account executive by region over a company’s history. As data has become a critical component of analytics & production systems, data engineers require more sophisticated tools to manage their data transformations. Tobiko is that solution. Tobiko enables modern data transformations by calculating only what’s needed.
Last week, public software markets suffered significant compression. MongoDB fell 24%; UIPath fell 36% ; Salesforce fell 15% ; Workday was down 11%. Weaker revenue projections tend to cause sell-offs. These large drops aren’t unprecedented. In 2016, valuations fell 57%. Is it different this time? Growth rates have changed meaningfully. The 25th, 50th, & 75th percentiles for public growth rates have halved in the last 18-24 months. The grey bar indicates Covid ending & marks the beginning of the slide.
If I asked you, “When someone turns in a work assignment, how accurate is it? 80%, 90%, 95% or perhaps 100%?” We don’t think this way about coworkers’ spreadsheets. But we will probably think this way about AI & this will very likely change the way product managers on-board users. When was the last time you signed up for a SaaS & wondered : Would the data be accurate? Would the database corrupt my data?
NVIDIA’s growth is an index on the growth of AI. “Compute revenue grew more than 5x and networking revenue more than 3x from last year.” Data center revenue totaled $26b, with about 45% from the major clouds ($13b). These clouds announced they were spending $40b in capex to build out data centers, implying NVIDIA is capturing very roughly 33% of the total capex budgets for their cloud customers. “Large cloud providers continue to drive strong growth as they deploy and ramp NVIDIA AI infrastructure at scale and represented the mid-40s as a percentage of our Data Center revenue.
For the past few days, I’ve been using the Mac ChatGPT app OpenAI demonstrated last Monday. It’s unquestionably the future of human-computer interaction. Conversing with a computer is much more natural than typing. Imagine speaking t a colleague with the entire internet at their disposal. But also a verbose colleague without much sense of social cues. Tapping a keyboard shortcut, the ChatGPT app loads & four little bars reminiscent of Google transcription software appear in the app.
Ramp published its quarterly spending trends & revealed how businesses are spending on AI. There are many great data points that underscore the growth in AI but there are important nuances in the patterns. First, AI growth rates across the most popular vendors have fallen 78% annually. On average, these AI businesses are growing customer counts 105% ; the median is 38%.1 However, this isn’t a uniform trend. Some companies have seen contraction in customer counts.
On Thursday May 30th at 9:30 Pacific time, Office Hours will host Benn Stancil. Benn co-founded Mode, a analytics platform acquired by ThoughtSpot. He writes about the future of data & analytics on his blog & asks deep questions about core tenets of software including Do software companies actually have good margins? & Disband the analytics team. With a unique lens on data & software, Benn will share his views on some of these bigger topics including BI’s third form.
The Theory AI Index of publicly traded companies continue to outperform more classical software companies since we published the index earlier this year. In addition to the delta in absolute multiple, is there a difference in the factors that drive valuation? Across the five most important metrics, AI companies’ valuation correlates isn’t that much different than non-AI companies. Revenue growth, sales efficiency, & cash flow are all relatively similar. So is net income margin, although the AI correlate is about half.
When a person asks a question of an LLM, the LLM responds. But there’s a good chance of an some error in the answer. Depending on the model or the question, it could be a 10% chance or 20% or much higher. The inaccuracy could be a hallucination (a fabricated answer) or a wrong answer or a partially correct answer. So a person can enter in many different types of questions & receive many different types of answers, some of which are correct & some of which are not.
The fastest growing category of US venture investment in 2024 is AI. Venture capitalists have invested $18.3 billion through the first four months of the year. At this pace, we should expect AI startups to raise about $55b in 2024. AI startups now command more than 20% share of all US venture dollars across categories, including healthcare, biotech, & software. In the preceding eight years, that number was about 8% per year.
Amazon announced their earnings yesterday. Like Microsoft & Google, Amazon’s Web Service business is seeing a surge of growth, up from 13% annual to 17% annual growth (16% when excluding the leap year). Aside from the overall growth of these clouds increasing, the massive investment in CapEx data centers, power plants, and GPUs is stunning. Cloud Capex in Q1 AWS $14 billion Azure $14 billion Google Cloud $12 billion These are not one-time investments, but part of a broader trend that started to occur after the introduction of GPT 3 in mid-2020 Amazon was the first to invest significantly.
What drives the acquisition market of startups? It’s the big deals. In the last decade, the total number of venture backed software M&A by count has remained relatively constant. The black line shows the linear trend across US venture backed companies with disclosed values of $50m or more. The average & median counts by year total 58 & 55 respectively. If there are any increases, they tend to be in the bigger acquisitions of $500 million or more - although the sample size there is sufficiently small to conclude the trend is significant.
If it wasn’t clear before, AI is the single biggest revenue driver in cloud. Microsoft’s Azure is winning share directly from Amazon. In Feburary, Microsoft grew 2% & Amazon lost 2%. Google is also taking share - 1% in the last year. A one percentage point share shift represents about $750 million of spend or about $5b in market cap. Quarter Microsoft Amazon Google Q4 2021 21% 33% 10% Q4 2022 23% 33% 11% Q4 2023 24% 31% 11% “We’re fundamentally a share taker there because if you look at it from our perspective, at this point, Azure has become a port of call for pretty much anybody who is doing an AI project”
Enterprises spend more on security but aren’t benefitting from the extra spend. Palo Alto Networks’ customers who buy security across 3 platforms spend more than 40x those that secure just one. Despite those dollars… “Adding incremental point products is not necessarily driving a better security outcome for them.” - Nikesh Arora, CEO of Palo Alto Networks The average enterprise uses upwards of 70 security products. Many of these products produce alerts identifying phishing emails or network access issues or odd device behavior.
Snowflake announced Artic, their open 17b model. The LLM perfomance chart is replete with new offerings in just a few weeks. Overall knowledge performance is asymptoting as expected. It’s hard to discern the most recent dots. One thing stands out from the announcement - the positioning of the model. “The Best LLM for Enterprise AI” Snowflake focuses on the model’s enterprise performance : SQL generation, code completion, & logic. This push will be echoed by others as models start to specialize.
On Friday May 3rd at 9:30 Pacific time, Office Hours will host Jordan Tigani, CEO of MotherDuck. Jordan’s path through data starts with being the founding tech lead on Google Big Query to SingleStore’s Chief Product Officer to MotherDuck’s CEO. He wrote the Big Data is Dead, explaining that a power law applies to most companies’ data. One of the most innovative things about DuckDB & MotherDuck is the ability to start querying with local data & then use the cloud when necessary.
In “Do software companies actually have good margins?”, Benn Stancil makes a case for a counterintuitive point : Software companies are much less profitable than they might seem. Why? Because the research & development costs associated with software should be part of their cost of goods sold. Software companies don’t invest once in R&D & then sell copies of the software as we did in the 90s on CDs. It’s software-as-a-service.
Ethereum generated $370m in profit on $825m in revenue for about a 45% net income margin. The chart above shows both the historical performance & also explains how web3 blockchains like Ethereum generate revenue & profits. How does Ethereum compare to other software companies? If Ethereum were to trade on the New York Stock Exchange or the NASDAQ, it would top the net income margin (%) charts, with Microsoft, Adobe and Veeva thereafter.
Over the weekend, Tobi, the founder and CEO of Shopify, discussed the major reason investors passed on Shopify in the early days : market size. Good supercut about why Shopify worked https://t.co/pw92KutpEn — tobi lutke (@tobi) April 14, 2024 I remember that financing round, & I remember having the same concern, & making the same mistake. Living in the valley & driving on 101, the billboards & logo-adorned headquarters of successful companies provide a daily infusion of all the mistakes in I’ve made in guessing how a company or a market might evolve.
Recently, I was at the library & stood in front of a book that might be a good read. I fired up Gemini, asked it for a plot summary & a digest of the reviews before deciding to read it. So I started a list : what are the ways I use AI that are new in the last year? I often publish tables within these blog posts. I write posts in markdown, a language that requires making tables in a unique format that looks like this
First Round’s Product Market Fit Framework adds dimension & texture to a concept we all use - Product Market Fit (PMF). PMF isn’t attained permanently, but sustained over time as companies & complexities grow. As they scale, startups move up-market, expand into new geographies, add new product lines, & develop new customer acquisition motions. These complexities are initially linear but eventually multiply. The idea is best illustrated by example. Looker is one of the companies highlighted 🤗 & a portion of their journey is detailed here :
On April 18th at 9:30am Pacific time, Office Hours will host Evan Cheng, founder & CEO of Mysten Labs, creator of SUI. Evan & the Mysten team were instrumental in the creation of Meta’s high-performance blockchain techology before leaving to start Mysten. There are few key ideas Mysten espouses about blockchains that have led to dramatic improvements in performance & usability. First, the Move language & virtual machine form the basis of the smart contract layer.
Product managers & designers working with AI face a unique challenge: designing a delightful product experience that cannot fully be predicted. Traditionally, product development followed a linear path. A PM defines the problem, a designer draws the solution, and the software teams code the product. The outcome was largely predictable, and the user experience was consistent. However, with AI, the rules have changed. Non-deterministic ML models introduce uncertainty & chaotic behavior.
Situational Management is a framework for deciding how to manage a report depending on their skill level & motivation. Where does AI fit today? The answer is easy. It’s in the micromanage category where the motivation is high but the skill is low. The computer is relentless in wanting to users’ questions or complete their code. But the skill level is relatively low - about the same as a high-school student.
The average American attention span has fallen from 150 seconds in 2004 to 75 seconds in 2012 to 47 seconds in 2023 - a 5-6% annual rate of decline. Year Avg American Attention Span (sec) CAGR 2004 150 - 2012 75 -6% 2023 47 -5% How does this compare to these blog posts? In 2013, the average reader dwelled on this site for 47 seconds. Today, it’s 33 seconds, a 3.