Forums  > General  > palantir  
     
Page 1 of 1
Display using:  

akimon


Total Posts: 566
Joined: Dec 2004
 
Posted: 2016-05-11 11:07
Inside Palantir, Silicon Valley’s Most Secretive Company

Perhaps the 3rd party data analytics bubble may start to burst?

I think in the future, companies will increasingly perform data analytics in-house, using open-source technologies and published algorithms.

Kitno


Total Posts: 349
Joined: Mar 2005
 
Posted: 2016-05-11 11:48
1. Their burn rate is jaw-dropping. I don't think I've ever seen one that high (for their stage of development) if Wikipedia is to believed. They've been raising a few hundred million or 'units' each year since 2011. Each raise has sought a higher valuation. I separately read they had an estimated $250MM of revenues in 2011.

2. "On April 22, in an extraordinary move for a company that had prided itself on paying salaries below market rate, Palantir CEO Alex Karp announced a 20% pay raise for all employees who had worked there for at least 18 months. Karp also canceled annual performance reviews, saying the current system wasn’t working."

This is start-up 101. You can pay sub-market for a while with equity upside. As soon as you face a downround and consequent dilution you start seeing staff leave and it's a vicious circle.
Suspending appraisals and a 20% payrise is a sign of management desperation. You suspend appraisals because employee targets are not being met - unattainable targets are the first warning of bad management.

High burn rate, failed targets and management weakness - downround incoming. The article mentions high profile client losses. Well that again points to potentially a great product, but it's not delivering what those clients want.

At the end of the day they have a government client backstop but it aint at a unicorn valuation...

Salut toi, je vais au Social Club avec des amis ce soir, c'est au 142 rue Montmartre. J'ai mis ta robe préférée. Viens me trouver.

NIP247


Total Posts: 543
Joined: Feb 2005
 
Posted: 2016-05-11 12:26
Additionally, I've always been wary of business models based on software AND consulting, you generally end up Neither, Nor...

On your straddle, done on the puts, working the calls...

Maggette


Total Posts: 1001
Joined: Jun 2007
 
Posted: 2016-05-11 13:33
IMHO the "software + consulting services" problems appear in cases where software pretends to do stuff that can't be done that easily.

For example the ESB/ bussiness integration stuff. They sell you a software solution...and on the first day they show up withh 100 consultants and 150 developers to actually get their software to work in your specific case.

Any software that pretends to reduce solving complex bussiness and software problems (and analytics is complex) to a simple drag and drop GUI more often than not won't deliver.

Another example of this IMO is this one
https://www.talend.com/products/big-data

Ich kam hierher und sah dich und deine Leute lächeln, und sagte mir: Maggette, scheiss auf den small talk, lass lieber deine Fäuste sprechen...

akimon


Total Posts: 566
Joined: Dec 2004
 
Posted: 2016-05-11 13:48
Any software that pretends to reduce solving complex business and software problems (and analytics is complex) to a simple drag and drop GUI more often than not won't deliver.

Unless it is offered by Google. Wink

Kitno


Total Posts: 349
Joined: Mar 2005
 
Posted: 2016-05-11 13:59
Akimon I was just thinking the same as you posted that. As tech companies go Google is my favourite through their continued innovation and execution. Apple for example on the other hand is mostly focused on ergonomics and 'packaging' which is fine as a sale conduit but less intellectually impressive.

I especially the Google VC maxim: it's worth investigating (potential investment) if people use it at least twice a day.

Salut toi, je vais au Social Club avec des amis ce soir, c'est au 142 rue Montmartre. J'ai mis ta robe préférée. Viens me trouver.

jslade


Total Posts: 1115
Joined: Feb 2007
 
Posted: 2016-05-12 05:08
I don't think anything that has happened to Palantir recently is any different from things that have happened to them over their history. Their burn rate has always been bad, and customers do come and go.

The pay raise; long time coming. DS in the valley is paying absurdly more than Palantir is.

That said, it's a consulting company valued at software company prices. Silly.

"Learning, n. The kind of ignorance distinguishing the studious."

Nonius
Founding Member
Nonius Unbound
Total Posts: 12720
Joined: Mar 2004
 
Posted: 2016-05-12 10:42
You guys see that DataMinr was dropped by Twitter as a partner. From my perspective, I'm not sure I'd advertise to the world that the intelligence agencies are my clients.

Chiral is Tyler Durden

Kitno


Total Posts: 349
Joined: Mar 2005
 
Posted: 2016-05-12 12:07
jslade, about their high burn and payrise/appraisal suspension it points to a common theme seen in 'start-ups'. The necessary entrepreneurial management for product development, build-out and rapid growth is often unable to successfully morph into an established player management. Private Equity will often replace management a few years before IPO, or earlier, for this reason.

The CEO is noted in that article as saying they don't want to IPO because it's not right for his business - too true. The kind of funding they have from ultimately UHNW angels has negated the strictures of PE management scrutiny yet these investors will be asking for improved metrics and EBIT - it's been 12 years since incorporation. The response to employees shows they need to change their management style for their development cycle.

Salut toi, je vais au Social Club avec des amis ce soir, c'est au 142 rue Montmartre. J'ai mis ta robe préférée. Viens me trouver.

ast4


Total Posts: 390
Joined: Aug 2007
 
Posted: 2016-05-12 14:47
I grew up with one of the co-founders of palantir as well as a few people in senior management who they ended up bringing on board. All pretty smart guys.

Akimon, in regards to companies taking things in-house, I think you may be over estimating 1) the ability of the general talent and 2) management/hr's ability to recruit candidates who are actually talented. Running a scaled data analytics/ml operation/team is non-trivial IMO - generally you need to find people who can not only do the math, but actually understand how computers work.

Here's an example of what I'm talking about, a friend of mine who works for a certain tech company was complaining to me about the abysmal state of their "data scientists". Apparently these guys are storing all of their data in CSV format which takes hours to parse. That team's solution? Obviously the most reasonable one... fire up X instances using EC2 to speed up the process. This type of nonsense ends up eating into the bottom line obviously, but beyond that these guys don't even seem to understand how to solve a problem without brute force.

/rant

All this being said, I haven't really kept up with the palantir guys. Last time I checked their biggest client was the US gov't which seems like it should be fairly sticky money. They also attempted to launch a financial product which seemed to fail miserably.

"Mathematicians are machines for turning coffee into theorems!"

tbretagn


Total Posts: 252
Joined: Oct 2004
 
Posted: 2016-05-12 14:58
Regarding the government use (military), they were very happy (ie palantir really helped on the ground operations).

Et meme si ce n'est pas vrai, il faut croire en l'histoire ancienne

akimon


Total Posts: 566
Joined: Dec 2004
 
Posted: 2016-05-12 15:08
ast4,

Thanks for your insights from your knowledge of the situation. I agree with you on your points regarding the ability to attract talent in general for this space.

One of the things on my mind these days is with the exodus of talented people from investment banking and hedge funds. I actually think it might not be difficult for many traditional "quants" working for investment banks to retool themselves to become "data scientists" (man, I kind of dislike this term..) for Fortune 500-type companies who would normally pay a few hundred $k/month for these type of consulting services. There are certainly many elements of skill overlap here, like the ability for quants to work with really unimpressive people from the main business side.

Nonius
Founding Member
Nonius Unbound
Total Posts: 12720
Joined: Mar 2004
 
Posted: 2016-05-12 19:53
" I actually think it might not be difficult for many traditional "quants" working for investment banks to retool themselves to become "data scientists" (man, I kind of dislike this term..) for Fortune 500-type companies who would normally pay a few hundred $k/month for these type of consulting services. There are certainly many elements of skill overlap here, like the ability for quants to work with really unimpress"

I think you might be on to something.

there's an old NPer here who did that a loooong time ago, by the way. Jeff Hammerbacher. Apparently he even coined the term "data scientist".


Chiral is Tyler Durden

Kitno


Total Posts: 349
Joined: Mar 2005
 
Posted: 2016-05-13 01:11
From Feb. It seems the horse had already bolted (in response to that article) hence the management actions.

Morgan Stanley Marks Down Its Stake In Palantir 32%

Don't interpret my comments about management style as lack of intellect. We each have our own styles; few can change them.

Salut toi, je vais au Social Club avec des amis ce soir, c'est au 142 rue Montmartre. J'ai mis ta robe préférée. Viens me trouver.

katastrofa


Total Posts: 378
Joined: Jul 2008
 
Posted: 2016-05-27 00:31
In my uneducated opinion, Palantir doesn't have to be profitable to exist, because it's being secretly funded by CIA, NSA or both.

rftx713


Total Posts: 100
Joined: May 2016
 
Posted: 2016-07-05 21:11
>Genuine curiosity: If you are very talented and able to get jobs with either, who would choose Palantir over the CIA or NSA, particularly for the line of work being discussed here? The main question being: what in the world would the CIA, or particularly the NSA, actually need these guys for?

jslade


Total Posts: 1115
Joined: Feb 2007
 
Posted: 2016-07-06 00:47
" The main question being: what in the world would the CIA, or particularly the NSA, actually need these guys for?"

The same thing they need Booz Allen Hamilton for (Snowden's employer): to actually get things done without red tape, bypassing GS-regulations on salary, seniority, and not having the overhead of government administration, pensions and non-ambitious pointy headed manager.

Government is a poor innovator. They really need contractors to solve problems. I remember looking over some guy at NASA who built these amazing spectrometers which musta cost a half million bucks a piece. Great craftsmanship; a real throwback to the 1950s. But you can buy smaller, nicer ones for $10k from spectra-physics or Newport.

"Learning, n. The kind of ignorance distinguishing the studious."

Rashomon


Total Posts: 172
Joined: Mar 2011
 
Posted: 2016-07-06 02:44
akimon: Perhaps the 3rd party data analytics bubble may start to burst?


What's going to happen to IBM when the analytics bubble bursts?

"My hands are small, I know, but they're not yours, they are my own. And they're, not yours, they are my own." ~ Jewel

Maggette


Total Posts: 1001
Joined: Jun 2007
 
Posted: 2018-04-07 11:36
The German state Hessen has brought a Palantir product. The size of the deal was not revealed

Ich kam hierher und sah dich und deine Leute lächeln, und sagte mir: Maggette, scheiss auf den small talk, lass lieber deine Fäuste sprechen...
Previous Thread :: Next Thread 
Page 1 of 1