Thursday, October 2, 2014

Paul Graham On "Before the Startup," And The Startup Bubble

Paul Graham has written his first new essay in almost a year, "Before the Startup". It seems like his essays have had a long interruption, ever since he wrote one that was politically correct by 2010 standards by not politically correct enough by 2014 standards.

I like the new essay, and his thoughts on startups in general, but he is an uber-bull on startups.

I think the flaw in his thinking is that, while a startup may be making something that users want (all-important in his view), lots of people want things at a price of $0 that they wouldn't want at a price of $n>0.

I've never seen him write an essay about what type of products, originally given away for free, will subsequently be possible to charge for.

People say that these free products will be advertising supported. Let's see, if it's very cheap to start a website that will attract users who can be served ads, what does that say about: (1) the future prices of ad placement, (2) where the rents from advertising will go?

Being able to serve ads was historically a great business, but that was when there was a very high fixed cost and low marginal cost of serving the ads.

Google has a market cap of $400 billion and Facebook has a market cap of $200 billion; $600 billion combined. Google has EBIT margins of 23%, Facebook (which is a purer look at advertising margins) has EBIT margins of 36%.

Global advertising spending is about $550 billion. Only about a quarter of that is digital. That implies that digital advertising is creating about $50 billion in operating profit.

I get the impression that the glut of capital from VCs has led to something on the order of 5,000 startups. It seems as though what's left of the advertising pie isn't enough to justify the valuations at which these have raised money.

My other big problem with advertising-supported giveaway businesses, besides the nonexistent barriers to entry, is that advertising is a bet on middle class discretionary spending. We already know that global elites have decided to gut the middle classes. Do the VCs have a variant perception, or is it hard to see from inside their Silicon Valley bubble, just like it was hard to see the real estate bubble from inside Manhattan?

6 comments:

Viennacapitalist.com said...

Interesting, also the larger the share of online ad-spending the more cyclical (advertising being pretty cyclical) the operating profit of google, FB a.s.o. should be, which means they should derate, as their profits will not be as recession proof as in the past...
P.s. thanx for linking to my recent post

Josh Maher said...

Reading Paul's essay, I don't see where he says that all startups should give things away for $0.

I do see where he says that startups need to make things people want - that is a pre-requisite to having a successful business. Paul isn't going into business models in the essay though so the amount one should charge, how they should obtain revenue, and so on are left for another conversation.

Nick said...

Completely agree although to be honest I've always thought the problems were a lot bigger than just all these incremental start-ups chasing ad dollars...

If you look at how little variance there generally is in ad spend/GDP and think about the way most companies allocate ad spend (i.e. as a fixed % of sales), and you consider the exploding range of places to place ads, it only adds to these pressures.

To my mind, the only thing that could potentially offset this is the argument that targeting and data-driven ad spend increases the ROI on ads (which it should)- but I have a hard time believing this is enough to keep prices even flat....

CP said...

Vienna, I like your blog. You and youngmoney are two good new blogs.

CP said...

http://www.creditbubblestocks.com/2014/11/the-techcrunch-bubble-index-parsing.html

CP said...

bubble bubble toil and trouble