Why Twitter's IPO is Bad for Startups
jackealtman.com
jackealtman.com
That sounds richly valued to me, but not insane. IHG (disclaimer: shareholder and double super platinum for life due to often staying there for business travel) has been trying to move, for years, into a capital structure which Airbnb had from day one: they want a cut of every night's stay without actually owning any hotel property because that's a capital-intensive location-dependent slow-growth business. They've got a substantial scale advantage on Airbnb (~600,000 rooms which each rent 300+ nights per year, whereas Airbnb does about 50 million nights a year at systematically lower price points), but 5~10% a year is great growth for them, and Airbnb is growing at about 100% every 6 months.
There exist significant risks to the company, from the legal issues to merely not sustaining the hockey stick for long enough to justify the valuation, but the people with money on the line are mostly rich and sophisticated. If they lose their investment, oh well, capitalism happens.
It's categorically better than when $X00,000 of pet food sold at a $Y0 million loss was valued in the billions.
The threats to airbnb are much greater (and of course different) than the threats to IHG.
IHG has a lock on locations because of contracts in addition of course to locations which they own.
IHG has brands.
And they have standards. You book with them because you expect a certain level of service when you visit one of their hotels (regardless of who owns the property). They have a functioning marketing machine which fills up hotel rooms with conventions and corporate travel.
Otoh, airbnb is much more susceptible to a threat from someone else doing something similar, slightly different, or essentially the same.
There is a barrier to entry in being IHG as they have infrastructure and an established brand. To me that makes them valuable.
Since airbnb is effectively an exchange, how would another exchange move in to replace it? Marketing and branding is one way, particularly since airbnb hasn't fully saturated the market. One strategy might be to include hotel rooms in listings since airbnb is unlikely to do that given their niche. Another way might be a series of negative events (eg. a serial stalker or rampant bed bugs) which could really tarnish their brand.
That is the double-edge sword of being a tech company. Non-capital-intensive, but fast growth. Make it work, and make your money fast, because what you're doing has almost no barrier to entry other than code and server time.
Trying to knock a major bricks n mortar hotel chain off it's perch is 1000x more difficult
What delusion do you tell yourself to discount everyone else that talks about the importance of network effects? They're the crazy ones?
Sorry, just blows my mind how people think in such black and white terms. Just because one thing is easier to displace doesn't mean it's immediately a "commodity" that can be replaced by the equivalent code and servers farms. Might be great dev-think but definitely gets you nowhere fast when trying to actually build a successful business.
For one simple fact, Holiday Inn is a physical entity providing service to people with massive brand in a industry which is capital intensive and requires endless customer focus.
The easiest example is Elon musk trying to make a car company Vs building Pay Pal. Difference is so enormous, though both are difficult, they are yet no way comparable.
I was illustrating that building a primarily online-based business (including code, servers, network, marketplace, branding, awareness, good people, etc) is faster and simpler than a classic business like a hotel chain (which requires all of the above, plus significant investments into geographically disperse physical structures.
I'm absolutely not saying it's "easy", I'm just reflecting on the market valuation of an organization that (in my opinion) is more at risk of competition than a classic business.
Yes, it takes time for your marketplace to reach critical mass. But there isn't anything that ties people there if there are viable competitors. If AirBNB doesn't have the inventory where I'm looking, I just open a couple of tabs and check out other sites with the same model. Just like I do with airfare.
I actually can't think of an example of an established e-commerce tech company with a two-sided marketplace that this has actually happen to.
eBay, Amazon, Paypal, Google, Alibaba etc. - even Silk Road quickly took up a monopoly position and required the feds to make it go away, it was very resistant to clones that were cheaper, faster, better etc.
It is because they have network effects ^ 2. It is proven that they are resilient, there is tons of research in the field:
http://www.hbs.edu/research/pdf/12-024.pdf
This is the reason why these business models are so attractive to investors, the companies are incredibly difficult to bootstrap, but very lucrative.
One thing that is different about airbnb vs. the companies you mention is the amount of interaction as well as the transaction size and importance.
The network effect at, say ebay or amazon, is bolstered by repeated interaction of small (and of course some large) transactions. That repetition is important because it keeps the service in the forefront of your mind as the go to place for many things, not just one thing. I am in the habit of using amazon so it's my go to place for things. I just bought a ladder from Lowes only after checking amazon first. Because I buy all sorts of small things from Amazon and rarely need to visit the Lowes site.
Airbnb otoh has the same thing but to a much lesser degree. The habit part in other words. Because the amount of time you need a room is less there is less reason to think about it and more of a chance that (because of the lack of habit) you might try something else.
Habits (and small transactions) are very important. Starbucks for example hates to close their locations for renovation because they have found that once people break the habit many times they fail to return (finding another acceptable morning place to get coffee.)
I don't use AirBnB, but my understanding is they take 12% no matter what. In some ways, AirBnB is more like Hotels.com than the local Holiday Inn.
IHG is InterContinental Hotels Group, one of the largest hotel operators around. And given that their net profit margins are around 25% annually (that's $500 million in profit, and it's growing YOY) [1], I wouldn't be too quick to dismiss them as "a shit business" in favor of AirBnB's 12% surcharge. They seem to be doing pretty well for themselves.
Seriously, folks: take the two seconds to look things up before you snark. You'll look smarter.
[1] https://www.google.com/finance?q=NYSE%3AIHG&fstype=ii&ei=SHp...
I think the answer is trivial, and as a result they should not be considered equivalent.
Can you perfectly substitute the holdings of IHG with the software of Airbnb over the long term? Could Airbnb create a market equilibrium point for non-resident housing in absence of brick and mortar hotels?
Airbnb and Evernote, for example, both make sense of me. They provide immense value to people (with obvious cash streams) and are not immediately replaceable.
Meanwhile, I'd put Twitter and Path in with the Pets.com crew.
I don't know if they can capture that value very easily. There is certainly immense risk. I'm not buying their stock. But there is definitely value there.
The trouble is, if they charged this, then a great many people who add value to my experience of Facebook and Twitter would drop away, reducing my own experience of the networks and probably making me question whether it was worth the money for me.
(Google is a separate question because the network effect is less of an issue for web-apps, email, calendar, and so forth - they are still valuable for my personal use even without the sharing and networking features.)
Many people already pay similar prices for the higher end versions of Dropbox or Evernote, but again the "web app" vs "social network" factor comes into play.
It doesn't benefit me as a Dropbox or Evernote user if more people use Dropbox or Evernote, except in very indirect ways, whereas it does benefit me as a Facebook user if my friends all use Facebook.
So naturally you'd pay hundreds of dollars a year for a subscription, right?
Twitter is valuable because of the huge audience. Making everyone pay will lose that.
BUT, there is an important subset of people who will easily pay hundreds for a subscription.
LinkedIn is a good example: free for everyone, but recruiters pay for extra access. It's easy to imagine brands paying Twitter for a similar model.
But unfortunately I don't think even most Twitter power users will pay. Even if brands will pay, it's a fundamentally different equation than LinkedIn's in that there are a lot more recruiters out there than brands.
That's why I said it's hard for Twitter to capture that value.
I'd put Twitter in with NBC/CBS/Fox and other TV stations that sell advertising against entertainment. That's a pretty well-proven business model with an obvious cash stream.
The archetype story in my mind of how public companies work is: A company has some proven way of turning capital into revenue. They go to capital markets. Raise money Make the revenue stream and pay it to investors as dividends. This revenue stream has some risks and companies can always fail, but ultimately they risks are small enough that companies can be treated by investors as living forever.
Obviously that story gets complicated by reality. But, it's a starting point and it has some core of truth when looking at companies like IHG.
Companies like Twitter or AirbnB essentially don't have capital investments in the same sense. They have something like R&D costs (though even calling it R&D is weird if you compare it to the R&D of companies like pharmaceuticals) but those are covered by VCs earlier on. They do have revenues and they have even more potential revenues so they have value. But, they are clearly not going public to raise capital with which to grow the business. They are going to get liquidity for their shareholders.
The other big difference with IHG is the risk profile. Twitter & AirBnB do have much more rapid growth potential. They also have much more spectacular failure potential. Even a solid company like Google could be obliterated in a few years by an upstart search engine. Facebook, Twitter & AirBnB have even moor 'gone tomorrow' potential. Twitter or Facebook might just become unpopular, giving way to the next new thing. Most users even expect them to.
I think the recent examples from Blackberry & Nokia will be repeated, with even younger companies. I'm not implying that tech companies can't last for generations. I just think they have much higher chances of losing all their value in any given 5 year interval than a bricks and mortar company.
All the underlying things are different. The concept of a company as machine turning capital, inputs and labour into revenue doesn't apply. The concept of IPOs as a way of raising capital doesn't apply. Growth potential is completely different. The idea of companies as semi-immortal doesn't apply in the same way. All those things are supposed to add up to a guess at a companies intrinsic value, the rational valuation of the company. Intrinsic/expected value is a pretty abstract concept. It's a lot more abstract when we're talking about lower chances at a bigger prize.
I don't know what that means for how companies are valued. I suspect there is more systemic risk as this stuff moves to public markets.
This cycle has renewed itself many times in the Bay Area, with recent examples likes Nescape, eBay, PayPal, Google, Facebook, and LinkedIn. I feel this refertilization is not nearly as strong in many other cities like NYC.
I wish I started sooner, and was on the wealth climb Warren Buffet had been on. Unlike him though, I wouldn't donate to the Gates Foundation though. I'd call Elon and ask, "How much faster can we get to Mars?"
</rant>
So essentially the "good" people leave and you end up with a gutted company lacking the most important people that made it special in the first place.
I'm going to be nice and just assume that you're so completely surrounded by competent and successful people that you overestimate the capability of "your average programmer" and underestimate the difficulty which Twitter's variety of "some scale issues" would pose to "your average programmer."
The community is the externally visible thing which gives Twitter value, but that community wouldn't exist without some really awesome technology providing it the capability to exist.
Its about branding, name recognition and value.
That is why Twitter is not overpriced at all. Neither is Facebook and that website sucks :)
I challenge you to go set up a twitter clone using only software which existed in, say, 2008 or earlier, or software which you write 100% yourself. To test it, rent a small swarm of AWS servers and get 10,000 fuzzing clients to hammer on the thing. Should be a fun weekend project, right?
I agree with the sentiments about the brand and the community, but anyone who thinks that just magically appeared like it was some inevitable fatalistic force of history is dead wrong. It took a ridiculous amount of ridiculously hard work by ridiculously smart people. And if anything the fact that I have to point this out is a testament to how excellent and valuable their work really is.
An average programmer couldn't build Twitter.
Even with only 100K users, you start to hit interesting scaling problems that are specific to activity streams. (For example, should you use inbox or outbox model to construct the streams). They are well understood now, but they were not well understood in 2006-2007, when Twitter (and Jaiku too) had constant scaling issues.
1. Of course an average programmer could build twitter. It would just take a very long time.
2. "Not well understood in 2006-2007" is the key. Its easy to say "I could build blah-blah-blah in a weekend" now. Because that's after the non-average programmers at companies like twitter figured out all the hard problems.
But point number 1 is a bit like saying "of course you can factor a 2000 digit number into the two primes which created it, it would just take a while."
This is an awfully oversimplified way to put it, but without launching into a long diatribe wrought with lots of smaller generalizations, I'll just use one big one: it's been my experience that the average programmer is typically no better than a very buggy compiler for a fairly low-level subset of natural language.
I come to this perspective as someone who does a fair amount of hiring for a government contractor. Government contracting is the place where the average to sub-par programmers go to find work. The pay is "good enough" but crappy compared to private sector, and the people you work for range from disgruntled (if you're lucky) to incompetent. But it's a very stable and reliable way to earn a living, as cost-plus contracts have promoted an environment where anyone with a resume can get a job, and where your pay is dictated by how many years you've been in the business (correlated strongly to how old you are).
Very few people have the "dive in and solve the problem" mentality that it takes to build something like Twitter, even using bleeding-edge tools. Fortunately we live in a world where the number of people who do have that talent is quite large, it's just that the number of people who don't is much, much larger.
Could a single average programmer, starting today with cloud technology available today, build a Twitter clone in 7 years if he would get the current +200M users at the same pace as the original Twitter got? Twitter clone here meaning a replica of Twitter 2013, with all supporting functionality (ad management etc.).
I say no, (s)he wouldn't be able to build it.
The opposite has been shown to be true (and correlated with my own experience) for early stage companies. The biggest startup valuation bubbles happen when there are no public tech companies for money to be invested in (eg. the '07 startup bubble and more recently in '11).
With Facebook, LinkedIn, Pandora, and now Twitter etc. public early stage startup valuations have shrunk again.
Where it does get competitive is in the later stage rounds, but that was happening anyway as the mega-funds like DST competed with traditional VC and mezzanine funding (the business model here changed, as companies wanted to go public later).
> Snapchat and Pinterest are worth around $4 billion dollars each. Evernote and Airbnb are each worth over $2 billion. Path is worth $500 million.
Snapchat never raised on $4b, that was a rumor (and it continues to get cited as signs of a bubble even though the round never happen). The only other valuation here that would be out now is Path, and that is only because that startup has struggled. Were it a real bubble, Path would not have any problems raising a new round at $1b, fact is they are laying people off and searching for bridge loans from friends).
> “People who think the company is overvalued,” they scoff, “just don’t get the power of Twitter.”
There is nothing magic to understand, the business model is centuries old - sell services to individuals whose attention you have. Twitter currently has 250M subscribers, larger than most traditional media companies (that required printing presses, expensive TV spectrum to be purchased, entire studios like 30rock, dealing with the politics of media ownership, expensive distribution etc.) and growing quicker.
I am a twitter user, I don't watch television anymore and I don't pick up newspapers or magazines. For advertisers to reach me and people like me, they need to find me on Twitter (or others on Facebook - this is the competition).
The Twitter IPO is great for startups, for a few reasons: first, it returns money to investors who will inturn invest in the next generation of startups, it reinforces the hit parade and that the industry depends on, and it will bloom and entire new generation of angel investors who will be supporting the ecosystem (in the same way former Google employees did 10 years ago and FB employees did 2 years ago).
edit: and I think the Buffett quote argues for Twitter, i've seen that quote used before to argue against Bitcoin, but not against an actual company that is producing hundreds of millions of dollars in revenue and growing at almost 100% a year.
In fact, lots of points made in this post seem to sound as though they are factual, but I'm not sure they are and would likely argue the opposite point of the author:
"The real problem here is that we increasingly live in a world where technology companies are valued on emotional whims and promises of unbelievable future growth."
Technology companies are almost universally about promise and potential and always have been. Read Crossing the Chasm to understand why that is. At the time that the private market investors care about them, they are so early in the selling motions that their early success in a tiny subset of the market is indicative of later potential. That's why VCs exist. To see something with early potential, pay a premium (at times) to get a part of it, foster it's development, and then ride it into reality and receive a terrific return for the foresight and fortitude.
The pattern recognition there is unmistakeable. Where it gets challenged is in the fickle and finicky consumer market where externalities are more likely to cause massive changes in momentum (See: Path can't maintain the momentum of fundraising).
"A surge on IPO day is good of course, but 15-20% is more than enough; 75% simply means the company was mis-priced and that Twitter left over a billion in cash on the table."
This is also not a factually true statement. Because a small float was offered, and demand was prescriptive by the bankers, it's impossible to say that because of how it ended, trying to raise the price to $40 would have resulted in the same outcome.
If Twitter wanted to fully maximize their first day gains, there are dutch style IPO auctions that can be done to better match demand for an offering. Google did this.
I'm not going to continue, but I guess this is a great headline with a poorly made point. I'm not sure what the goal here was.
A startup (or any company) is worth something like the integral of of the graph of probability vs. outcome. The problem I see right now is that people are overvaluing the positive outcomes and discounting the low end.
"Because a small float was offered, and demand was prescriptive by the bankers, it's impossible to say that because of how it ended, trying to raise the price to $40 would have resulted in the same outcome."
I'm not completely sure what you mean by this, but I spent time on an ECM desk one summer and I they aren't aiming for a 75% first day pop. The bankers get paid less, the company raises less. You want a pop to satisfy buy-side clients and to keep a positive public perception, but 75% is too much.
I see fewer and fewer comparisons like these between tech companies and brick and mortar/commodity companies and I think it is a shame. To me such comparisons do more to highlight how broken the financial markets are, though I am not exactly sure why. It seems like investment priorities or consumer demands are out of whack when electrical utility companies (PEPCO) are valued less than social media platforms (FB).
This is really not a logical comparison. These are 2 companies in different fields and have very different characteristics. Investors almost certainly have different reasons from buying one or the other. The comparison between their value is not really relevant.
If a utility company which provides basic necessities were to make a lot of profit, wouldn't it be sort of a bad outcome for humanity? Isn't it better if basic necessities are provided to us at the most affordable prices?
My point is that the valuation comparison OP posted has nothing to do with "life without utility companies and imagine life without FB".
Share value isn't based on value to humanity; it's largely based on profit and potential profit.
My only point was that comparing the valuation of a single utility company against a tech outlier doesnt really make any sense.
Look around, small time companies demand they want the brightest algorithm minds. And then when they hire them, they make them build HTML pages. This is really a bad signal.
I fear the day when people working in companies that make stuff like medical electronics companies decide there isn't much money in innovating in those areas. In many ways this is already happening.
Just sell for less. If everyone else demands high valuations, it's easier to sell, not harder.
Your investors gave you $50 million at a $500 million valuation. Google wants to acquire you for $100 million. Your investors may not let you.
Or you can cancel some of your own shares and take the $50M.
Or if you still have runway and believe in the business, keep going and raise that $100 million.
It sounds like the problem of having scaling issues: it's a problem, sure, but it means you're already a success. If you never got that $50 million to begin with because valuations weren't high, then you might be a failure now instead.
Be sure to check your term sheet. It could be that your investors have a liquidation preference that includes a multiple. Based on my (admittedly limited) understanding, if your investors gave you $50 million for 10% of the company, with a 2X multiple liquidation preference, they would be entitled to the first $100 million of any sale, plus 10%. So you may be able to sell for $100 million, but you wouldn't see a dime.
You lost me with the trickle-down argument, “So if Twitter is worth $25 billion and its size is X,” thinks the VC to himself smugly, “this other startup must be worth $250 million because its size is 0.01X.”
Is the other startup doing something as unique and compelling? And does that startup have the opportunity to own that equally unique and compelling space outright?
Twitter's valuation is born from its opportunity to be THE real-time information platform for the web. I would argue that Snapchat's opportunity (to be yet another photo social network) is nowhere near as compelling.
Sure, valuations are too high across the board, but I do not see Twitter's IPO as bad for startups. If anything, it shows that unique concepts with the opportunity to own their market globally are highly rewarded by the public markets.
As is often the case, I agree with much of your content, but take issue with your sensational headline :)
- who are you going after here? In your post the people I see are the unsophisticated investors who go for emotion/speculation rather than business perspective. Those that say with their gut "I love this product, it's so influential, therefore it's worth $50/share" and do off-the-cuff calculations "Facebook is priced at X so Twitter is around X, too" instead of doing hard research. One can see how it's bad to have these kinds of people involved in a relatively small field (ie VC), but again, your headline is towards startups and not the investors just noted.
- don't startups stand to profit most from these valuations? The problem I see is in having the aforementioned investors. If we have people running around trying to get a slice of the pie, it looks like they'll pump-and-dump. They'll pump their cash into whatever startup seems to have a chance of opportunity and then walk out when it's profitable enough. Win/win, except that means the investors don't really care what the company is doing or up to, they just want to profit.
- with what I just mentioned, startups stand to profit because more opportunity is available. The downside I see is there being a higher chance of shallow investments. When investing in a company, you want to really understand it and know it, but with these valuations and profits, it seems the headline shouldn't be "Why Twitter's IPO is Bad for Startups" but "Why Twitter's IPO is Bad for Sophisticated Investments."
note: It's pretty difficult trying to write out thoughts on here, so hopefully I made some sense. If not, just ignore it as beginner's mind.
my bet is that these billion dollar speculative valuations will continue amongst growing startups until a few of them fail to build solid profitability and fold - then we'll have a classic bubble burst through lack of share market confidence, and availability of cash and growth with slow.
then, cue blog posts about "the virtues of being an organic-growth hacker"
Isn't the main justification for these valuations growth and potential maximum size? Surely an investor can't extrapolate a $250MM valuation based on relative size with a straight face, especially as their metrics will also look quite different.
I think that's only viewable in Canada, but Canadians, play the video and watch the first pitch.
These kids develop an iPhone app for tracking and recommending your workouts. They're asking $100,000 for 10% of their business. Their plan for generating revenue is converting users from their free app to pro at $4.99/mo, $29.99/yr. So far, 12,000 downloads on their app, and they say 1% are converting to paid. That means 120 paid users, averaging say $15 each, so they've made roughly $1,800 in revenue.
What the hell? $1,800, and they value their company at one million dollars? They're not even paying themselves, and they're in one of the most saturated niches in the world, with already countless others well established and successful in that market. What blows my mind even more, one of the investors offers $100,000 for 20%, with 25% royalties until they're paid back. They turn it down.
- There is biggest horse carriage company
- And a dominant railway company
Here comes { Ford, GM, Cadillac........}
Every investor at that time can make a case that how could a transnational railway/carriage company is worth less than any motor company.
Software has a much harder time in the secondary market.
Here's the real reason Twitter's IPO is bad for most startups:
Their API is no longer going to be free in the future. That or they're going to impose harsher rate limits.
Not making money isn't a bad thing for startups? I must be really out of the loop, I didn't realise startups run on Unicorn smiles and jellybeans.
Some also make money.
But VC-backed startups, which is really the only connotation of startup that we use here on YC, are a special breed of startup (generally focused on massive growth) where operating capital derived from revenue is not a pre-requisite for success.
Twitter's revenue so far in 2013 is $422 million [1].
[1] http://www.itworldcanada.com/post/twitter-loses-growing-sec-...
What crack is he smoking?
http://techcrunch.com/2013/11/07/path-loses-biz-head-amid-ru...