Category: business

betting on failure

~ ginsu ~ Leave a comment

It’s interesting to watch reaction to the news of Twitter’s financing at a $1 billion valuation.  The vast majority of commenters seem appalled (or at least cynically amused) at such a lofty valuation for a company with no meaningful revenues.

The shocked reaction misses an important point:  Everyone believes that investments in companies like Twitter are likely to fail, including the investors in Twitter.  For the most part, people who invest their money in companies like Twitter are not putting their life savings into a single company; they are investing their portfolio (or an allocation of it) into high-risk, extremely-high-return-potential companies.  For that high-risk portfolio, it could be rational to invest in companies with a 90% chance of failure, if there is sufficient return for the other 10%.

Now, there aren’t many actual portfolios that are (intentionally) structured with any allocation to a class of investment with a 90% failure rate.  But it would be completely typical if every single non-employee investor in Twitter made their investment from an allocation that has a greater-than -50% failure rate.  In other words, most Twitter investors believe that it’s likelier than not that Twitter will fail.  (Here, “failure” means that the investment will fail to reach the modeled return, not that the company will completely go out of business.)

It’s easy to say that Twitter will probably fail, but how many critics are confident that there is a less than 10% chance at a 10X return?  Investors in Twitter don’t bank on Twitter, they plan that either Twitter or one of the other companies in that allocation of their portfolio will make an outsized return.  Many of those investors have been right time and time again about their projected portfolio performance, which means that as a reward they will continue to invest in companies that are likely to fail.

losing my privacy

~ ginsu ~ Leave a comment

Another burst of news about privacy online, with Ars Technica explaining that removing personal information from data isn’t enough to protect anonymity, and The Monitor giving an overview of how we’re losing our privacy online.

But these people seem to talk on and on about privacy with some seriously flawed assumptions.  They assume that everyone agrees on what privacy is, and that everyone wants privacy in exactly the same way.

I’ve become enamored of the comparison between privacy and religion.  Even without being religious scholars, most people have a basic notion of what the word “religion” means.  And most everyone understands that different people can have very different views about how to practice their religion, or whether to practice any religion at all.

Privacy is the same way, isn’t it?  There is some shared understanding of what the term means, but the specifics of the meaning and practice of privacy can be very different among different people (especially across generations).  Some people don’t believe privacy is important at all, choosing to live without it.

Both religion and privacy deserve the protection of our laws, and for very much the same reason:  the practice of these matters according to one’s own belief is essential for building and maintaining a sense of meaning in life.  In simpler terms, a personal view of these things are required elements of the pursuit of happiness.

Our laws protect religion (and atheism) without saying that “religion” must include a single deity, or prayer at sunset, or robes or hats or ritual.  It’s a mistake to think we should protect privacy by defining exactly what data people should consider private.

Breaking my tradition of linking privacy posts to ’80s songs, because this early ’90s song has perfectly apt lyrics:

Every whisper
Of every waking hour I’m
Choosing my confessions
Trying to keep an eye on you
Like a hurt lost and blinded fool, fool
Oh no, I’ve said too much

I see you, you see me

~ ginsu ~ 1 Comment

Does anyone care about online privacy?

The New York Times thinks so:  just since I’ve been paying attention, I’ve noticed – 1 2 3 4 5 6 7 8 – eight articles about the threat to consumer privacy posed by increasingly effective online behavioral ad targeting.

Jeremy Liew is concerned that the recent public interest push for privacy regulation will threaten startup media companies, suggesting that the ad networks should band together to lobby against online privacy regulation.  He says “While it is always hard to argue against privacy, the impact of this level of restriction would be enormous for companies relying on online advertising.”

It’s not that hard to argue against privacy, it’s just . . . delicate.  And I think simply saying that a lot of money is at stake isn’t enough of an argument.  So I’ll try to make a better argument for why privacy legislation of online advertising is likely to cause more harm than good.

I’m actually a huge fan of the Electronic Frontier Foundation and Consumers Union, and I think their hearts are in the right place on this.  I’m generally in favor of legislation that protects consumers from predatory practices in the marketplace.  But although privacy is a special value, it is not something that is well served by detailed regulation.

The problem is that privacy means many different things to different people, so everybody’s expectations can be quite different in terms of both substance and process.

The substance of privacy is the content of what you want to keep private.  Some people don’t care if you know whether they are male or female, but they don’t want to reveal their age.  Some are ok with gender and age, but not job and income – etc, etc.

The process of privacy is about the availability, collection and use of the information.  Some people want to opt-in to every interaction, some prefer to have opt-out control.  Some are ok with information used in the aggregate but not the individual, or even vice versa.  Some are ok with information being used by private parties, but not the government, or for a day or a month, but not a year or a decade.  Etc ad nauseum.  Few of us are ever thinking about exactly the same thing when we think about privacy.

Privacy may be a fundamental right, but it’s more like the right to freedom of religion than the right to trial by jury.  The latter is a specific procedural right, which we want everyone to have in a very clearly defined way.  The former protects an abstract and highly personal set of values, which each person may regard in a different way.

In the US, we don’t protect religion by telling people what it means; we protect it by saying that the government won’t promote any particular form of religion, and people can exercise any form they choose.  The failing of the public interest proposal on online privacy is that it presumes to define privacy for everyone.  That’s a dangerously unsophisticated view of a standard that varies from person to person and evolves across generations.  A time-traveler from before the Internet would not recognize what the average Facebook user calls “privacy.”

So how do I think privacy concerns should be addressed?  Well, by the market, of course. Don’t get the wrong idea:  despite my love of entrepreneurism and therefore capitalism, I don’t believe that the market is infallible, nor do I believe a free market must be unregulated.  But where you have complex consumer preferences and an infinite variety of potential solutions, a market is often the best way to satisfy the most people.  Think again back to religion:  people basically make their religious choices in a free market as well.

Consumers should have a large variety of choices about how their personal marketing-relevant data is collected and used by advertisers.  The role of governmental regulation here should be limited to traditional consumer protections about clear and full disclosure, contracts of adhesion, and anti-competitive practices.

The government just needs to make a level playing field.  People do care about privacy, and companies that can address those concerns correctly and creatively will make a lot of money.   And that matters not just because it’s a lot of money, but because it’s a case where consumer interest and the pursuit of money can be aligned.

[Apparently I’ve decided that my posts on online privacy must be titled by reference to 80’s hits.]

unforgiven

~ ginsu ~ 1 Comment

Apparently Chris Dixon is my new blog crush, a potential successor to worthies such as Pmarca and Steve Blank.  And I’m not alone:  Venture Beat picked up on Chris’s suggestion that the 2-and-20 compensation “rule” in venture capital compensation deserves to be revisited.

But my thinking about this whole conversation is best summed up by the immortal (and surprisingly moral) William Munny:  Deserve’s got nothing to do with it.  People talk about venture capital compensation as if there is some moral justification for what they make, or conversely a moral reason why they shouldn’t make their money.

Now, I don’t belong to the amoral school of thought that says that people “deserve” whatever the market will pay them – sometimes the market is wrong, sometimes the conditions are unfair, so sometimes there are outcomes that are morally unjust.  However, this is not one of those times.

Venture capitalists negotiate their compensation terms with extremely sophisticated investors.  Those investors are willing to pay VCs an amount that still gives the investment portfolio an expected return that is correct at the time of projection.  Sometimes the investor projects incorrectly, but that’s not a moral misjudgment, it’s just people not being good and/or lucky at their jobs.  And if the VC sector underperforms expectations, compensation will get adjusted over time.  Deserve’s got nothing to do with it – this is a case where the market is perfectly capable of taking care of itself.

iocane advice

~ ginsu ~ 3 Comments

Chris Dixon and Fred Wilson provide a very special kind of bad advice on the topic of equity grants in startups.  Now, Dixon and Wilson are both very smart and very successful, and what they say about equity grants is absolutely true, so the advice is not bad due to its supporting expertise nor its substantive merits.  The advice is bad because nearly everyone who attempts to use this advice will use it to their own harm, and the few folks who cannot be harmed by this advice have already lived a life full of preparation and savvy choices.

Dixon emphasizes that the most important thing about equity grants is the percentage of the capitalization granted, and Wilson adds that the implied valuation of the grant (number of shares times share price of most recent financing) is also useful.  While these things are true, my objection is that the probable audience for this advice is composed of prospective startup employees, and the use that they will make of this advice is to try to choose a job based on the value of the equity grant.

This is a bad idea for two reasons.  First, valuing an equity grant is only secondarily about determining the percentage of the company – it is primarily about determining the exit value of the entire company, an exercise at which professional investors in the field routinely fail. (Fred himself will tell you that 2/3 of venture investments in a successful fund will break even or lose money.)  If you are thinking about joining a startup, and you have 2 choices, you are very unlikely to have any rational basis for believing that 0.1% of one startup will be worth more or less than 0.2% of the other.

Second and more importantly, if you want to work in a startup, you should not choose where to work based on compensation.  You need to pick the project and the people that get you most excited, period.  Without a belief in the mission and an authentic fit with the team, you will not be successful anyway, so any compensation will be a waste of your time and their money. If you have other employment options, you should explain that to the place you want to join, and if they want you they will make the comp work within their range, and you should accept.  Or, if you simply want to work at the place where you will be paid the most, you should not work at a startup. (Don’t be offended, this isn’t a test of character or a judgment of your soul – if you’re not a startup person, that doesn’t make you any worse or better than the people who are.)

Dixon actually gives really good advice in his post, for those who are paying attention:  “If management tells you the number of shares and not the total shares outstanding so you can’t compute the percent you own – don’t join the company!”  As I’ve said before, the reason to have a detailed conversation about equity comp with your manager is to test management’s clarity and forthrightness in general – not because you have any hope of making a correct equity valuation.

I would be willing to bet that neither Dixon nor Wilson has ever made a choice of company to join or invest in based on equity percentage.  They made their choices from their interests in the market, the product, the team – and then later, after a decision to join/found/invest has essentially been made, they did some optimization around the equity.  Choosing the other way around is about relying on luck, not successful choices and preparation.

[Bully for you if you know the reference for the title of this post!]

trademarks gone wild

~ ginsu ~ Leave a comment

I try to avoid drawing parallels between trendy tech issues of the day and my own past experiences – generally I believe that to move forward you have to treat most of your past as irrelevant.

But the parallels are too strong in watching Twitter make a controversial attempt to trademark the term “tweet,” bringing them into a cycle of uncomfortable conflict and limited accommodation with their own developers.

Second Life faced exactly the same issues – a passionate and well-meaning developer community using many terms associated with Second Life that the company hoped to protect as trademarks.  We ultimately came up with a comprehensive policy that was and remains a subject of derision in the SL community (see comments to the linked blog post).

It can be very difficult to engage in a productive conversation about trademark law, because even the basics are hard for nonlawyers (and some lawyers) to absorb, and yet because we’re just talking about using the English language, it seems like anyone who speaks English good should be able to comment intelligibly.  [Yes, the usage error in that sentence was intentionally ironic.]

I think everyone – the company and the commentators – could make better progress by ignoring the legal issues, and just focusing on the marketing questions.  Now, marketing is another one of those disciplines that requires a lot of expertise, and is nonetheless discussed with fervor by anyone who has a couple of IQ points to rub together.  But I think the marketing questions here are simple enough even for me to understand.

1st question:  Is there a name for the product or service that the company should be able to control?  The answer to this question is almost always yes for at least one name – companies are generally better off when they control the primary name for their offering.  Once you reach that answer, following trademark law in order to implement that answer is a straightforward process, and having good customer communication around that process is a requirement.

2nd question:  When there are words associated with the product or service that facilitate the use or adoption of the service, is that facilitation improved or hindered with greater company control over those words?  Marketers and lawyers almost always have the same bias for control (though for different reasons).  The bias itself is always wrong – I don’t mean that it’s always wrong to have that control, I mean that it’s always wrong to approach this question with bias.

Does it really do any good for Twitter to own the word tweet?  Some brand marketers and lawyers will raise the specter of genercide (basically, losing control over your brand name), but this fear should not be the primary analysis unless we are talking about the primary name.  When we are talking about those strongly associated words that help spread the gospel of the company, the analysis should not be of the law and certainly should not come from a place of fear.

The analysis should dispassionately examine whether unrestricted use of the words will help spread that gospel.  And it will often make sense to have less control over these words, not more.  If religion were a business, it would probably make sense to trademark “The Holy Bible” – but trademarking “Christ” would probably make for a lot fewer Christians.

fighting the good fight

~ ginsu ~ Leave a comment

Bernard Moon pointed out these slides on the culture at Netflix, which may be the best presentation on company culture that I’ve ever seen.  But does that mean that Netflix actually has an effective culture?

Of course not, you can’t tell from a slideshow how a company really operates.  Employee comments are helpful, but not conclusive – Netflix has public reviews at Jobvent, Telonu and Glassdoor, which show a mixed approval rating.  But from the outside you never know if the complainers are malcontent underperformers, or if the fans are deluded Kool-Aid drinkers.

At Linden Lab, we spent a lot of time on company culture, creating and periodically revising the Tao of Linden.  That document was similar to the stated Netflix culture in emphasizing a high degree of both choice and responsibility.  I loved the culture we built, as did many employees, but I can’t say that it’s a culture that works for everyone.  And I won’t say that there’s any single best way to run a company (though there are many undeniably wrong ways).

I’ve worked in some centralized, command-and-control environments, and cultures based on internal competition and depersonalization to the point of dehumanization.  And I’ve had plenty of fun in most of these places.  I’ve come to believe that the single most important thing about a company culture is whether or not management truly believes the culture matters.

Every management team will give at least some lip service to company culture.  The companies that stop at mere lip service end up with hollow words engraved in the lobby – these are the truly miserable places to work.  The companies that put real time and thought into their culture, in the firm conviction that a great culture is required for enduring success – these are always great places to work, almost independent of the actual values of the culture.

Commitment to the culture, a genuine determination to fight the good fight to make the company a place with a certain cultural identity – this always leads to a great place to work for some set of people.  A culture of choice and cooperation works well for certain kinds of people.  A culture of command and competitiveness works well for others.  Even a culture based on greed and amorality can work, depending on the industry.

Which is not to say that anyone can work in any culture – in fact I’m saying just the opposite:  you should understand what preferences and constraints your own personal values carry, for this determines what kinds of cultures you will enjoy.  And then it will be easy to identify the companies that express your cultural values.  The hard part will be determining whether the leadership is really committed to fighting the good fight.

hey facefeed, let’s just be friends

~ ginsu ~ Leave a comment

I can’t help wondering if Facebook’s acquisition of FriendFeed isn’t an overreaction by the giant social network, in response to the deafening buzz around Twitter.

It must have irked Facebook that the tech blogosphere has been obsessed this year with Twitter Twitter Twitter – Facebook’s growth has been just as impressive, arguably more so since it’s rarer to grow a large base much larger than it is to grow a small base into a medium-large base.

So in the past year, Facebook has tried to buy Twitter and has copied features of both FriendFeed and Twitter.  And this acquisition appears to be about bringing the values of FriendFeed to Facebook.  Among those values are an emphasis on product openness and sharing beyond your circle of friends.

But what if users don’t want to be more open?  Could it be that Facebook grew so fast because its users regarded the service as a safe place to share their lives with only a close circle of friends?  If Facebook becomes more like Friendfeed, will the service become less attractive to a mass audience?  (I’m certainly going to have to rethink my social media use.)  Maybe it’s only folks like the 250 that believe that everyone wants to share everything all the time.

Oh sure, Facebook has and will have a variety of privacy settings that give people choices about what to share – but these are terribly confusing and difficult to use.  More importantly, a company has to choose a single dominant brand image.  Will Facebook remain the place where friends can share their lives?  Or will it continue to morph into a knockoff of its less popular competitors?

Early indications are that Facebook will integrate FriendFeed’s staff, which is likely to lead to shuttering the FriendFeed service.  I think that could be a lost opportunity.  Facebook might do well to reaffirm its core brand as a more private place for friends, and retain FriendFeed as a brand extension that focuses on open data and public sharing.  That way they can serve the mass market and the avant garde with different product philosophies and branding.

weighing your options

~ ginsu ~ 1 Comment

A friend asks:

I’m working through a start up analysis based on a web-based software app. What is “options consistent” with a start up?

The short answer:  show the prospective employee a 5-figure number, and convince the employee that it makes sense.

The long answer:

This is highly dependent upon the type of startup and the position of the employee. I’ll answer for a typical situation of a VC-backed startup and a developer from entry-level to senior (sub-exec) level. You could have a very different answer for startup that was not VC funded, or an executive level position. You might even have a slightly different answer for non-developers, certainly in the areas of marketing, administration, customer support. (I’m putting aside for the moment the question of whether it is “right” to treat execs or different functional areas differently.)

Many propspective employees seem to take a highly illogical view of evaluating startup options. They compare the raw number to that in their other offers, or to offers that their friends received. (“Well I think I deserve at least 30,000 shares in this company, because my friend Jonny got 20,000 shares in his company, and he’s an idiot!“)

This seems nonsensical because the value of the options must be calculated with respect to the specific company situation, especially in terms of the company’s existing capitalization and prospects for liquidity and growth. Having 20,000 shares in a company that has 10 million shares outstanding is, absent other facts, five times more valuable than 20,000 shares in a company that has 50 million shares outstanding. That’s simple enough, but it is by far less important than the other main factor. Having 20,000 shares in a company that is about to go public might be much more valuable than having 20,000 shares in a company that has just started.

Might be. Or might not.  What if the company that has just started is the next Google, as they all think they are? The problem here is that you have to evaluate both distance to liquidity and prospects for growth. Figuring out which startups will be successful and when and how big they can get is extraordinarily difficult – these are things that professional money managers routinely get wrong. A prospective employee has very little hope in getting this evaluation right.

So let’s look at it from the other side: how do companies decide how many options to offer employees? Typically a company budgets a particular target of dilution from issuances of options over 12 to 18 months. For an early stage startup, this target is often around 15-20% of total capitalization (including the options pool). A one-year hiring plan in that stage might call for something like one new executive, 3 senior employees, and 12 employees from entry to mid level. So the company would budget its options accordingly, obviously also aligning grants with external market conditions.

A prospective employee who wants to know whether an offered grant is “fair” really has no better method of evaluating this than by asking the company to explain how they came up with the number. Ideally as an employee you’d want to ask:

  • What’s the fully diluted capitalization?
  • How far is the company from liquidity? What type of liquidation event does the company anticipate?
  • What are the company’s business prospects for the current year?
  • What is the options range for my position, and where am I in this range relative to other recent hires?
  • How far into the hiring plan is the company for the current year? How many and what positions will be hired?

And you ask these questions not because you can actually value the company based on the answers.  You ask as a test to see if the hiring manager has thought through the offer, and sounds as if there has been rational thought behind your compensation.  You want to work at a place where the management can provide sensible answers to these questions, independent of whether the answers can add up to a company valuation.

Many candidates do not feel comfortable asking these kinds of questions. Worse, some companies will not answer them, and will view the asking of such questions as a sign of impudence.  I’d say you should think twice about working for any company that would be insulted by the asking of these questions, but unfortunately that company attitude is not uncommon.

As a result, the best guideline to fall back upon for many prospective employees is back to good old Jonny:   What have I been offered at other companies, and what are my friends getting at their companies?

Which turns out to be not such a dumb way of evaluating offers, because many companies use more or less the same budgeting processes and have similar investor and advisor networks. So most companies end up in a similar range of options for similar positions. For most mid-to-senior positions, this will be a 5-figure number, and as long as that number can be justified to the employee, then you can move on to more important questions, such as why anyone would want to work at this company in the first place. There should be a lot of answers to that question, and the options offer should be only a very small piece of the puzzle.

virtually great currency

~ ginsu ~ Leave a comment

The acquisition of SuperRewards by Adknowledge is a notable milestone in the evolution of virtual currency business models. This is the first time an independent virtual currency platform has been acquired by a company outside of the virtual goods category, and so the first time that a virtual currency has achieved monetization for someone other than its creators and users. We’ve moved into the peak of the third phase of business models for virtual currency.

The first phase was a sort of prehistory where virtual currency was a gameplay feature of massively multiplayer online games – points that players could gain through the completion of tasks, and use to acquire in-game items that were valuable for further progress in the game. Although points have been a feature of most videogames since the inception of the medium, the relevant new thing about MMOGs was the operation of a “persistent” online economic environment. That meant that even when particular players weren’t online, the service constantly maintained an environment where items of value could be acquired and traded. Much of the trading of items for value was “off-service” – often against the game rules – but this was the first step in virtual currencies breaking free of gameplay rules.

The second phase started when online services that were not solely game-oriented used virtual currencies to encourage trading of service assets – this time trading currency for service items wasn’t against the rules, but specifically designed to encourage sales within the service. Korea’s Cyworld was a pioneer in this use, with “Cyholics” using “acorns” as a medium of exchange for digital presents that users could buy for themselves and each other. Chinese Internet portal Tencent built QQ coins into a $900 million economy, while in the U.S., Second Life users are heading towards $450 million (in U.S. Dollars) of Linden Dollar transactions. The authorized use of virtual currency within these services led naturally to implicitly or explicitly authorized use of their virtual currencies outside of the traditional boundaries of the service, demonstrated by Chinese users buying real-world items for QQ coins and Second Life users setting up 3rd-party currency exchanges and virtual goods stores. (As an illustration of the differences in culture, it’s interesting to note that the Chinese government eventually banned the use of virtual currency for “real” items, and that Linden Lab rebuilt or acquired the third party services.)

In the third phase, we have businesses that were natively built as a platform for virtual currency to be used on other services (rather than a feature of an economy within a more comprehensive service). Some have stayed closer to virtual currency’s MMOG roots, like PlaySpan and LiveGamer, while others have tried to ride the wave of social media apps platforms, like TwoFish and SocialGold. SuperRewards and OfferPal brought a new twist by using marketing offers as the underlying value to the virtual currency.

This part takes a little bit of explaining. For any currency to gain favor with a user base, there must be some underlying value to the medium of exchange – from a consumer point of view, this is sometimes expressed as a demand that the currency be “backed” by something of value. In ye olden days, governmental currency was backed by precious metal; in theory you could turn in your dollars to the government in return for equivalent value in gold. Most governmental currencies came off the gold standard decades ago, and are now backed by the declaration of the government that the currency is legal tender. The meaning of this declaration is a little murky both in theory and in practice.

Suffice to say that there are virtual currencies that emulate most of the historical models of real governmental currencies. e-Gold tried the gold-backed model, to disastrous result. Some virtual currencies are run as essentially stored value systems for governmental currency, so ultimately they are backed by the same declaration of the government. QQ coins to some extent, and Linden Dollars to a greater extent, are free-floating media of exchange that are backed by the commercial viability of their operators – a private rather than governmental declaration of value (this is not as revolutionary as it may seem, since in many ways it’s similar to airline miles and other customer loyalty programs).

By using marketing offers as the underlying value, virtual currency operators can sidestep some of the difficulties involved in demonstrating that a currency is sufficiently “backed” to satisfy customer demand for stable value. This technique introduces significant complexity and cost by introducing many additional parties to the value chain, but now SuperRewards has demonstrated (to its investors if not yet a skeptical public) that this kind of backing does create a valuable virtual currency. OfferPal is not far behind, and of course is now far ahead in terms of its ability to maintain an independent business.

So what’s coming for the fourth phase of virtual currency business models? That’ll have to be the subject of another post. But for now the developments to watch are the competition between Facebook and MySpace in their own virtual currencies, app developer currencies from companies like Zynga, and the continued progress of OfferPal.