We published an overview of the Palantir (PLTR) basics already so see Parsing Palantir – Part I for that if you want more background.

With the direct listing coming today (Wednesday, September 30th) it’s time to circle the wagons and look at the facets of the company and the stock and develop a feel for where the pricing may settle out.

Snowflake (SNOW) is also in the “data” space and came public in a resoundingly hot deal. Snowflake is the epitome of a Silicon Valley technology company. Palantir is very different in terms of their culture, their business model, and the customers they serve.


To explain what’s unique about Palantir is that it’s an *operational* system *based on data* rather than a data-based system. So what does that mean? It boils down to three functional elements wrapped in their “high touch” delivery model:

  1. At the center of a Palantir deployment is a large-scale model of the organization. This defines key objects, their properties and relationships. Palantir can do much of this automatically based on original source data and show their work.  That last part is a big deal. It’s now easier to build a model from existing data than think one up. Using technology to do it puts the heavy lifting on technology and frees the user to wield it instead of building it.
  2. Without bringing in too much computer science here Palantir allows clients to make changes and redefine this model as they need to. It is more dynamic and fluid than what you might be able to set up with traditional databases or enterprise applications like SAP in manufacturing or EPIC in healthcare.
  3. Security, version control, and full-offline deployment are core to the offering and not partially implemented afterthoughts. These are obviously critical for applications in government and intelligence domains but they are also important to enterprise customers. The ability to deploy systems for offline use allows air-gapped security along with synced operations. The fact that these things are built-in at Palantir makes them different.

I’m not sure that I like the term “OS” for what they do for an organization but they do turn data and information analysis into a more operational system. This diagram captures the idea that there is a dynamic model at the center that serves different purposes – review, analysis, reporting, extraction, and discovery.


Critics of the company and the deal tend to surface these concerns:

  1. They are more like a services company like Accenture (ACN) than a software company. This is partly historical because the early days of the “product” were indeed more custom than off the shelf. They still deploy engineering talent to the front lines working directly with customers. This doesn’t mean they are a product company but they are closer to the old IBM model of ensuring customer success.
  2. Their heavy reliance on government business and “anti-Valley” sentiment has cost them some access to engineering talent. The increasingly divided political environment in the U.S. and the world may also limit their market opportunity and ability to grow.
  3. They are losing tons of money and are likely to need to raise substantial capital in the future that could be dilutive to highly dilutive to shareholders depending on the future stock price.
  4. Their business model doesn’t “fit the times” and is a bespoke, long-sales cycle, high priced model rather than the ubiquitous “land and expand” model of so many SaaS and usage-based models.

Stock considerations

Having said all that it’s clear that retail demand for Palantir is high – it’s the #1 viewed and voted on IPO for this week. It’s likely that investors will “look through” the stunning losses the company is making and not think much if at all about valuation.

We’ll provide some comparables here anyway because we’re old-fashioned that way – feels rather quaint in this market.

The comps are skewed by SNOW which is still basking in the glow of IPO ardor.

Looking at more sensible comparables like ESTC, PLAN, and CRWD and using a round number of $1B in current revenues the starting valuation of PLTR at $11.50 isn’t terrible. That puts it at 20-25x sales. I’d also say that PLTR has more of a deep technology advantage than companies like SNOW so at these prices would prefer it.



Sign Up!

Get pure IPO Goodness - no ads, no fluff, no SPAM.

Powered by ConvertKit

Leave a Reply