Business Models

Those who forget the past are surely doomed to repeat it - TechCrunch:

The SecondMarket Facebook shares auctions are back on after a holiday break, and the valuation is up big time. The last auction prior to this one closed December 15 at $22.75/share. Today it hit $28.26 per share. With 2. 5 billion or so shares outstanding, that’s a $70.65 billionish valuation. A month and a half ago shares were trading on SecondMarket at a $50 billion valuation.

If anybody recalls as far back as 2000, the gig then was to IPO a tiny amount of the company's shares, the resulting scarcity forcing the price higher and higher as overhyped shares were bought by people with more dollars than sense. And that was at least in an open market on NASDAQ!. When you get to a point that a company with un-audited books can be accepted uncritically as being worth $70 bn then you have to either believe in a New Economic Paradigm, in the Wisdom (Madness?) of Crowds, or that just perhaps a good old fashioned Pump n Dump scheme is on the go - Wikipedia.

Pump and dump schemes tend to take place either on the Internet including e-mail spam campaigns or through telemarketing from "boiler room" brokerage houses (for example, see Boiler Room). Often the stock promoter will claim to have "inside" information about impending news. Newsletters that purport to offer unbiased recommendations then tout the company as a "hot" stock. Messages in chat rooms and email spam urge readers to buy the stock quickly.

Unwitting investors then purchase the stock, creating high demand and raising the price. This seemingly "real" rise in prices can entice more people to believe the hype and to buy shares as well. When the people behind the scheme sell their shares and stop promoting the stock, the price plummets, and other investors are left holding stock that is worth significantly less than what they paid for it.

Fraudsters frequently use this ploy with small, thinly traded companies—known as "penny stocks," generally traded over-the-counter (in the United States, this would mean markets such as the OTC Bulletin Board or the Pink Sheets), rather than markets such as the New York Stock Exchange or NASDAQ—because it is easier to manipulate a stock when there is little or no independent information available about the company.

So, to reiterate - tiny numbers of shares changing hands in shadowy secondary markets do not constitute price discovery, despite the promises of promoters, and using them as a basis for valuation of Facebook is about as logical as discovering witches by seeing if they float like ducks (Cue Monty Python sketch above). I really thought Mr Arrington/TechCrunch recalled all this, but this article makes one wonder......

(Our view of what you have to believe for a $50bn valuation is over here - do you believe it?)

Today some news emerged about Facebook's 2009 financial figures - WSJ:

“According to people familiar with the document, Facebook had net income of $200 million in 2009 on revenue of $777 million. Figures for 2010 weren’t disclosed, but analysts have said the company’s revenue last year could be as much as $2 billion, fueled by advertising growth.”

If it is making up to $2bn today, with reportedly c 550m users (Highest sourceable estimates I have seen), that implies average revenues per user of c $3.60 per annum. Facebook at the end of 2009 had about 275m users, ie about $3 per user per annum. However, these figures will under-estimate actual revenue per user as it is dividing total annual revenues by end of year numbers. Assuming a roughly linear growth allows us to calculate the actual mean running revenue per user was c $4.65 in 2010

Going back to 2008, Facebook ended the year with c 140m users and revenues between $325m (Facebook guidance) and $210m (eMarketer). Splitting the difference gives 2008-9 average revenue per user per annum of about $3.65, in other words Facebook has added a mean $1 per user per annum over 2010.

This allows us to start thinking about "real world" valuations, and comparing it with Goldman Sachs / Facebook's valuation of $50bn.

(1) Revenue ratio-based Valuation

A market cap to revenue ratio of $50bn on $2bn revenues gives a ratio of 25:1. To give an indication, Google is at between 5 and 6:1 today as a mature company, and at its IPO had a ratio of about 13 ($20bn on c $1.5 bn trailing 12 months revenue) and it was growing at a similar rate to Facebook. To believe that Facebook can be a c $50bn company when mature, at the Google ratios, you have to believe it can achieve revenues of about ($50bn/5) = $10bn, ie a 5-fold growth over 2010 revenues.

(2) Margin based Valuation

Assuming that last year's ratios hold true - ie 2009 net income of $200m on $777m revenues, ie c 25%. Translated to 2010 means net income of c $500m on revenues of $2bn*. That gives a price/earning (P/E) ratio of ($50bn/$500m) of c 100:1. Google's P/E ratio today is about 18, and was about 120 at IPO. Assuming Facebook can keep a 25% net income margin, as a mature company, with a Google-ish P/E ratio of c 20 means it needs to generate at least $2.5bn net income, ie revenues of about $10bn per annum - similar to the outcome above.

So it comes to this - to justify its valuation, you have to believe Facebook can achieve c 5-fold growth in about 5 years (after that, NPV is near zero for all practical purposes) at current margins. So where will this growth come from?

- It has a about 550m users, there are about 4bn people online today, of which c 2bn are in the OECD or similar "rich" nations, ie have incomes to support the sort of Ad revenue per user that it's current customer base is worth.

- It has an average revenue per user of c $4.65 per user per annum. Apparently the average user spends about 7 hours a month on Facebook. Assuming most revenue is Ad based, a CPM of c $0.5 (about the norm for the Social Media market) that's about 9,600 Ads that have to be served to each person per annum, or about 800 per month, or about 120 per hour (7 hours/month) or about 2 a minute. (Our preliminary data has Facebook running at about 2 minute per page viewed, so it needs to serve c 4 ads/page)

Looking at how it may get to $10bn revenues:

(i) Assuming average revenue stays the same - Facebook needs a c 5-fold increase in users, ie about 2.75 bn users - or more than 1/3rd of the planet's population, about 2/3rds of the 4bn online, many of those who are earn far less than necessry to justify OECD ARPU.

(ii) Assuming number of users stay constant - Facebook needs a 5-fold increase in revenue per user, is about $23 per user on average per annum. At $0.5 CPM that is 10 ads a minute that each user has to see, or it has to expand to 35 hours a month at the minimum required Ad-rate.

(iii) Assuming a rough doubling of users and revenue per user doubles to c $9.50 - Facebook grows to c 1.1 bn users (a mere 1/6 th of the world's total population) and a rough doubling of revenues per person - e.g Ad rates (2 times more Ads, or 14 hours a month, or $1 CPM).

Throughout I have used CPM as a proxy for the user's mean net present value across al services offered - yes, Facebook may sell virtual tat, or a freemium service, or whatever, but I can't see these being the main revenue drivers in the short to medium term any valuation is based over. In many ways, the fact that Facebook is going through this dodgy private IPO despite revenues of $2bn implies the same. In a way, it's main problem is the stickiness of its pages - too much time reading them, not enoght time clicking through them to get to new Ads. No doubt they are working on this, but unlike Google where you quick-click-to-find, on Facebook you are engaging with the page.

Anyway, that's what you have to believe for a $50bn valuation. My take - its do-able on a wing and a prayer, but the chances of Facebook underperforming is far, far greater than overperforming. Not surprised Goldman's own investment unit passed on it, and Goldmans in facts reserves the right to sell/hedge their stake without warning or notice.....

*Goldman's say first 9 months shows $355m net income on 1.2bn revenues - c30% - but there are likely to be annual costs that are not in that. The New York Times says $400m on $1.2bn - ie c 20% - so we will stick to our 25% estimate for the year for now)

....apart from this 'umble blog, of course. There is an interesting summary on Switched of what they believe are some of the better articles on the Web, for the Web in 2010. Most of them are thoughtful pieces, sanning a wide variety of topics. The benefit of human aggregation of editing is writ strong.

Compare this with Techmeme, which lists the 50 most trafficked stories of 2010, basically a litany of Apple/iPad/iPhone/Apple killers, Google, Facebook, and, er...thats about it. That is what automatic aggregation looks like.

This is not say either are wrong - Techmeme is a fascinating system, an algorithm based media selection engine (albeit moderated by, I must say, some pretty smart folks on balance). It is a behavioural psychology experiment writ digital, a barometer of where the digital vox of the populi is actually clicking. The Switched ist on the other hand is looking at where they may click to expand their horizons, but it does highlight a major problem with "popularity" based search algorithms (ie measuring quantity of links) as a measure of quality.

What is more interesting is to think about why there is such a gap between "good" writing and "popular" writing on line.................

(Afterthought - there are plenty of interesting stories on Techmeme, but they tend not to bne teh most popular, most commented on etc, though I must say I found it becomeing more of a "product pimping site in 2010 - whether its Techmeme policy, PR hacks gaming its algorithms, or just the impact of the increasing commercialisation of the blogosphere is hard to tell)

Article by Kara Swisher on why there are so few senior women in all the new Silicon Valley startups, complete with TED talk by Sheryl Sandberg. Well, being modelling wonks, I naturally look at the flow of the supply chain system (see above).

Look at the above model. Assume that for every board seat there are 3 candidates (a Neck of the Bottle thing), ie a 67% attrition rate. For there to be only 15% women on the Board, that means you need c 45% to be women in the next level upstream (the middle boxes). If each of the initial starting boxes were 50% full of women (ie 150% women in total start the flow), then in order to get only 45% candidate representation for Board level you are looking at a 70% attrition rate to those middle boxes.

One can argue that the attrition rate between stages is higher for women, one can argue that some routes (eg Boardroom clubbiness) are not open to women, but as you can see, if the starter boxes were primed at 50% each, you have to believe in a massive attrition rate through the system - the sort of losses that the most profligate of a First World War generals would be proud of (and that led to mutinies).

In fact, we can use this model to deduce something approximating the "real world" attrition rate. If I assume that about 25% on average (rather than 50%) of all the people starting out as middle managers, technologists and entrepreneurs are women, then to deliver a 15% representation at board level (today), still assuming a 33% final attrition, then the actual attrition rate is more like 40%

Now clearly, there are issues that impact this attrition rate, and Sheryl goes into them in good detail (this is PC territory for us men, so let me just re-iterate her points and say they tally with my experience too):

- Sit at the table (ie don't be backward about coming forward
- Don't drop out early (don't make the decision to stop striving when you are thinking of having kids, make it when you have them - keeps options open)
- Equal partnerships at home - its hard to be a CEO if you are also doing 67% of the housework (to this I'd add, don't be shy to use some of the extra income to bring in helpers).

But you still have to start with the initial flow, as that is where the leverage is - assuming the "real" attrition rate is something like 40% as above, if you then have 50% of all starters being women you should immediately see a 30% board representation by women.

Which brings us back to the older chestnut - why are there so few women coming into technology and entrepreneurship areas?

Twitter trying to explain why Wikileaks never trended:

Twitter Trends are automatically generated by an algorithm that attempts to identify topics that are being talked about more right now than they were previously. The Trends list is designed to help people discover the 'most breaking' breaking news from across the world, in real-time. The Trends list captures the hottest emerging topics, not just what’s most popular.

...

Sometimes a topic doesn’t break into the Trends list because its popularity isn’t as widespread as people believe. And, sometimes, popular terms don’t make the Trends list because the velocity of conversation isn’t increasing quickly enough, relative to the baseline level of conversation happening on an average day; this is what happened with #wikileaks this week.

Far be it for me to suggest this may not be entirely true, but I have seen a conference topic trend more than once with just a few hundred people tweeting. I eagerly await the first analysis of the twitter traffic of #wikileaks vs some of the topics that did trend in over the last few days

Doesn't it gladden your heart when trusted things come through? Yes, it would seem that Porn and e-readers were made for each other - Slate:

As I write this, the most downloaded item for Amazon's Kindle is a novel by Jenna Bayley-Burke called Compromising Positions. Here is part of the plot description: "David Strong knows how to do a lot of things—run an international fitness company, finesse stock portfolios and stay out of emotional entanglements. That is, until he gets tangled up with Sophie Delfino and her Sensational Sex workout....

......

You won't find Compromising Positions anywhere on the New York Times or USA Today best-seller list. So how did it become the No. 1 item on the Kindle, slightly outpacing Jonathan Franzen's Freedom? Price. To buy the paperback on Amazon costs $10.20, plus shipping and handling. The Kindle version? $0.00, which includes instant delivery. Christina Brashear, publisher of Samhain Books, explains that she usually makes one title in a series available as a freebie for two weeks, betting that some readers will pay for future titles. Based on the performance of August freebie Venus in Blue Jeans, it's reasonable to assume that some 35,000 people will download Compromising Positions during its freebie run. While few would dare try Samhain's giveaway tactic with a physical book, the tactic seems to be working digitally.

Many—including the publishing industry—would not label Compromising Positions pornography, and with good reason. For decades, romance and women's fiction novels have featured fairly explicit sexual passages, which readers apparently feel comfortable with as long as they are surrounded by what one Amazon commenter labeled "a tastefully written story of deep love and emotional commitment."

But you know the rest is coming (fnarr fnarr), don't you. It will be interesting to see what directions e-Readers (and by definition tablets like iPad) take as Porn starts to drive adoption of technologies in the delivery system.

A quick Flash, anyone?

Mike Arrington on the "why".

The truth is I was tired. But I wasn’t tired of writing, or speaking at events. I was tired of our endless tech problems, our inability to find enough talented engineers who wanted to work, ultimately, on blog and CrunchBase software. And when we did find those engineers, as we so often did, how to keep them happy. Unlike most startups in Silicon Valley, the center of attention at TechCrunch is squarely on the writers. It’s certainly not an engineering driven company.

AOL of course fixes that problem perfectly. They run the largest blogging network in the world and if we sold to them we’d never have to worry about tech issues again. We could focus our engineering resources on higher end things and I, for one, could spend more of my day writing and a lot less time dealing with other stuff.

The (rumoured) $40m price of course never entered into it

That aside, Mr Arrington points to a very interesting problem that all startups will face at some point - it stops being about the technology (unless they are a pure technology startup) and starts to be about the product, and product managers etc etc rather than technology, and in The Valley that can be a tough transition.

And if the technology is not yet bedded down that is a huge risk to the business, and a major diversion of resource. I know of what he speaks from small companies I have run!

But what I can also tell them for free is that it is no picnic integrating into a large 24x7 infrastructure with huge heritage systems - never mind integrating a small, independent, "with attitude" culture into a large machine. We wish them luck - they will need it!

What is interesting is that they didn't opt for an Amazon or similar 3rd party infrastructure solution. But then those don't hand out $40m

(I think it works out to $25m to buy, $15m earnout, but it is, as I said, all rumours - as is the story that Mr Arringtom moved to Seattle some months ago to avoid California asset sale taxes)

Tonight, Vince Cable, UK Business Secretary, blurted out the inconvenient truth when he noted that:

Vince Cable will tomorrow launch an aggressive attack on capitalism with a speech that warns that the current system "takes no prisoners and kills competition where it can". In an echo of Denis Healey's famous 1974 pledge to "squeeze property speculators until the pips squeak", Cable will unveil plans to shine a "harsh light into the murky world of corporate behaviour".

The business secretary will announce the launch of a major consultation on takeovers and executive pay, with the intent of ending "corporate short-termism".

"Let me be quite clear," Cable will tell the Liberal Democrat conference in Liverpool. "The government's agenda is not one of laissez-faire. Markets are often irrational or rigged."

Cue howling by all the big biz meedja poodles and proxies on the TV, newspapers et al. By this time tomorrow Mr Cable's character will have been assassinated more times than a working girl in a conclave of archbishops (or is that vice, versa?) . Meanwhile, back in the real world, as Michael Arrington writes, Capitalism is red in tooth and claw - if you are outside of the closed, smoked filled dining rooms that is....

This group of investors, which together account for nearly 100% of early stage startup deals in Silicon Valley, have been meeting regularly to compare notes. Early on it was mostly to complain about a variety of things. But the conversation has evolved to the point where these super angels are actually colluding (and I don’t use that word lightly) to solve a number of problems, say multiple sources who are part of the group and were at the dinner. According to these souces, the ongoing agenda includes:

- Complaints about Y Combinator’s growing power, and how to counteract competitiveness in Y Combinator deals
- Complaints about rising deal valuations and they can act as a group to reduce those valuations
- How the group can act together to keep traditional venture capitalists out of deals entirely
- How the group can act together to keep out new angel investors invading the market and driving up valuations.
- More mundane things, like agreeing as a group not to accept convertible notes in deals (an entrepreneur-friendly type of deal).
- One source has also said that there is a wiki of some sort that the group has that explicitly talks about how the group should act as one to keep deal valuations down.

At least two people attending were extremely uneasy about the meetings, and have said that they are only there to gather information, not participate.

So what’s wrong with this?

Collusion* and price fixing, that’s what. It is absolutely unlawful for competitors to act together to keep other competitors out of the market, or to discuss ways to keep prices under control. And that appears to be exactly what this group is doing.

It was ever thus, hence anti-trust laws etc etc, problem is the regulators are always one step behind. Private Equity of all stripes is still outside of most regulators' ambit, and no OECD government has yet had the balls to resurrect Glass Steagal, which sort of sets the whole tone for Cable and Arrington's duet.

Still, true or not, this one is going to be interesting to watch play out.................

*A collusion is when two cartels run into each other

We have often explained that Apple's strategy is always to enter (create) a new market and own the 15% or so of the most profitable market sectors, and have done so since the Apple II. Many people have taken umbrage with this, misunderstanding revenue for profit. Revenue is what startups boast of, profit is what businesses still alive 5 years later have. Superprofits means you are still around 40 years later. The graphic above explains all you need to know about this strategy. It's from Fortune:

We are also impressed with Apple's ability to monetize its innovative products through selling high-margin consumer products that drive strong earnings results and growth trends for Apple shareholders. A case in point is the mobile phone market, where most handset OEMs struggle to post a profit or even 10% operating margins (except RIM and recently HTC), while we estimate Apple boasts roughly 50% gross margin and 30%+ operating margin for its iPhone products.

In fact, Nokia, Samsung and LG combined sold roughly 400M mobile handsets worldwide in the first half of 2010 with a combined value share of 32% of handset industry profits, while Apple sold roughly 17M units over the same time period and captured an estimated 39% of industry profits, or greater than the top three global handset OEMs combined. Apple leads the industry in every metric except for unit share...

The quids*, as they say, est demonstratum.

*quid = bucks, dosh, wonga etc etc - specifically in British £ denominations. The above is, in fact, a word play on the Latin......

We have been fascinated by virtual goods ever since reading more pieces of furniture were sold online in Korea in socnets like Cyworld than in real life. You may also recall that a number of companies experimented with virtual goods in 2nd Life (c'mon, you can't have forgotten 2006 - unless you were one of those 2nd Life marketeers ). Anyway, hey ho we are off again.

.....some large companies are testing whether they can raise awareness of their brands — and sell more actual goods — by creating and offering their own pretend merchandise. Volvo Cars of North America, the clothing retailer H&M and MTV Networks are among the diverse brands entering the market for virtual goods — the make-believe items offered on social-networking games, smartphone apps or fantasy Internet sites.

................

So far, the virtual goods market largely consists of micro-purchases. Consumers typically pay $1 to $3 while playing games like FarmVille or Mafia Wars, both created by the social-gaming company Zynga, to get a jump on game rivals. Users also can give a gift, like flowers, or build a collection of items — just as collectors do in real life.

Those impulses will be worth nearly $2 billion in revenue or more this year, according to ThinkEquity, a financial research firm in San Francisco. Its analyst for new media and games, Atul Bagga, said his research found that the market could reach $2.6 billion next year.

.................

To succeed, “branded virtual goods have to be identifiable and have a real world relevance,” said Ravi Mehta, vice president for products at Viximo, a social gaming platform provider. “They are driven by the relevance to the purchaser. Paris Hilton has people who buy her virtual goods because they are fans and want to identify with her, her hair, her place in pop culture.”

I think he is actually serious about Paris Hilton.....

Actually, we've been arguing for a while that this is the best model for Facebook, but whether people will buy virtual trainers etc on SocNets any more than they did(n't) on 2nd Life is still an open question. Those "decorate your own room" socnets have largely passe'd away. All the evidence is that virtual goods are bought in pursuit of another (typically game based) objective, not "real world relevance" - unless maybe they can be converted into free drinks, like 4square mayoralties at Starbucks.

I await with fascination to see if Nike will be selling under-armour gear on World of Warcraft, and Wilkinson Sword will sell...... but I think this is far more likely to be more akin to the 2nd Life experience, as corporate clumsiness leads to egg on faces and rivers running with red ink.....

Still, as these goods cost nothing to make and distribute, they could be the last hurrah of Freeconomics