Palantir: A SaaS Company (Part 2)

Pivoting and succeeding in a landscape filled with competition

Last week we covered Palantir’s journey breaking through the political process, lobbying the government, changing legislation, and paving the way for govtech as an industry. Because of them, several US government agencies have started taking big bets on startups.

Palantir’s ride to the top is nothing short of phenomenal - but let’s take a closer look at some of the partnerships they had during this growth. The startup started out with the public sector but even took some private sector contracts during this time. You’ll start to notice a pretty obvious trend here.

Early on, Palantir was the X-Factor during the War On Terror, working with the Pentagon and CIA in Iraq and other areas in the Middle East. When the conflict in the Middle East started, the US Military used various drones and balloons to gather tons of surveillance footage. This raw data is then washed through Palantir to help avoid IEDs and identify terrorist hotspots.

Non-conflict uses through the government also existed. Palantir has worked with the Department of Health and Human Services (DHHS) to offer Gotham as a tool to detect Medicare Fraud. By combing through Medicare claims data, Palantir was able to visually map hotspots where doctors were overprescribing and allow the DHHS to crack down on prescription abuse. These efforts were made to tackle a $100 billion fraud market.

JPMorgan infamously used Palantir to quite heavily spy on their employees. Gotham ingested tons of emails, browser histories, GPS locations, printer activity, and transcripts of phone conversations to flag employees for potential abuse of corporate assets and point out which employees were getting disgruntled. 

I think you’re starting to get the gist of it - Palantir has enabled numerous agencies to ingest and analyze amounts of raw data that were previously unimaginable. Sometimes even if it meant toeing the line of ethics or blatantly disregarding it.

All of this sounds awfully familiar right? Isn’t that what Google and Facebook have been doing for years? Contrary to what all this might be indicating, Palantir firmly believes that it does not collect and monetize off data at scale.

The Problem with Gotham

We build digital infrastructure for data-driven operations and decision-making. Palantir’s platforms tie these together by bringing the right data to the people who need it, allowing them to take data-driven decisions, conduct sophisticated analytics, and refine operations through feedback. We license this software to organisations, who receive secure and unique instances of our platforms in which to conduct their own work on their own data.

If you believe what they say, Palantir does not collect data, but rather simply enables those who do to better manipulate the data in ways more useful to their business. 

Palantir’s first product, Gotham, was built around the rather archaic requirements of the government. When pivoting to other private sector use cases, this meant limited flexibility in pricing and structure around the product model.

Here’s how Gotham worked. If you wanted to contract Palantir, you would first need to pay for a server which comes with 1 year of maintenance. Then you’d need to gauge how many people in your organization might need access to Gotham. Let’s say it’s just 10 - maybe you need a 1 core system. That’s around $150k for 1 core before maintenance. For every year after the first, maintenance is $30k. For just 3 years, you’re already paying $200k before software updates, training costs, power, security, IT support staff, etc. All of these can easily double the cost.

There’s one more cost to mention. Operating Gotham requires onsite engineers, called “Forward Deployed Engineers”. These employees are essentially just onsite contractors that work fully on the specific business use case for a particular customer. As part of their contract, the customer is required to pay the $150/hr salary commanded by the FSR.

All of this was just way too expensive for the prospective Palantir customer. Large agencies with obvious use cases could deploy these expenses but enticing new customers was nearly impossible. Even the Street didn’t really love Palantir. The company was back to IPO levels just two years after going public. With Gotham as the start of the show, Palantir was not expected to make any money unless they were able to get massive government and private contracts several times, every year, forever.

To turn heads and create profit, Palantir needed to find a way to offer the same product for a substantial operating margin. Palantir had to create a SaaS product.

Competitive Advantages with Foundry

Welcome to Foundry. No more onsite engineers, core server costs, maintenance and software update costs, nothing. At its core, Foundry is a data ingestion and integration platform that was designed with cloud infrastructure in mind. As a data engineer, you are able to use a web platform to link multiple data sources to better use unstructured data.

Ok, if you put a magnifying glass to Foundry, you might be reminded of competitors like Snowflake, BigQuery, Databricks, Azure, and other data aggregators. These companies allow you to ingest multi-platform data into one data lake and get a holistic view what you want. But let me advocate for Palantir here. Palantir does not want to be seen as a replacement for Snowflake or Amazon or Microsoft. Similar to how they built Gotham, Foundry also has a “sky compute” engine that makes it easier for business level employees (non-technical) to manipulate data sources similar to Tableau.

Palantir has been in the AI space long before it was seen as cool. Similar to NVIDIA handling the hardware infrastructure that’s powering AI, Palantir has developed the software infrastructure for the past 20 years at this point. In 2019, Palantir won a bidding war for a US Army contract to create a dashboard that would pull together data from human resources. You know who they were competing against - Microsoft, Deloitte, Accenture, and Ernst & Young. Isn’t that crazy? Just 5 years ago, they were bidding against companies that you would never consider software business - EY is literally an accounting firm!

It’s impossible to project where the data aggregator industry is headed in the next. In 2021, the total addressable market was already $119 billion with the potential of doubling in the next few years because of developments in AI. Palantir, Databricks, and Snowflake - all leaders in the industry - all sit at a $45 billion valuation. Even without killing the competition, Palantir can thrive simply by growing its business. And if they can figure out how to further pivot away from Gotham, they can once again pave the way as the leading company for data mobilization

That’s all for this one folks! If you’re interested in more content around the data cloud, let me know by liking this post or dropping a comment. If you enjoyed reading, please feel free to share with your friends. Reach out to us by replying to this email if you have any insider takes or leave a comment on our Substack page!

Thumbnail from Medium