Revisiting the “Is Software Expensive” QuestionThe current median revenue multiple for the software universe is 6.4x, which is ~17% below the long term average multiple of 7.8x. However, when looking at growth adjusted multiples, the current median is 0.48x, which is ~75% above the long term average of 0.28x (you can see this chart later on in this post in the EV / NTM Rev / NTM Growth section)! So on a pure revenue multiple basis the software universe looks “slightly cheap.” But on a growth adjusted basis, the software universe looks “super expensive.” Below is a chart that shows how growth has really trailed off the last few years. During the period where the long term average multiple was 7.8x, the median projected growth rate was 27%. The median projected growth rate today is 14%The pi

Going in to this earnings season I expected a big driver of stock price performance to be how 2024 full year estimates change pre / post earnings. Since the macro feels more “stabilized” it felt fair to assume “are numbers going up or down” to be the question most correlated with stock price performance. After a few software companies have reported we can start to look at the data behind “are numbers going up.” Surprisingly, despite the huge moves in stock price ($Palantir Technologies Inc.(PLTR)$ / $Confluent, Inc.(CFLT)$ both +30% and $Cloudflare, Inc.(NET)$ was +20% after hours yesterday), we really haven’t seen forward estimates change that much. The chart bel

If AI delivers on its promise, it may spell the end of the SaaS business model as we know it. Historically, cloud software businesses charged a recurring fee based on the number of users of their software - the SaaS model. Then infrastructure / dev tools software companies took a different approach - more of a consumption based pricing model (I’m generalizing, not all infra is consumption based). Look at $Twilio(TWLO)$ - they charge per message sent. Or $Snowflake(SNOW)$ who charges based on compute used or data stored. At the end of the day, software vendors want to align value delivered with price charged. Back to the Twilio example - let’s say I, as a solo developer, built an application on Twilio that

Q2 earnings season is now behind us. I’ll provide a more comprehensive recap later, but for now, let’s look at how the future outlook has changed after Q2 earnings. One metric I like to examine is how much companies adjust their full-year guidance. Generally, software companies follow a beat-and-raise model in their forecasts. The “raise” part reflects how much the future outlook (i.e., the guidance) changes. When we compare full-year guidance for 2024 from the Q2 call versus the Q1 call (~3 months ago), you’ll see in the graph below that, for the most part, full-year guidance didn’t change significantly. The median full-year “raise” was only 0.3%.Looking at Q2, the median “beat” (i.e., how much a company’s Q2 results exceeded consensus) was around 1.5%. So, while the median quarterly beat

This week the 3 hyperscalers reported ( $Amazon.com(AMZN)$ AWS, $Microsoft(MSFT)$ Azure and $Alphabet(GOOG)$ $Alphabet(GOOGL)$ Google Cloud). What did we learn? Most importantly - they ALL called out still being meaningfully capacity constrained. CapEx guides are going up, data center builds are going up, power constraints are meaningful. This isn’t the telecom bust where the world laid fiber that was “dark” (ie unused). GPUs are being used the second the come online…Here are the numbers:AWS (Amazon): $132B run rate growing 20% YoY (last Q grew 17%)Azure (Microsoft): ~$93B run rate (estimate) growing 39% YoY (last Q gre

If I had to simplify software market cycles, I’d say they come in two phases: the expansionary phase and the consolidation phase.In the expansionary phase, buyers scoop up software almost indiscriminately. There’s little concern for cost or efficiency, what matters is speed. It’s about accelerating product development, capturing market share, or outspending competitors to stay ahead, all under the assumption that growth will take care of everything else. During this phase, public markets shift their focus entirely to growth over profits. Take a look at the multiples chart I post later on breaking out multiples by high, medium, and low-growth companies. You can see the high-growth bucket has seen multiple expansion this year, while the mid-growth bucket has seen steady contraction.In the co

The AI FactoryIf I zoom out for a moment and look at the current trajectory of AI infrastructure, it’s hard not to see an entirely new pattern forming. Inference keeps getting faster. Inference engines keep getting smarter. And the ecosystem around them keeps getting more modular and open. What once felt like specialized machinery locked inside a handful of labs is now drifting into the hands of every company with a GPU budget and a few strong engineers.Neoclouds like $CoreWeave, Inc.(CRWV)$ and $TOGETHER PHARMA LTD.(TGPHF)$ have rewritten the economics of GPU access. Inference clouds like Fireworks, Baseten and fal have done the same for reliable serving (and we’ve already separated into separate infere

In the world of traditional software, products were built on determinism. You shipped a feature, QA tested it, and expected it to behave the same way tomorrow as it did yesterday. Releases were milestones (that admitedly have seen the time between them drastically fall over the years as software is shipped daily or multiple times a day). But regardless, once something was stable, it generally stayed that way.AI breaks this assumption. AI products now look more like living systems vs static ones.Models drift as data changes. AI providers push new models that improve performance on average but can regress in specific cases. Or, the updated model just “feels” different and users don’t like it (kind of similar to when Facebook back in the day would update the UI of the newsfeed, and people wou

This week in enterprise software: Top 10 #SaaS #Cloud multiples as of today's market close $Palantir Technologies Inc.(PLTR)$ $Cloudflare, Inc.(NET)$ $CrowdStrike Holdings, Inc.(CRWD)$ $Samsara, Inc.(IOT)$ $ServiceNow(NOW)$ $Guidewire(GWRE)$ $Palo Alto Networks(PANW)$ $Rubrik Inc.(RBRK)$ $Snowflake(SNOW)$ $Veeva(VEEV)$ ImageThis week on Clouded Judgement: Authentication

This week on Clouded Judgement: New Rules for a New Game- Median software multiple: 5.3x- High Growth software median: 21.8x- Mid Growth software median: 8.5x- Low Growth software median: 4.1x- 10Y: 4.2%ImageThis week in enterprise software: Top 10 SaaS Cloud multiples as of today's market close $Palantir Technologies Inc.(PLTR)$ $Cloudflare, Inc.(NET)$ $CrowdStrike Holdings, Inc.(CRWD)$ $Zscaler Inc.(ZS)$ $Snowflake(SNOW)$ $ServiceNow(NOW)$ $Guidewire(GWRE)$ $

This week in enterprise software: Top 10 #SaaS #Cloud multiples as of today's market close $Palantir Technologies Inc.(PLTR)$ $Cloudflare, Inc.(NET)$ $CrowdStrike Holdings, Inc.(CRWD)$ $Guidewire(GWRE)$ $Snowflake(SNOW)$ $Zscaler Inc.(ZS)$ $Rubrik Inc.(RBRK)$ $Datadog(DDOG)$ $Shopify(SHOP)$ $ServiceNow(NOW)$ ImageThis week on Clouded Judgement: The AI Operating Model- Med

AWS report is out - seeing some reversal in incremental share gained from $Microsoft(MSFT)$ Azure to $Amazon.com(AMZN)$ AWS (45% incremental share -> 41% / 30% -> 34%)Recall - AWS is the bell weather for many SW stocks. Many of them will be up tomorrow!ImageAWS crossed a $100B run rate growing 17%! Truly MASSIVE scale Quarterly YoY growth trends in below chart.ImageOne of my favorite charts - this shows the Quarterly absolute change in revenue YoY. So the most recent data point shows Q1 '24 AWS rev - Q1 '23 AWS revImageBelow is an estimate of net new ARR added over the last few quarters.ImageSome additional charts on the aggregate hyperscaler data from Q1. First, the chart below shows the YoY growth

Model distillation might be the most important shift happening in AI right now—and it’s reshaping the entire tech industry. It's increasingly becoming a MASSIVE topic. DeepSeek's R1 model released yesterday only reinforced thisModel distillation is a process where a smaller, simpler model (the "student") is trained to replicate the behavior and capabilities of a larger, more complex model (the "teacher"). This is achieved by using the teacher model's outputs (e.g., predictions or reasoning processes) as training data, allowing the student to inherit high performance with reduced size and computational demands.So why is this important? For large AI labs, capital and scale were moats. It took literally billons of dollars of compute and data to pre-train a state-of-the-art model. Let alone al