It’s time to start assessing the early winners of the AI race – here is how I do it.
ChatGPT launched nearly two years ago, marking the start of a new era where AI has become the central theme in tech investing. Since then, we’ve seen virtually every tech company seemingly “pivot” towards an AI-based business model.
With the AI race now in full swing, I think it’s time to have a serious conversation about the landscape of investing in AI. As an investor, I often find it challenging to distinguish which companies are genuinely leveraging AI to drive value and which are merely “faking it.”
Recently, I’ve started using the Rule of 40 as a key financial indicator to assess whether a company’s AI efforts are beginning to translate into monetization. I touched on this in my last article about Palantir, where I highlighted the company’s impressive Rule of 40 score as one of the reasons I am doubling my position in the stock. Following that article, a few readers reached out, seeking more insights on the Rule of 40 and my perspective on investing in AI.
In this new article, I will explore what, I believe, are the main applications of AI today—from chatbots to military software. Additionally, I will review the Rule of 40 scores for companies claiming significant investments in AI to gauge whether they are successfully monetizing these efforts.
The analysis will reveal that while AI can be effectively monetized and used to accelerate growth in established businesses, the path to monetization is often complex. Notably, a few key players are currently capturing the lion’s share of benefits in the AI space. I see AI as a catalyst that will intensify competition in the already volatile tech sector, potentially paving the way for new tech leaders to emerge while some of today’s giants may struggle to keep up.
The 3 categories of AI applications and their key players
AI is a broad term that has recently gained popularity in the business world to describe a range of technologies that have actually been in development for decades. Prior to the launch of ChatGPT, terms like ‘machine learning’ and ‘chatbots’ were more commonly used to refer to specific applications of what we now collectively call ‘AI.’
In this context, it can be difficult for investors to understand what exactly are the applications of AI and how they relate to businesses. I find online material on the subject to be confusing, as often it consists simply of a list of applications by industry. I do not think this kind of clustering is helpful to truly understand the impact that AI can have on a business.
I personally consider all applications of AI today as falling into one of three macro-categories:
-
Creative work.
-
Data analytics & Cybersecurity.
-
Automation & Robotics.
There is also a fourth category, which I call “Picks & Shovels.” This includes companies selling the hardware needed for AI tools and models to be developed, such as NVIDIA Corporation (NVDA) or Advanced Micro Devices, Inc. (AMD).
I believe AI applications in the three categories above concern all industries – from agriculture to healthcare and information technology. In the next paragraphs, I will illustrate my thinking with a few examples for each category.
AI applications in creative work
AI applications under the category of “creative work” are what most consumers are familiar with, for example generating text using Large Language Models (LLMs) such as ChatGPT. Other examples of “creative work” include:
- Copywriting and ghostwriting
- Image editing
- Language translations
- Gaming
- Personalized education and learning
- Writing news pieces
I see LLMs such as ChatGPT, Google Gemini or Microsoft Copilot as the main examples of the AI tools used for creative work.
In terms of companies operating in this space, I see Adobe Inc. (ADBE), Autodesk, Inc. (ADSK) and BuzzFeed, Inc. (BZFD) as some of the key players in the space. Adobe has recently included services based on LLM technology in its suite of products. Customers of Adobe can now generate or edit images using AI in photoshop, as well as asking Adobe’s AI assistant to summarize or extrapolate information from a PDF. Additionally, Adobe is leveraging AI to sell Adobe Cloud, a suite of products that competes with more established CRM systems (something I will cover in the next category of AI applications).
BuzzFeed has recently announced it was able to lay off 12% of employees, replacing their jobs with AI tools. Most of the jobs impacted by layoffs are related to copywriting and writing news. While I do not consider BuzzFeed a company that is really playing the AI game, I do consider as a company that has benefited from AI earlier than others.
Tech giants such as Microsoft Corporation (MSFT) or Alphabet Inc. (GOOG) are also operating in this first category, not least because they develop the very AI tools that make AI applications in creative work possible.
AI applications in Analytics & Cybersecurity
Under the “Analytics & Cybersecurity” category, I include all AI applications that aim at transforming big data into actionable output. I see prime examples of companies with AI and analytics at their core in Palantir Technologies Inc. (PLTR), Snowflake Inc. (SNOW) and CrowdStrike Holdings, Inc. (CRWD). All of these companies market full-AI products, such as Charlotte AI for CrowdStrike, the AI Data Cloud for Snowflake and AIP for Palantir.
Other examples of companies in this category include Salesforce, Inc. (CRM) and HubSpot, Inc. (HUBS), both active in the CRM (Customer Relationship Management) industry. The key about CRM is the ability to leverage user data to engage with customers and prospects at the right moment, with the right message, predicting behaviors and needs. While this is often referred to as “automation”, in CRM, I consider it an application more closely related to analytics. I prefer to use the term “automation” for stricter AI applications – something I will cover in the next few paragraphs.
More generally, I see the following as the main examples of AI applications in Analytics and Cybersecurity:
- Customer Relationship Management
- Supply Chain Optimization
- On-field military aid and tactics support
- Algorithmic trading
- Network protection
- Spam filtering
AI applications in Automation & Robotics
Finally, I define “Automation & Robotics” as AI applications that have to do with substituting complex human labor with more cost-efficient and cost-effective artificial input. Examples of what I consider AI applications falling in this category include:
- Virtual Assistants
- Targeted Advertising
- AI-powered Search Engines
- Robotaxis
- Humanoid robots
Established tech giants such as Apple Inc. (AAPL), Microsoft, Meta Platforms, Inc. (META) and Alphabet are all primarily active in this category, in my view. Microsoft, for example, is actively pushing to include AI tools in its product offering. This includes the recent launch of Microsoft Copilot in its suite of Office products. Similarly, Google and Apple have recently embedded AI assistants in their respective mobile Operating Systems.
Google and Meta also use AI automation to improve the targeting capabilities in their advertising. These latter examples are fundamentally very similar in terms of technology to what CRM’s players use to predict customer behaviors. I have classified Meta and Google as primarily active in the Automation & Robotics category, but I believe major tech giants do increasingly play a role across these categories.
A final example in the category of Automation & Robotics concerns Tesla, Inc. (TSLA), which I see as a major player and potential beneficiary from AI technology. Tesla is allegedly developing robotaxis and humanoid robots, as part of what the company calls “the Tesla Ecosystem.” While neither of these two products have yet been launched, I do see the company as a potential major player in the AI space.
AI applications at a glance and a note about my thinking
I have divided AI applications into three (plus one) categories, summarized in the chart above. This split is not perfect, not last because the technology behind different applications is ultimately very similar and there are some overlaps between categories. For example, LLMs can be used for copywriting or generating images, but they can also be used to analyze a vast quantity of data.
The line between applications in Analytics & Cybersecurity and those in Automation & Robotics can also be ambiguous at times. The AI technology that Google or Meta uses to optimize advertising is fundamentally not very different from the technology Salesforce utilizes in its CRM systems. Similarly, large-scale data processing can “teach” a car autonomous driving or enabling a robot to walk.
In my classification, the focus is on the end-use of the technology rather than the technology itself. This approach is why I view major tech companies as operating across multiple categories. My categorization considers the primary applications of a company’s product offerings. For example, I primarily classify Microsoft within the Analytics category, but it’s also clear to me that Microsoft has a footprint in online advertising with Bing, engaging with AI in ways similar to Google and Meta for advertising purposes.
Ultimately, my hope is that this classification – albeit not perfect – can help investors better understand the AI landscape.
What the Rule of 40 tells me about investing in AI.
The Rule of 40 is a financial metric used primarily by venture capitalists to evaluate the performance and growth efficiency of SaaS companies.
It states that the combined growth rate and profit margin of a company should exceed 40%. This metric helps VCs assess whether a company is striking the right balance between growth and profitability.
For example, if a company has a 30% revenue growth rate and a 10% profit margin, it meets the Rule of 40. Conversely, a company growing at 50% but with a -20% profit margin would also meet the criteria.
VCs use this rule as a benchmark to identify companies that are managing their resources effectively and are likely to be sustainable and scalable investments.
I think the “Rule of 40” can be helpful to assess whether companies are “faking it” or “making it” in the AI race. I see AI as a new technology that can drastically improve productivity, as well as fuel the launch of new, revolutionary products. In this context, I see all Tech companies claiming to have AI at their core as similar to SaaS startups, in that they are quickly developing and scaling new products.
Using the Rule of 40 has some limitations – something I will cover in the risk section of this article. However, I find it comes very close to being the single best metric to gauge whether a company is successful in AI.
Rule of 40 scores: who are the early winners of the AI race?
The chart above summarizes the 12-month Rule of 40 scores for a selection of publicly listed companies that I consider major players in AI, which I discussed earlier in this article.
It’s important to note that the most recent Rule of 40 scores may differ from the 12 months data shown above. This is because many of these companies have experienced strong acceleration in their businesses over the last quarter or two, boosting their growth and profitability metrics.
For instance, Palantir, my largest single stock holding, recently reported a Rule of 40 score of 64, up from a 12-month score of 49. Despite this, I’ve chosen to present the 12-month scores as I believe they provide a more stable metric, less skewed by the impact of a single quarter’s performance.
One key insight from this data is that a few “Pick and Shovel” players are capturing the majority of benefits in the current AI landscape. Nvidia, for example, holds what I see as the equivalent of a monopoly on selling shovels during the Far West gold rush—not through government mandate, but due to its unmatched cutting-edge technology that no other microchip manufacturer has yet truly been able to rival.
Another interesting observation, in my view, is how established tech giants like Microsoft and Meta are growing at rates typically seen in scale-up companies. I see this as proof of the transformative impact AI can have on even the most mature businesses, demonstrating its potential to significantly accelerate growth.
My main takeaway, however, is that not all tech companies are equal in the AI race. If we, as investors, believe that an AI-centric business model is essential today, we must question whether the “Magnificent 7” will continue to dominate in their current form indefinitely.
It is apparent to me that companies like Google and Apple have yet to establish a strong footing in AI. Both have lagged behind their tech peers: Apple is only just releasing an AI-centered iOS update this September, while Google faces ongoing challenges in diversifying its revenue streams beyond Google Search Ads. Despite integrating AI into both Android and Search, Google has not yet seen significant improvements in growth or profitability from these efforts.
I expect increased volatility in the tech sector moving forward, with AI serving as a catalyst that could redefine the landscape of “best in class” tech companies. As AI continues to evolve and integrate into various business models, it has the potential to disrupt existing market leaders and elevate new contenders. I think investors should be prepared for rapid shifts as AI reshapes the hierarchy of top tech performers.
Risks to my thesis
While I believe the Rule of 40 is a valuable indicator of whether a business is “making it” or “faking it” in AI, there are notable limitations to relying solely on this metric.
Firstly, the Rule of 40 can be heavily influenced by a company’s strategic decisions. For instance, Tesla’s Rule of 40 score is currently just 2, which I attribute to the company’s strategy of sacrificing margins to grow the EV category, rather than reflecting any shortcomings in its AI efforts. Much of Tesla’s AI work is focused on future products like robotaxis and humanoid robots, which, in my view, the market is prematurely discounting.
Another limitation of the Rule of 40 is that it doesn’t provide insights into a company’s valuation or implied risks. For example, CrowdStrike has an impressive 12-month Rule of 40 score of 66, yet it is currently grappling with significant litigation risks due to the recent global outages its product caused. Despite its high score, I would not invest in CrowdStrike at this point in time, given the difficulty in quantifying these legal risks.
Ultimately, I view the Rule of 40 as a gauge of whether a company is effectively leveraging AI to accelerate its business. However, it should be seen as just one data point among many. Investors should avoid basing their decisions solely on a company’s Rule of 40 score. Instead, they should consider this metric alongside other critical factors, such as the company’s product strategy, AI vision, and stock valuation.
Conclusion
I expect AI to drastically change the lineup of tech companies we consider world-leading today. From analyzing the current AI landscape and assessing how companies perform in terms of the Rule of 40, it’s clear to me that only a few are truly managing to adopt AI as a core part of their business.
As an investor, I generally prefer a diversified approach to tech, which is why my primary exposure is through QQQM, a Nasdaq 100 ETF. However, I also hold a few significant single stock positions, including Palantir. When choosing these bets, I increasingly rely on the Rule of 40 as a key financial metric to gauge whether a company is genuinely monetizing its AI initiatives.
For investors looking to select single stocks as AI bets, I recommend starting with a deep understanding of the company’s products or services and the role AI plays in enhancing them. Using the Rule of 40 can then help assess whether the company’s product strategy is translating into tangible financial results.
Editor’s Note: This article covers one or more microcap stocks. Please be aware of the risks associated with these stocks.
Credit: Source link