Why Palantir Makes My Head Hurt

While I’ve blogged before about Palantir Technologies (e.g., Beware the Spectacular B-Round Valuation), this will probably be my last post about them because, since leaving MarkLogic, I am no longer terribly involved in the Intelligence Community space nor engaged against them as an indirect competitor.

I initially became interested in Palantir for several reasons:

Part of the marketing was making controversial claims, such as:

  1. We have no sales.  (e.g., at minute 5:40)
  2. We have no marketing.
  3. We have no services.  (Our field technical staff aren’t consultants, they are forward-deployed engineers.)
  4. Positioning as a billion-dollar company when sales were probably in the tens of millions.
  5. Talking about valuation on funding rounds.

Now, as a credibility-is-key marketer, these kinds of claims bug me at two levels:  first, that people would make them and second, that the media would report them.  Here’s my take on these 5 claims:

  1. Whether you want to call it sales or not, someone meets customers, talks about what your software does, discusses how to price it, negotiates and signs a contract.  In the normal world, that is called sales.
  2. Whether you want to call it marketing or not, someone made the website, spent money to sponsor a party, setup the Charlie Rose interview, and designed and paid for the DC subway ads.  In the normal world, that is called marketing.
  3. Whether you want to call it services or forward-deployed engineering, you are sending smart people with engineering and computer science degrees to customers’ sites and helping them solve problems using your software.  In the normal world, those tasks are called pre-sales engineering and consulting, depending on whether they happen before or after a sale.
  4. The standard way, in the real world, to refer to a company’s size is by revenue.  The one and only time I frequently heard people referring to company size by market capitalization (or valuation) was during the Internet bubble.
  5. While this is primarily a style issue, most companies do not disclose valuation on venture funding rounds.  I believe those who do are trying to generate hype.  (And for a company that insists it has no marketing to want to generate hype is doubly irritating.)

At the big picture level, Palantir reminded me of MicroStrategy:  big claims and hype, DC-centricity,  elite school hiring focus, youth focus, a large field technical team, and a work hard / party hard ethos.

At this point I should admit to having some scars from having run marketing at Business Objects during MicroStrategy’s rise.  Let demonstrate what a day in life looked like:

  • Dave, MicroStrategy says their mission is to “purge ignorance from the planet.”  How come we can’t say anything visionary like that in our mission?
  • Dave, Michael Saylor says he’s going to build a modern-day Versailles just outside of DC.  How come our CEO never says stuff like that?
  • Dave, MicroStrategy says they’re building a service where people will wear tiny microphones in their ears and it will notify them if their house catches fire.  How come we don’t have product vision like that?
  • Dave, MicroStrategy just did a $52.5M deal in an industry where average sales prices are $250K and a big deal is a few million.  Why can’t we do huge deals like that?
  • Dave, Michael Saylor says that there will be riots if his software doesn’t work and that this year people will die — literally — because they didn’t buy his software.  How come we’re not mission-critical like that?

To which for several years I had to say “it’s all bullshit, it’s all bullshit, it’s a barter transaction and they’re double counting, and it’s all bullshit.”

It turns out being a naysayer isn’t fun work:  for three years you sound like a whining, doubting-Thomas constantly on the back foot, constantly playing defense and then one day you’re proven right.  But there’s no joy in it.  And the naysaying doesn’t help sell newspapers so you don’t get much press coverage.  And, in the end, all people remember is that “MicroStrategy was pretty cool back in the day” and “Dave’s a grump.”

It was during this period that I got interested in Corporate Cults because MicroStrategy struck me as one.

  • Hire young people with similar profiles from the best schools (e.g., MIT)
  • Work them long hours
  • Isolate them from friends and family
  • Blur distinctions between work life and personal life (e.g., company cruise, work hard / play together)
  • Tell them they’re the best
  • Tell them naysayers don’t get it

Six steps to make MIT engineers cult members.  Thus, in addition to other MicroStrategy parallels, Palantir struck me as a corporate cult.  Kind of a Logan’s Run (where no one is over 30) meets the Apple 1984 commercial (conformism à la the black jackets).

Since I left MarkLogic in January, Palantir got tangled up in the HBGary WikiLeaks mess (proposal here), generated some positive press in Forbes, and raised a $60M round of financing at a valuation rumored to be as high as $3B, bringing the total invested capital to an estimated $175M, a lot of money for an enterprise software company.

This begs the perennial question of “if they’re doing so well, then why do they need so much cash?”  While there are potentially both good and bad answers to that question, my guess is the answer is roughly:

  • Because they can raise it at huge valuations for relatively little dilution.  (Peter Thiel may be a huge help on this front.)
  • Because they intend to spend it to continue hiring and perpetuate the lavish-spending culture and hype machine.
  • Because they are executing a go-big or go-home strategy that is cash intensive and will, they hope, result in a huge exit valuation.

But why does Palantir make my head hurt?

  • Because, despite my general skepticism, I believe they get some things very right.
  • Because, despite their intent, they may have created a new kind of company.

Because I can be perceived as a Palantir detractor, I’ll say it again:  Palantir gets some things very right.  Which things?

  • They hire brilliant people.  They deliver on the hype in this department.
  • They solve hard problems.  I hear customers are generally quite happy.
  • They solve the whole problem.  They don’t just drop software in your driveway and run away.
  • They aren’t afraid to ask for huge checks, order of magnitude in the tens of millions.

Personally, I don’t buy the argument that all field technical staff are “forward-deployed engineers” as opposed to pre- or post-sales consultants.  But I would believe that you can hire better people by telling them they’re engineers as opposed to pre-sales consultants.  And, I could even believe that someone could convince himself — if perhaps not his accountants — that field technical staff are not customizing an application but instead developing a product.

That last point is important.  Why?

  • If field technical staff are engineers, then the associated revenue is presumably license fees and the cost is R&D.
  • If field technical staff are consultants, then the associated revenue is services and the cost is COGS.

Why does this matter?  Because most software company boards and investors see the world in a pretty black-and-white way:

  • License revenue is good.  Services revenue is bad.  (Largely because gross margins run 98% on the former and 20-30% on the latter).
  • R&D expense is investment and ergo good.  Cost of goods sold is bad.

Almost all Silicon Valley boards will want an emerging enterprise software company to run with a consulting business that’s no more than about 20% of total sales.  In practice this means a company can have at most about 1.5 consultants (pre- and post-sales) per salesperson.  Any work that can’t be done either as R&D investment or by that small consulting team needs to get handed off to partners.  This means the vendor loses control over customer success (which customers don’t like) and the vendor doesn’t end up owning all the IP required to solve the whole problem.

Now, my guess is that Palantir’s board doesn’t care about any of the preceding four paragraphs, probably because of cult arrogance:  we don’t care what pedestrian accountants say because we are changing the world and building the ultimate set of products.  Accounting, schmaccoutning.

This works well as a private company, particularly if you don’t plan on going public.  But the constraint on consulting growth hamstrings most enterprise software companies forcing them into a component-orientation, a drive-by license sales model, and a disregard for customer success — the traditional negatives that helped the drive the SaaS movement.

But, regardless of the reason, Palantir is a different type of company.

  • Like a system integrator (SI), they have a small sales force, a large field technical staff, solve whole problems, and ask for big checks.
  • Like a software company, they hire world-class engineers and try to capture everything in product.

Is Palantir an enterprise software company with no sales, marketing, or services (as they would like to believe) or are they the first SI to figure out how to build a world-class software business as most SI’s aspire?

You can argue the difference is just semantics, but I’d argue the latter.

28 responses to “Why Palantir Makes My Head Hurt

  1. Dave, a terrific post and best synopsis of Palantir I have read ;)

  2. Hey Dave,

    So many good things about this post; but you had me with “eg, butter/olive oil,” on so many levels. Aside from the obvious distinction that olive oil tastes good on salad and butter does not, one of the downsides of capturing everything in product is the danger of putting Extra Virgin Olive Oil (EVOO) on *everything* because it is trendy or some BS, when what you really want melting in your skillet is half and half butter and a less expensive second pressing.

    As far as I can tell, the problem with SI, anyway you slice the bread, is that you *can* keep on pressing until you’re squeezing water out of your presses, but the taste is terrible – you still need supra-genius engineers to produce stellar results. I am wondering if there is a typo here:

    >>> Is Palantir an enterprise software company with no sales, marketing, or services (as they would like to believe) or are they the first SI to figure out how to build a world-class software business as *most* SI’s aspire?

    I think you meant “all”…

  3. You have really hit the nail on the head. I’ve been to one of their government events (at the Ritz in McLean), and they do seem to spend like sailors.

    I also agree that the “no sales people” and “no PS” is BS – I know for a fact that they have sales guys, and you’ve already talked about the “forward-deployed engineers.” A lot of their customers in the government market believe they’re arrogant, too – it’s not just their competitors. :-)

    One other downside to the “hire ’em young and work ’em hard” culture you didn’t mention was that, due to hubris or inexperience, they do tend to needlessly reinvent the wheel. I attended a technical session on their new distributed database backend, which they were proud to point out had been built by four engineers during their “hack week.” I don’t know about you, but if I’ve got enough data to worry about sharding over multiple machines, I’m not sure how good it would make me feel to know that such an important subsystem had been “hacked” together.

    All that said, they are doing something right – their deployments by and large do work, which is saying a lot.

  4. So to say, palantir tech seems to be co., with die hard engineers who work earnestly for the growth of the company

  5. Dave, you crack me up! Joe

  6. Here is an interesting article a reader sent me about both lobbying and large-deal hunting which I believe supports my next-generation SI hypothesis. http://www.politico.com/news/stories/0611/58051.html

  7. Pingback: With $5 Million In Their Pockets, Ness Has Quietly Built A Subjective Search Engine | Internet Billboards

  8. This sounds awfully like the Trilogy of the yore… 90 developers, 1200 consultants, $2B valuation at it’s zenith, $54M deals, work-hard-play-hard, beating MSFT in college recruiting, and on and on and on…. eerily similar!

  9. Dave – great post – totally like Trilogy in many ways.


  10. Excellent post. Palantir has created a culture out of hiring only the best in the same way that Google achieved, many years prior. It’s a great lesson for startups. However, I met Stepen Cohen, co-founder of Palantir, this weekend and he did say that hiring super-bright engineers comes with a price: You need to keep them engaged with extreme challenges or they’ll get bored. As such, Palantir has lots of “hacked” projects to demonstrate abilities.

    PS: I’ve studied Microstrategy and BusinessObjects case studies for launching a new database company. It would be awesome to chat with sometime about your views on the future of disruptive enterprise software.

  11. it’s sad but you just don’t understand how the new startup model works. nothing about what they are doing sounds strange to me. this is the new generation model.

    • Given the anonymity of your email I’m not sure I should bother responding but just in case others have this viewpoint I have a one-word answer: wrong.

      It may well work for Palantir (or they end up a smoking crater) but there is nothing new about the model except two things: the enormous capital consumption, the SI-hybridization. The point of the post was educate on that very fact — see MicroStrategy, see Trilogy.

  12. Sounds like early days at TIBCO.

  13. Pingback: pmworks change: we can now test our Evolutionary Vision Model | pmworks:blog

  14. Pingback: Dan Shapiro » Companies that would do best without venture capital

  15. Nice article, Dave.

    Not sure if you noticed that your word processor auto corrected MicroStrategy to Michael Saylor a couple of times.


  16. Pingback: Companies that would do best without venture capital « Blog of Intellectual Capital

  17. Wow. That was a great post! Thanks for the insight…it will help. Amazing that the company / brand isn’t as well-known in the DC market unless you are in the same tech circles or IC. I’d still love to see a follow-up post…the playing field looks like it’s changing again at least here in DC.

  18. Pingback: What is the most underhyped startup in Silicon Valley today? Why? - Quora

  19. So, would you work for or with Palantir?

    • In short, no. While the lines between cult and culture and hype machine and great marketing machine can by blurry and are certainly subjective, for me, Palantir falls on the wrong side of those lines. If you’re considering working there, I’d ask a lot about the job and pay little attention to the name — often these types of companies dress-up mundane jobs with big names and sell (often naive / graduating) employees on a role that is actually quite a bit less sexy than it sounds. As a customer, I might work with them because, as noted in the post, they do seem to be a very high-end SI — the only question I’d have as a customer is viability. To the extent they are spending a fortune and raising a lot of money on premise 1, and to the extent there is a lot of hype built into premise 1, then what will happen to the company is premise turns out to be false.

      Finally, the bloom does appear to be off-the-rose as I’ve heard little from the hype machine of late, which for a hype-y company, can be an indicator of troubles. Or perhaps a new leaf.

  20. Amazing post Dave!

  21. Pingback: Pros and Cons of Palantir? : Beyond Search

  22. It’s not particularly useful to draw a comparison with Saylor. You certainly weren’t the only critic – at the time, he was a caricature of himself, and made grand pronouncements that Palantir, or any sane person, would never make. Palantir’s reasons for success are more concrete, and they certainly don’t hype about being the next GE or as great as Thomas Edison.

    The characteristics of these companies that you and others posit are the signs of ‘tragic doom’ (a la Trilogy, MicroStrategy, etc.) are no different at SV’s successful tech companies. And I would agree with the above comment that the model of startups in SV is very different from before. You see it at Stanford, you see it among investors, you see it in the media; sure, some of the characteristics (work-hard-play-hard, etc.) aren’t novel, but they are intensified. Why? Because it delivers results. Palantir must be doing something right.

    I’d also add that neither MicroStrategy nor Trilogy have gone out of business. Sure, they peaked at the height of the dot-com bubble, but it’s really not impressive to criticize them now, given that they were a product of the hype that everyone recognizes. Old news, you know. MicroStrategy in particular has over $1 billion market cap today – a far cry from the bubble, but nothing to sneeze at, and it’s steadily climbing. Who knows? Perhaps in 10 years they’ll be competing with Oracle et al

  23. It is a great post. I have a couple of points that perhaps you could further elaborate.

    You wrote: Any work that can’t be done either as R&D investment or by that small consulting team needs to get handed off to partners.
    Question: Why should that (handing to partners) be necessary? No other alternative?

    You wrote: But the constraint on consulting growth hamstrings most enterprise software companies forcing them into a component-orientation, a drive-by license sales model, and a disregard for customer success — the traditional negatives that helped the drive the SaaS movement.
    Question: Why are they forced to component orientation? And why the anthithesis to customer success?

    Thank you in advance.

  24. Interesting. I felt the same about Palantir – more of a high end SI model then a product company. By the way great product engineers don’t necessarily want to be Forward Deployment Engineers so it is tough to hire laterally – freshers are more receptive to their marketing. I felt they had huge Govt. backing thanks to In-Q-Tel funding. My mantra is “anything that looks too good to be true is not true” Only the real facts are hidden – sort of not adding up.

  25. Pingback: Why consumers run the enterprise -

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.