Stock idea: In pursuit of Alpha with Palantir
Updated: Aug 9
Alpha, the ability to outperform, is something we all search for in our personal and work life. We strive to achieve more and exceed expectations. Companies and investors obsess over it. Alpha is also what my most recent stock idea is focused on. Palantir (NYSE: PLTR), the sometimes controversial data platform company, describes its software's purpose as enabling their customers to generate alpha. Judging by the ambition and achievements of customer use cases on Palantir software I don't see this as hyperbole. They are helping their customers pursue or achieve moonshot goals, including building safer cars, uncovering human trafficking rings, preventing money laundering, fighting hunger, delivering lower carbon energy, accelerating cancer research. For heaven's sake the US military uses Palantir to plan and execute its missions.
The thesis for investing in Palantir (or not) is pretty straightforward. You either believe they have the best software for helping companies understand, decide and act on their ever-growing treasure trove of data; or not. And, that the value Palantir is able to extract from the messiness of an organization's data is so important customers will pay headline grabbing sums of money. From what I've been able to decipher after studying this company for the last couple of months is that they simply offer the best widget for turning data exhaust into fuel that helps companies generate differentiated value and outperform their competitors. The alpha Palantir generates for its customers should lead to alpha generation for its investors.
The audacity of what Palantir seeks to achieve with it's software traces back to its origin story. Palantir was founded to help the US government fight terrorism without infringing on civil liberties. Palantir was born out of a necessity the U.S. government had following the 9/11 terrorist attack. While it was Peter Thiel, Paypal co-founder and famed venture capitalist, who had the initial idea to use a data driven methodology to help the U.S prevent terrorism, he tasked his former law school roommate Alex Karp to lead the company as CEO.
The methodology borrowed heavily from what Thiel and other members of the PayPal mafia used to reduce fraudulent behavior on PayPal. Thiel decided that it was now going to be codified to help America with its war on terror. Palantir's first product helped intelligence analysts spot patterns from billions of different data points being collected across email, SMS, documents, phone calls, and many other signals. Palantir software was built to make data intelligible, which is especially hard to do at the exponential rate data has and continues to grow. By spotting patterns Palantir's software was able to help intelligence officers identify bad actors. Meanwhile, to keep the government accountable citizens (as represented by other branches of elected government officials and attorneys) needed to verify the government was legitimately tracking down a potential terrorist and not nefariously spying on its civilians. This is what Palantir helps solve - spotting incredibly powerful patterns from an over abundance of data, while complying with strict data privacy and security standards.
Prior to Palantir, how did this unique problem get solved? Not very well. Teams of computer and data scientists would be thrown at these data analysis projects to aggregate a organization's data and write complex algorithms that only they understood. This is not a formula for making sense of and acting on insights from mountains of data, at scale. Emancipating an organization's data and democratizing the ability for all to discover insights thru easy to use tooling is required. To address this need Palantir created an out-of-the-box platform that organizations put on top of their disparate data repositories so that anyone can be an expert at finding meaningful patterns buried deep in data.
But having more powerful data insight tools is not enough to generate alpha. Returning to Palantir's first mission to help win the the war on terrorism, one of the main breakdown's that led to the worst terrorist attacks on U.S. soil was a lack of collaboration between the necessary government agencies. The post 9/11 commission found the point of failure that resulted in the September 11th World Trade Center terrorist attack was various government agencies inability to share intelligence. The challenge was one of privacy and security. Palantir built sophisticated security and privacy technologies that met the requirements of the most demanding customers.
Since their very first government customer, the central problem of helping organization integrate, analyze, collaborate on and act on their data has only gotten harder and more important. More disparate systems, the explosion of social media, billions of sensors and things getting connected to the Internet, accelerated growth rates of data. The threat and security landscape has gotten more complex, while privacy requirements have gotten stricter.
Palantir has risen to the challenge. It morphed into a software-based solution from its days as more of a technology consultancy. Given the secretive nature of government use cases and its eccentric CEO Alex Karp, they have had their fair share of controversy. While there are concerns about their ability to penetrate the commercial market, I believe Palantir offers investors one of the most appealing risk/reward quotients and it has potential to be a multi-bagger.
If you are a regular reader of The Wealthy Owl you'll know I have a set of characteristics I use to assess tech stocks as potential multi-baggers. I review Palantir in the context of these six characteristics below to see if it's worthy of consideration for growth investors.
1. Riding a megatrend
I have written extensively on the topic of my favorite investing thematics for the next decade. One of the most compelling, "Data is the new oil", is the tailwind behind Palantir's growth. It is worth reviewing and then outlining what Palantir is doing to take this trend to a whole new level.
The amount of data being generated in our society today is mind bending.
In 2010, Google CEO (at the time) Eric Schmidt famously put this in perspective:
Every two days now we create as much information as we did from the dawn of civilization up until 2003
That was in 2010!
How far have we come in the last 10 years? In the last two years alone, astonishingly 90% of the world’s data has been created (Source: IORG). And, the exponential growth of data is only expected to get steeper as more people come online, social tools get further adopted, more things get connected to the Internet, more of the human genome gets sequenced, and more cat videos get uploaded to YouTube. IDC forecasts that there will be a 10 fold increase in the amount of data on the planet by 2025.
Data is now considered to be one of the most valuable assets on the planet. This shouldn't come as a surprise as you think about the millions of use cases in which data is being used strategically by organizations today. Data is sold to marketers so they can better target advertising and sell more products. Data generated from sensors is used by operators to track machine performance, do predictive maintenance and to optimize delivery routes. Data is being used by companies to develop their next generation of products and services. Banks use data to identify fraudulent transactions. Pharmaceutical companies are using data and analytics to aid in the discovery of new life saving drugs and to accelerate their time to market with these new drugs. Data is used by health care practitioners to dramatically improve health outcomes and to drive down the costs of care.
The ability to harness data is a competitive differentiator that is driving growth for companies that are leading the way. In a McKinsey study they state:
Analytics capabilities have become a differentiating factor in industry competition, as leading players use data and analytics to grow revenue, to enter or even create new markets, to change the nature of their relationship with customers, and to increase organizational efficiencies. Organizations that are lagging behind will need to adapt quickly before the gap grows wider.
The business case for companies to invest in data analytics is survival. Research has found that investing in data and analytics capabilities has high returns. Organizations can use data analytics capabilities to achieve productivity gains of 6 to 8 percent, which translates into returns roughly doubling their investment within a decade. This is a higher rate of return than other recent technologies have yielded, surpassing even the computer investment cycle in the 1980s (Source: Jacques Bughin, “Ten lessons learned from big data analytics,” Journal of Applied Marketing Analytics).
Despite the significant opportunity with data and analytics, much of its inherent value is still going uncaptured. McKinsey in their 2016 Age of Analytics study found most companies are capturing only a fraction of the potential value from data and analytics.
If there is so much opportunity and proven success stories in the area of data analytics, why has value capture of it proven elusive?
The reason organizations struggle to realize the insights and value embedded in their vast pools of data is how this data is being captured and stored today. It's messy. The diagram below is a very accurate picture of what the data estate of most organizations looks like today. Data is disconnected, spread out across hundreds (if not thousands) of point solutions and custom applications, and downstream data lakes. Each new system or custom application compounds the problem. Outside of hiring an extremely expensive army of system integrators that take way too long and are typically unsuccessful in achieving their objectives, there is no easy way to create the connections across disparate data sets that are necessary for finding the insights that create value. Organizations simply aren't able to yield the massive value buried deep in their data.
This is the problem that Palantir is focused on helping their customers solve. Unlocking the value of data to differentiate and generate alpha. Palantir software allows organizations to recast their siloed systems into contributors to a unified data ontology, transforming their data into insight that can then be acted upon from within the Palantir platform. One of the most appealing aspects of Palantir's software is that it does not displace existing systems. Palantir helps customers squeeze out more value from existing investments. As the world changes and technology evolves, institutions can add to their data model and integrate new systems, rather than rebuilding everything from scratch.
2. Market opportunity
Palantir defines its market opportunity across commercial and government sectors, consisting of two platforms - Gotham for government customers and Foundry for commercial enterprises. They have estimated their total address market to be $119 billion, $63 billion government ($26 billion U.S government, remaining from international governments) and $56 billion commercial. Palantir's total addressable market (TAM) is difficult for Palantir and analysts alike to estimate. There is really no comparable to Palantir's software. They have estimated their TAM using third-party industry publications' projected software spending in areas such as data management software, integration and orchestration middleware, analytics and artificial intelligence, security, system and service management, application development, and applications for enterprise resource management, supply chain, and content and workflows. Palantir included subsegments that are applicable to its platforms within these categories, commercial firms with $500 million+ in annual revenue (6,000 companies), estimated government spending based on software and consulting spending, and they excluded potential spending based out of China and Russia. Palantir only does business with US government and it's Western allies.
While these estimates may be overblown, I believe they are directionally sound and I'm pretty certain Palantir's TAM will grow as new products are launched and new use cases emerge. Both the government and commercial markets are massive opportunities. Let's dig into both in more detail.
Palantir was established to serve the government market, it has been their largest and fastest growing market. In their latest quarter, Q1 2021, Palantir grew government revenue by 76%, fueled largely by 83% growth in their US government business. At $616 million in government business in FY20, Palantir has only realized 1% of the $63B TAM government sector. They are still very early.
The government market has some tailwinds that Palantir is very well positioned to benefit from.
First is a 2018 legal victory in federal court that requires the US government to consider commercially available products, such as Palantir's software, before attempting to build their own customer solution. Palantir generated a total of $51.9 million in revenue from their U.S. Army accounts between 2008 and September 2018, when the federal court ruling came into effect. Since the 2018 ruling, Palantir generated $231.8 million in revenue from those same accounts as of December 31, 2020. Just take the the U.S army as an example of how Palantir can accelerate growth in a government account once they win the trust of the customer. You can see why the future is very bright for Palantir in the government sector.
Second, there is an Artificial Intelligence (AI) arms race happening between nations, their governments and commercial sectors. A recent U.S. National Security Commission AI report, which was aimed at ensuring the U.S. retains its #1 Military world super power status amid the rise of China through its development of AI, explains:
"to make recommendations to the President and Congress to “advance the development of artificial intelligence, machine learning, and associated technologies to comprehensively address the national security and defense needs of the United States.”
"China stands a reasonable chance of overtaking the United States as the leading center of AI innovation in the coming decade."
The report includes a recommendation to increase military spending on AI by an additional $32 BILLION by 2025.
Palantir is very well positioned to benefit from a likely increase in government AI spending because of it's growing AI capabilities and its unmatched security and privacy. Palantir is the only SaaS company that has achieved IL- 6 clearance for the U.S. Department of Defense.
Palantir is building its capabilities in AI to meet this growing demand from government. Most recent is Apollo for Edge AI. This technology is helping the US government operationalize AI at scale today. Using an approach they call micro models, Apollo for Edge AI enables customers to train, manage and deploy multiple independently versioned AI models to the edge with ease. What does that mean exactly? One Palantir customer taking advantage of Apollo for Edge AI provides a useful illustration of how the technology is being used in a real world example. The customer is Sarcos, which is using Apollo for Edge AI to push AI into the exoskeletons they build, which is a battery-powered industrial robot that combines human intelligence, instinct, and judgment with the power, endurance, and precision of machines. Think Jarvis, Iron Man's Tony Stark digital assistant available in his Mark V armor suit. Now, think about how the US defense agencies and army might use this technology to fight wars from "mud to space" in the future. We are talking SkyNet.
I believe for Palantir to be able to reach its full potential it must continue to penetrate the commercial market.
Palantir is relatively new to the commercial market, launching its enterprise platform Foundry in 2016. Despite Foundry being used in 40+ different industries, one of the common bear arguments about Palantir is that they have too few commercial customers. Palantir has only penetrated eight of the Fortune 100 companies, 12 and 24 of the Global 100 and 300 companies respectively.
The counter to this bearish view is to consider how much greenfield they have ahead of them. It's not that commercial organizations are not getting value from Palantir's software, just check out some of these juicy headlines about their commercial customer wins:
An energy supermajor deployed the Palantir ERP suite in hours. Within two weeks, they generated $57 million of cash savings and expect to generate $1 billion on an annualized basis
An aerospace customer signed the largest commercial deal, a $300 million five-year renewal in the midst of a pandemic that shook their entire industry
A Fortune 100 consumer goods company deployed the Palantir ERP suite to quickly respond to COVID-related disruptions. Now, Palantir is connecting the entire value chain – from procurement to distribution
A top five pharmaceutical company is linking data from more than 2,000 clinical trials in Foundry to uncover trends across trials and securely analyze outcomes at a population level
Palantir and 3M expand relationship to build dynamic supply chain
Palantir and bp (NYSE:BP) have extended their partnership to support bp as it works towards its ambition to become a net zero company by 2050 or sooner
Germany’s Merck Taps Palantir for Big Data Health Initiative
So if Palantir's Foundry is providing massive benefits and getting renewed in deals with many additional zeros more than the initial contract, why do they have such low penetration rates in the commercial market? To answer this we have to better understand Palantir's “Acquire, Expand, and Scale” sales cycle.
Palantir's sales cycle is long and typically produces negative unit economics until they reach the scale phase with a customer, once some initial value has been realized by the customer.
If you consider Foundry is only five years old and it takes on average 6 - 12 months before the beginnings of material revenue from an account start to appear non the income statement it shouldn't be shocking Palantir has a limited number of commercial customers.
Palantir's upfront business development investments are well placed, as once customers reach the scale phase the unit economics do a complete 180.
It is in the Scale phase of our partnerships with customers that we generally see contribution margin on particular accounts improve. In 2019, we generated $565.7 million in revenue from customers in the Scale phase, with a contribution margin of 55%. In H1 2020, those same customers generated $296.3 million in revenue, with a contribution margin of 68%.
Source: Palantir S-1
The company is also optimistic about the conversion rate from the acquire to scale phase and its abilities to continue to grow margins:
We believe that all of our customers will move into the Scale phase over the long term. We also believe that contribution margin for Scale phase accounts will increase further as we become more efficient at deploying our software platforms across the entirety of our customers’ operations and at managing and operating our software…In H1 2020, the top 25% of customers by contribution margin in the Scale phase as of the end of 2019 had a contribution margin during the period of 89%.
Source: Palantir S-1, emphasis mine
As the commercial business matures and Palantir converts more customers from Expand to Scale, margins will improve, costs will compress, and their current revenue concentration will diversify. Palantir is also investing aggressively to optimize its deployment process and tooling, leading to shorter sales cycles and accelerating time to scale.
If you want evidence of Palantir's ability to move it's customers through it's sales cycle look no further than their 2021 Q1 earnings report where they share the growth they've achieved in average revenue per commercial customer.
Other positive signs include the growth in their commercial market sales pipeline. In Q1 2021 Palantir reported that their commercial opportunities in the US and the UK have increased by 2.5x since the prior quarter. In the same period, active commercial pilots have more than doubled.
Another anchor on their commercial efforts has been the lack of a direct sales force. For a long time their CEO Alex Karp was personally involved in their sales process, not a scalable model. Palantir is changing that and finally scaling their sales engines. The company has been increasing its direct sales force, while also expanding its distribution channels thru new partnerships with system integrators like IBM, and cloud providers. These investments should increase their commercial reach considerably.
Palantir now has some jaw dropping value creation numbers from multiple reference accounts that have reached the scale phase. For example, imagine being the aviation sales executive at Palantir with the industry changing work Palantir did with Airbus to help the airplane manufacturer, its suppliers and 100 airline customers prevent maintenance issues, achieve better fuel efficiency, and optimize their team's time. Their first call is to Boeing. And, with all the legitimate hype Palantir's work with Airbus has generated in the aviation industry why wouldn't Boeing pick up the phone?
3. Better mouse trap
"Build a better mousetrap, and the world will beat a path to your door"
- Attributed to American philosopher Ralph Waldo Emersonin in the late 19th century
There have been a lot of tech companies that have promised to deliver game changing insights from the data exhaust organizations are spewing every second as part of their daily operations. There were great riches to be unearthed. The value proposition of these data and analytics tools, to help find the needle in a haystack, has gone unfulfilled for decades. That is until Palantir.
I see a lot of similarities between the Palantir product story arc and the impact Google had when it launched in the late 90s. At the time search engines were a dime a dozen. How did Google beat out a dozen competitors in the then nascent search market back in the late 1990s? Competitors such as Excite.com, Lycos, AOL, Go.com, Yahoo, AltaVista, Magellan, Infoseek, Webcrawler, HotBot, Open Text and Ask Jeeves were all left in the dust when Google came out with its search engine that blew the competition away. Google simply had a better way to connect users with the most relevant results, across millions and now trillions of web pages. They had a better mouse trap. And now, Palantir offers a better way for enterprise users to connect with insights buried across millions of terabytes of data.
There are many shorthand's that have been given to help people better grok the complexity of Palantir's software. The central operating system for the modern organization. A platform that re-imagines how people use data. The central decision support infrastructure for any organization. A full stack data and analytics platform. Vertically integrated software solutions that cover the entire data lifecycle.
The explanation that piqued my interest the most was how COO Shyam Sankarn described Palantir as an "aggregator in the Ben Thompson sense of the word" in the company's first ever public investor presentation. I was motivated to better understand what this meant and if it was accurate because successful aggregator businesses at scale realize enormous economic value capture (see Facebook and Google).
Like Google is an aggregator that helps users find what they are looking for across the vastness of the Internet, Palantir is an aggregator that helps enterprise users find and work with all of its constantly growing in scale data assets fragmented across the vastness of the enterprise and its disparate systems. Palantir goes a step further by standardizing the data, ontologizing it (a set of building blocks that map business concepts to the data that describes them allowing everyone in the organization to speak a common language) and keeping it secure.
The next function an aggregator plays is bringing a critical mass of users onto the system. Palantir's Foundry and Gotham platforms are becoming the central place where enterprise users can make operational decisions and collaborate with colleagues on company data. Once critical mass adoption of Palantir is achieved in an organization, the Palantir platform becomes the most logical place to build the next marginal application (much like ads in the above diagram). This creates a network effect where more users attracts more apps and more apps pull in more users, making Palantir extremely sticky in an organization. Palantir provides both a set of pre-packaged solutions for common business processes that customers can implement quickly (e.g., resource optimizer app, case management app) and low code/no code tools so citizen developers can build more custom apps quickly.
So is this aggregator like phenomenon actually occurring with Palantir customers that reach the scale phase of its lifecycle? This is important because it's when platform lock-in occurs.
Based on anecdotal evidence it appears Palantir very much achieves aggregator status within a customer account over time. While Palantir's software starts with the data integration use case, it typically expands to include additional business units, functions (beyond IT), and use cases. Take the examples of two European banks that started using Foundry at different points in the product's maturity.
Bank 1: They moved onto Foundry when it was first released in 2016. They began with one use case case and now have over 70+ use cases live on Foundry. License revenue in the account has scaled in lock step, growing from $3 million per year to $20 million per year.
Bank 2: This bank moved more quickly as they deployed when Foundry was more mature. Bank 2 reached 60 use case in half the time it took bank 1 to get to the same threshold. They now have over 7,000 weekly users spread out across many different parts of the bank.
Other examples of Palantir's sales expansion model working as intended:
A major payment processor started using Palantir for powering its merchant retention initiatives, and now uses it across their entire business including reporting, billing, pricing, fighting fraud and cybersecurity.
An energy super major uses Foundry's advanced simulation engines to automatically identify opportunities to increase production. Palantir is also used downstream by finance and planning teams to evaluate future investment decisions, work plans, and capital allocation plans.
Palantir software is helping Doosan, a South Korean heavy equipment manufacturer, enable a transformation into a data-driven decision making and connected organization by integrating data across the value chain, from product development to production, from sales to quality maintenance and product support.
Each of Palantir's two platforms are vertically integrated software solutions that cover the entire data lifecycle, from raw data in a source system to decisions made at the highest levels of an organization. There are three layers to each solution - a data integration platform, a comprehensive suite of analytical tools, and a platform for building applications that get consumed by business users to drive important outcomes. CEO Alex Karp is fond of saying their two solutions each consist of 40 different products rolled into one.
Both Foundry and Gotham feature a dynamic ontology which transforms structured and unstructured data from across disparate systems into objects and associated properties that represent real concepts in the organization (such as people, organizations, places, documents, and events) and the relationships that connect them. In its simplest sense, once all of the organization's data is brought into Palantir (across ERP, CRM systems, thousands of sensors installed on oil rigs, military vehicles or other assets and systems organizations rely on) through out of the box, point and click data connection, integration and transformation, a digital twin of an organization is now available to power all sorts of high value scenarios. What is a digital twin? It is a digital representation of your operations, process and assets. Why do organization's need a digital twin you ask? Well, they help better predict the future and detect issues before they arise because organizations can run simulations with a digital twin, without having to run them in the physical world which would force downtime.
The ontology is bi-directional, meaning all insights and decisions generated by users in Palantir's software are written back into the ontology, immediately available for others to benefit from and build upon. It's like operating the company from inside a video game with colleagues, collaborating on ways to make the organization better. Because it's just a video game (read digital twin), if the players make a move that ends the game, that's ok it's just a simulation. And, you get the cheat codes to make it work in the real world.
Because we've discussed the data platform capabilities of Palantir's software a fair bit so far, let's go deeper on the product capabilities further up the stack or as Palantir describes it "everything north of ontology".
Since the launch of Foundry in 2017 significant new capabilities have been added each year:
ERP Suite: A turnkey solution optimized to run on Amazon Web Services (AWS) that enables rapid data integration across various ERP systems through Foundry based data connectors. The ERP Suite helps customers get more value out of their ERP investments by freeing the valuable data stored in it. The suite has been adopted by a number of enterprises, including bp which reported $50 million in working capital cost savings within two weeks of implementing.
Simulation: Foundry includes a simulation engine that allows business users to test decision before making them in the real world.
Low-code/No-code: Palantir has introduced new capabilities to enable just about anyone to easily and quickly create new applications on top of its platforms.
Once a government customer gets its data in Gotham, whether its data from the ever expanding set of sensors deployed across military hardware or government infrastructure, weather data, flood information, census data, Covid-19 and medical data and countless other sources, the platform offers some high end analytics capabilities and tools that help automate or shorten the time it takes to get to valuable insights. On top of the core platform there are a number of modules that unlock new workflows, insights and intelligence. This includes an investigative assistant that runs around the clock to surface insights and alerts from data that no human would ever be able to crunch.
One of the really cool modules of Gotham is an operations module that provides shared, live maps to track what's happening on the ground in real time so that everyone acts on the same information. Users can drag and drop objects from other Palantir Gotham Modules and applications directly into the Operations module so that planning and intelligence gathering happens in concert with analysis. The power of this is hard to describe and imagine. It's like the sort of software CIA agents or military officers use in movies like Enemy of the State (starring Will Smith and Gene Hackman), on steroids. And it's all available today, this is not military fiction. If you are big fan of these sort of spy thriller movies this Palantir produced demo reel for Gotham is for you.
There is a lot to get excited about with Gotham. Palantir's investments in Gotham is making it it the "substrate" the next war will be fought on. In May 2021 Palantir signed a two year, $111M contract with the United States Special Operations Command (USSOCOM) to provide it with an AI-enabled mission command platform. It is is designed to connect data scientists with battlefield decision makers, to enable collaboration for finding insights and auto-generating courses of action. Watch COO Shyam Sankar explain it. Makes you excited about what's to come and helps illustrate what a good salesman Shyam is.
Palantir's third platform, Apollo, operates behind the scenes. It's the infrastructure that powers both Foundry and Gotham. It is a continuous delivery platform that manages and deploys Foundry and Gotham anywhere customers need Palantir's software to run - across multiple cloud structures, on premise, or classified networks. It's enabled Palantir's move into Software-As-A-Service (SaaS) and brought the compelling economics of SaaS to any customer situation or environment without having to make any security or privacy trade-offs. Remember, Palantir started with the world's most demanding security customer - the U.S government. So Apollo has allowed them to scale and continues to drive new efficiencies for Palantir and its customers in terms of improving gross margins and time to value.
source: Palantir Technologies
Because Foundry has relatively low penetration for reasons discussed above and because Gotham is used by government customers in some pretty sensitive use cases, there is little user or analyst feedback on their software. Because Palantir software does not fit into existing categories or "magic quadrants" due to its breadth and uniqueness, analysts struggle to compare it with competitors. I have come across a very insightful user review of the power of Foundry on Reddit, and there are also published quotes from the most senior positions in their accounts - this level of CEO and defense secretary endorsement does not come easily, it must be earned.
We're not trying to see the future. We're trying to create the future. And I can't imagine a better partner to be doing that with than Palantir.
- CEO BP
“[PALANTIR] CAME UP WITH GROUNDBREAKING TECHNOLOGIES THAT HELP US MAKE BETTER DECISIONS IN COMBAT ZONES...YOU ARE GIVING US ADVANTAGES RIGHT NOW THAT WE NEED.”
- General James N. Mattis, Former Secretary of Defense
A lot of Wall Street analysts are not giving Palantir credit for the amount of optionality the company has. The big knock on Palantir is that they are just a government vendor and as such are constrained by the inherent ceiling on public sector technology spend. Not only is this perception untrue (Palantir has a large commercial business today), Palantir has its sights set on capturing entire industries and not just individual commercial organizations. In addition to pursuing the economic value of being an entire industry platform, Palantir is pursuing multiple, new segments including AI, digital twins, no-code/low code, and the mid-market segment.
For what it's worth, my gut feeling tells me Palantir is one of those companies that will find or create new opportunities to grow its business that we could never anticipate today.
The Palantir and Airbus partnership that built Skywise, an open data platform for the aviation industry, is a great example of a new type of emerging market for Palantir. In the case of Skywise, Palantir was able to stitch together various disparate components and data silos across the aviation industry. Through this unique go-to-market strategy Palantir has become a data platform standard for the aviation industry. Each of the 100 airlines and dozens of suppliers on the platform are either an existing or potential Palantir customer.
Enabling the digital transformation of entire industries with Foundry, like they did with the Skywise initiative, is a great playbook and one they are demonstrating they can duplicate. In a recent announcement Palantir revealed they are working with one of Japan's largest insurance providers Sompo Holdings and a soon to be public company Palantir has invested in called Wejo, to transform Japan's insurance industry. In their S-1 Palantir disclosed they are similarly working with customers and partners in the healthcare, automotive, and government sectors, to establish their platform as the central operating system for these industries.
This concept is an application of Ben Thompson's aggregator theory at an industry level. Instead of Palantir bringing together data across an organization's disparate systems, they are bringing together related data from across an entire industry. Instead of Foundry acting as an aggregation point for a single organization's users, it is attracting users from the entire industry. Greater value is created for all industry participants to share as a result of an entire industry supply chain collaborating together on an industry holistic data ontology in Foundry. In the case of Skywise, since all airlines are sharing their in-flight, engineering, and operations data in a secure ecosystem Airbus is able to offer more value to its customers. Whether this a more accurate predictive maintenance AI model or the opportunity for airlines to benchmark themselves against the entire industry on key performance indicators. This is all done using anonymized data in secure environments so confidential data is not being exposed to competitors.
I have observed three approaches that Palantir is taking to these massive industry platform initiatives, including a) working with incumbents like Airbus in the aviation sector, b) betting on emerging companies that have been built to disrupt entire industries, or c) a hybrid of the two approaches like the Japanese insurance example where they worked with an incumbent and a new, disruptive company.
These new, emerging companies that Palantir is not just partnering with but also investing in thru SPACs (special purpose acquisition company) offers some interesting optionality for Palantir investors. Over the past year Palantir has deployed substantial capital in the form of investments in six different SPACs. These are all public, or soon to be public, companies that are looking to use data to disrupt entire industries.
One of those SPACs, Wejo, offers a good illustration of the motivation and potential for these unique strategic investments. Along with General Motors, Palantir invested in a round a few months ago for a combined $100 million. With additional investments from Microsoft and others, the company will be going public in Q3 of 2021. Wejo is a very ambitious company with plans to "transform connected vehicle data into data for good, revolutionizing the way we live, work and travel". Wejo is onboarding and standardizing over 14 billion unique data points daily from live connected vehicles onto its proprietary data exchange platform. This data is being packaged up into insight and intelligence type products (e.g., traffic management, advertising, fleet management, usage based insurance, remote diagnostics, car sharing, etc.) to enable safer roads, less congestion, and reduced emissions. You can imagine how Wejo is relying heavily on Palantir for its data ingestion and interpretation. With their investment in Wejo and the other SPACs they've invested in, Palantir is not only creating long-term funded demand for its platform, it's adding a new distribution channel and providing investor upside if Wejo stock appreciates once it becomes publics.
Below are some of the companies Palantir has taken an equity position in. It is expected Palantir will add more SPACs on their balance sheet as they identify new innovative companies poised to disrupt entire industries and stacked with visionary leadership to bet on.
With the industry transformation Palantir is enabling with their platform, they are increasing the value created for the entire industry and therefore increasing the amount of value they are able to extract for shareholders. The alternative, as many software companies operate today, would be to work with a small handful of companies in the industry in a disconnected manner. Palantir's method is to work with ALL of the important players in an entire industry and expand the pie for all, while taking a healthy slice for themselves as a result.
- The Wealthy Owl
AI and AI at the edge
The McKinsey Global Institute in 2018 forecasted that AI would deliver $13 trillion of incremental economic activity by 2030. Very few technologies throughout history have had this level of societal and economic impact. We're talking the Internet, electricity, steam engine and printing press level impact. McKinsey's forecasts call for an additional 1.2 percent of additional GDP growth per year through to 2030. As a comparison, the steam engine's boost to productivity of 0.3 percent per year between 1850 and 1910, and IT in the 2000s delivered 0.6 percent.
Data is the fuel that AI needs to run. AI uses machine learning (ML) to get smarter. The learning comes from processing data. The more data the AI model is fed, the smarter it gets. Palantir specializes in helping organizations bring all of their data into a single platform, and so it goes without saying AI built on top of Palantir's data integration platform is likely to offer greater volumes of data to train the AI and become smarter.
While Palantir does offer customers important infrastructure to make AI/ML operational, I see a lot of upside to increase their capture of the projected $13 trillion AI market. This could include adding more AI services, beyond ML, such as natural language processing, computer vision and robotic process automation. Palantir can also offer customers more AI driven applications for common business processes such as an Anti-money laundering app that uses machine learning to detect financial crimes, or an AI predictive maintenance solution. Palantir is already offering these (and more) ready to implement AI applications as part of a use case catalog they offer to customers today. They appear to have a real advantage over AI vendors like C3.AI given the volumes of industry data getting stored in Palantir's platform. As Palantir wins more industry platform implementations they will have more data than any other competitor and therefore a smarter AI model.
Another exciting AI frontier that Palantir is just starting to make inroads in is their new Palantir Edge AI product. Palantir has created a suite of capabilities that make is easier to deploy and operate sophisticated AI models to edge computing. This means bringing AI smarts to vehicles, equipment and infrastructure such as drones, cars, buildings, oil rigs, aircraft, ships, wind turbines, robots, and satellites. New scenarios are requiring the application of machine learning models that empower equipment and infrastructure to make decisions throughout the course of their operation. This model is called machine learning inference. A trained algorithm gets deployed to an IoT device to make predictions, based on data generated at this endpoint, to guide decision logic applied by the device. A good use case example of where machine learning inference will be required is autonomous cars. Because self-driving cars need to make split second life or death decisions based on data the car is collecting in real time (e.g., the behavior of other cars on the road, hazards on the road, etc.), self-driving cars need the brains to make these decisions. They need an algorithm that's been trained extensively across millions of scenarios a self-driving car could face. The level of compute and data required to train this self-driving algorithm makes the central cloud (e.g., Amazon web services) the only feasible location today for this processing. It's not practical due to the latency inherent with the central cloud. Once the algorithm is trained to a suitable level, it can be used by autonomous vehicles to quickly infer and make the sort of local decisions that are necessary to keep their passengers safe. The real time nature of decision making required to make self-driving cars a reality requires an AI based edge platform so car don't waste milliseconds trying to communicate with a central cloud.
The U.S. government is already a Palantir Edge AI customer. What's exciting is how Palantir Edge AI has the potential to bust open the commercial market. These same AI and edge computing innovations apply to the commercial sector, including AI models on factory floors directing robots, enabling predictive maintenance, or better power utilization across the electricity grid.
No code/low code
I see a lot of opportunity for seat expansion in existing accounts with the recent addition of no code/low code capabilities to Palantir's platforms. As Palantir hides the complexity of building apps on its platforms, more non-technical business users will be able to get value out of Palantir and thus organizations will have a business case to add more users. More developers (even if they are non-technical) means more apps being built on the platform. More apps will attract more users and parts of the organizations, which will generate more apps to reach this ever expanding pool of users. And so the flywheel spins.
Palantir is tapping into a broader trend of citizen developer (non-technical users) and low code/no code platform growth. Gartner estimates that by 2024 75% of large enterprises will be using at least four low-code development tools for both IT application development and citizen development initiatives.
Early stage companies and mid-market: Palantir for Founders
Palantir has started to move downstream into the mid-market with their recent launch of Palantir for Founders. With Foundry for Builders, Palantir will sell Foundry to start-ups under a subscription model, first to companies connected to Palantir alumni, before expanding to other early stage companies. I expect Palantir to expand into mid-market beyond early stage companies thru future distribution partnership, much like they have forged with IBM to increase their reach in enterprise.
5. Room to grow
At a current market capitalization of ~$42 billion (as of July 30th 2021) and Price/Sales ratio of 35 you may legitimately wonder how much room does Palantir have to grow. Considering the amount of scaling Palantir can do, both in terms of revenue growth and unit economics, I see a lot of upside for Palantir and its long term investors.
Palantir grew revenue 47% in 2020. It's optionality, expected new distribution partners, further penetration in commercial due to increased awareness of Palantir and a growing sales team, plus recent wins entering the scale phase of their sales process are all tailwinds that I believe will accelerate revenue growth.
Palantir has a lot of room to improve its margins by engineering out some of the complexity today in its deployment process. The promise of its Apollo platform is to deliver scale economics by making their implementation engineers more productive, increasing automation and enabling greater customer self-sufficiency. Evidence of the impact Apollo is having on Palantir's unit economics is quite compelling as you can see the improvements in their contribution margins in 2020 across all phases of their sales process.
Margin improvements continued in their latest reported quarterly earnings showing the sustained impact Apollo is having. We are witnessing Palantir's successful transition to SaaS economics.
As a gut test I always like to ask my self how big a company can grow to from its current size. For example, at a current market cap of ~ $1.7 trillion, can I see Amazon approaching $10 trillion in the next 5 or 10 years? Probably not. In the case of Palantir, could I see them growing from $42 billion to $420 billion in the same time period? Palantir offers its customers game changing software. It holds the promise of being the most impactful software company to come along in decades. So, yes, if Palantir can scale successfully I would not be surprised if it were to be a 5 or 10 bagger from today's prices. This is not a prediction. I could be wrong. I just think it's more plausible than Amazon reaching a $10 trillion valuation.
6. Strong leadership
Given the historical secrecy surrounding Palantir, one of the critical activities for its leadership team has been to demystify the mystique and skepticism surrounding the company and evangelizing the power of its software. This was not only a pre-requisites for its recent IPO but as Palantir began moving into the commercial market it has had to explain and prove its relevancy to large company's. People don't understand enough about what Palantir does. Evangelism is possibly the most important use of their time that Palantir leadership can make at this stage.
While Palantir has stacked its leadership ranks with various industry veterans, including a former public health official and senior military officials, there has been no better evangelist for Palantir than is charismatic CEO. Alex Karp's unique personality has been well documented. Words like eccentric, genius, arrogant, unorthodox have all been used to describe Karp. If you want to get a sense of his personality, you gotta check out this one on one interview a Palantir employee recently conducted with Karp. The video is long, but even a couple minutes will give you a sense of Karp in his element talking with a Palantir employee.
The ying to Karp's yang is Palantir's COO Shyam Sankar. While Karp has been travelling the world as a one man Palantir sales machine, Shyam has been busy putting the technology, infrastructure and process to help scale Palantir in place. Shyam joined Palantir 15 years ago as a bit of a wonderkid at the time. The world really started taking notice of his talents in 2013 when CNN named him one of the world's top 10 thinkers in 2013.
Watching Shyam in various presentations, along with the product managers and technologists who have been given a lot of the spotlight lately, it's clear that Palantir has a very deep leadership bench.
The best way I know to get some insight on the quality of leadership at a company is to go to the source. What are the employees saying about the leader and company. Below are employee provided reviews of Palantir from Glassdoor. While the ratings in aggregate are good, I was a little surprised they weren't a bit higher given all the hype surrounding Karp. Having said this, Palantir is a very intense and demanding company and so that is bound to rub some people the wrong way.