Sunday, October 18, 2015

Organizational Structure Is All About Pragmatism

First time CEOs and entrepreneurs often call me to discuss how to structure their organization. Common questions include: Should I hire a COO or not? Who should the VP Marketing report to? How should I split up product and engineering? Should international build out its own functions or be matrixed with US headquarters?

There is often fear in the mind of the entrepreneur that there is a "right" answer to how to structure an organization and that if they screw it up by doing the "wrong" thing the implications could be disastrous. This is an incorrect perspective. Most of the time there is no "right" answer and org structure is really an exercise in pragmatism - i.e. what is the right structure given the constraints you face in terms of the talent available to your company, the set of initiatives you need to pursue, and a 12-18 month time horizon.

Here are a few key takeaways and things to keep in mind about org structure:

If you are growing fast, you have a different company every 6-12 months.
When I joined Google it grew from ~1500 to 15,000 people in 3.5 years. After my startup was acquired by Twitter, Twitter grew from ~90 to ~1500 people in 3 years. When companies grow that rapidly they are literally a different company every 6 months. This means every 6-12 months the company's org structure may change.

When choosing an organization structure for a high growth company, focus on the next 6-12 months. Don't try to find the "long term" solution as in the long term your company will be completely different and have radically different needs.

There is no "right" answer.
Often there is no "correct" answer to how to structure your organization, rather it is a series of tradeoffs. Two different structure may be equally "good" and "bad". Don't sweat it too much - ultimately if you make a mistake it is painful but you *can* undo it. 

Communicate to the team that as your company grows quickly things will shift around and it is normal for that to happen - it is a sign of your success and other companies that grow fast do the same thing.

Sometimes bandwidth matters more than perfect fit.
Sometimes executive bandwidth is more important than a traditional reporting chain. For example, Alex MacGillivray, the talented former General Counsel at Twitter, had User Support, Trust and Safety, Corporate Development / M&A, and other areas reporting to him at various times in addition to legal. Many of these areas normally would not report to a GC, but Alex was talented enough to take these on in the absence of other executives with bandwidth to own these areas. As new executives were hired or promoted things would transfer over to them from Alex.

As CEO, you should look at your team and allocate functional areas based in part on who has time and skill set to focus on the area and make it succeed. This does not mean the area needs to work for this executive forever. Remember, nothing needs to be permanent. There are also some cases that don't make sense from a tie-breaking or skill set perspective - e.g. your VP Engineering should probably not run sales and manage that team in addition to engineering. However your VP Engineering could potentially have the design team work for her or the product team if needed short term or if it makes sense to do so longer term. 

Org structure is often about tie-breaking.
Reporting chains are ultimately about decision making. E.g. engineering and product management have a natural tension between them, so where do you want most decisions to be taken if the two groups disagree? The person to whom both functions reports ultimately functions as the tie breaker between the orgs. This is a good heuristic to keep in mind when thinking about org structure.

Hire executives for the next 12-18 months, not eternity.
As an exhausted founder/CEO, the temptation is to try to hire an executive for a role who will last for the remaining history of the company. This leads to over-hiring/ hiring someone who will likely be ineffectual at the scale you are currently operating at. For example you do not need an engineering VP who has run a 10,000 person organization when you only have 20 engineers. Instead, hire someone who has led e.g. a 50-100 person team who can scale up your org to the right level over the next 12-18 months. Either that person will grow with the team or you will need to hire someone new in the future. 

Ben Horowitz has a good perspective on this.

Please note, a stable management team is only positive for a company. However, you should realize that even if the executive team only evolves slowly over time, the org structure may still change more rapidly.

There is no perfect organization structure for a company. A company is a living, breathing thing and will change with time - as will the organizational scaffolding on which it is built. As CEO, focus on a pragmatic solution to the next 6-12 months in the life of the company rather then the perfect long term solution. In part, focus on executive bandwidth - who on your team has the bandwidth to do more and where do you need to promote or hire talent to meet the company's execution needs?

Related Posts

Wednesday, May 13, 2015

Investor Update Emails

Entrepreneurs often spend an enormous amount of time raising and optimizing who is involved in a financing round.  However too few founders tap into their investor and advisor pool sufficiently after all that hard work.  One key way to keep investors involved and excited about the company is to send a monthly update.

Below are two formats for communicating with investors.  Note that once you get to a certain stage/size, you may discontinue or limit broad investor outreach.  At some point investors can start to become a source of leaks about the company.

This slightly longer email format (1-2 pages in an email) for companies with a launched product and some ongoing growth. Investor "Asks" should always come first (as some investors won't read much past the first section in the email).  This way if you need help, investors will know how to help.  This format usually works best for a company that is farther along (e.g. Series A company, or one producing revenue).

At some point you may decide to discontinue this format as the likelihood of leaks by investors goes up the later stage the company.  You can also skip any of the items below if there are no updates.

1. Asks.
What do you need help with?  List up to 3 items people can help you with.  This should always go at the top of the email, as people might not read the rest in detail.  If no relevant asks, skip this section.

2. 1-2 Key Metrics.
Month by month or other revenue metrics for the last 6-12 months.  Include a graph and growth rates.  Mention margin or other key financial metrics where relevant.  User growth, retention, or other key metrics.

3. Team.
Who got hired last month?  Write a sentence per person emphasizing the quality of the new hire.  If you are growing rapidly (e.g. 5-10 people a month) you can instead focus on mentioning 1-2 key hires, team size, and team growth rate.  E.g. "We are now 42 people (added 5 this month) including our new head of finance, ...."

4. Product.
Milestones, successes, key future items.  Could be 2-5 lines of text.

5. Partnerships.
1-2 lines on one or two most important deals, if any.

6. Industry news.
1-2 news items, if very relevant or provides background to your company.  If nothing big, don't mention anything.  Could also be links to articles about your company.

7. Burn/cash.
Current cash position and burn rate.  How many months or runway do you have left?

8. Other
Could include team photo from a fun event (e.g. Halloween) or other misc items (if any).

You can use this format either for earlier stage startups, or for broad lists of investors if later stage.  This allows you to avoid too much detail while still getting help (investor asks).

1. Asks.
What do you need help with?  List up to 3 items people can help you with.  This should always go at the top of the email, as people might not read the rest in detail.

2. Highlights.
What went well this month?  Could be a bulleted list of 2-5 items.

3. Lowlights.
What went poorly this month?  Could be a bulleted list of 2-5 items.

In order to maintain discipline on monthly emails, schedule some time on your calendar on e.g. the last Friday of the month to write the email and send it.  Once you have sent an update once, you can use past emails as a template and it should be quite fast.

Lead VC Vetos

Thursday, March 26, 2015

Career Decisions

Whenever you make a decision on what step to take next in your career[1], I think it is worth considering the following factors.  Depending on your stage of life or career, different factors become more or less important[2].

1. Network.
I view this as the most important thing to optimize for in a job.  Network includes both who you work with, and for, day to day.  But equally important, who are the network of people at and around the company (founders, employees, investors, advisors, etc.)?

In Silicon Valley networks of people work together repeatedly.  If you fall in with the right crowd, you will have outsized opportunities over time.  Being part of the original PayPal network exposed you to companies like LinkedIn, Yelp, Tesla, SpaceX, and Facebook.  Similarly, ex-Googlers are now senior executives or VCs involved with the top companies in Silicon Valley (Facebook, Sequoia, Dropbox, Pinterest, etc.).  People who were early at Google now know people at every major company in Silicon Valley.

The people who invested in, partner with, or advise the company you join may get to know you over time and impact your career in large ways.

Falling into the right network early on usually means a career full of interesting opportunities.  I would overweight this factor as well as market/growth rate below.

2. Market & Growth Rate.
Early in your career the market trend you ride may be the biggest future determinant for your success.  Joining a company in a great market means there will be tons of companies who want to hire you, and many high growth opportunities within your own company or across other companies in the same market.  People went from Netscape to Google, from Google to Facebook, and now from Facebook to other leading Internet companies.  In contrast, people who joined telecom equipment companies in the 90's are best case still at Cisco (if they are lucky).  Choose your market wisely.

Similarly, only go to companies where you expect a good growth rate over time.  Compounding growth creates new opportunities within the company itself, but also means the company is in a good market.

Genomics is a good example of a market today with great growth potential ahead of it.

3. Optionality.
Have you been doing the same type of job in the exact same industry over and over again?  Or can you find a role that sets you up for something new?

What are the new sets of future roles this job sets you up for?  Could you work in a new market (e.g. shift from enterprise to consumer or vice versa) or new role (can you work on the product team instead of operations)?  People dramatically underweight optionality and tend to stick to doing the same thing over and over.  This becomes more important later in your career.

4. Brand.
The branding of a company matters mainly if you have never had a name brand on your resume.  If you went to MIT and then worked as an engineer at Google, people will assume you are a great engineer.  The institutions reputation will rub off on you.  Once you have 1-2 brands, each subsequent brand is less important and I would optimize for the other factors above.

1. Role.
What will you be doing day to day?  What will you be learning?  Early in your career, you should optimize for going to a high growth early stage company rather than the exact role you could get.

E.g. if I wanted to be a product manager but could only get hired on the operations team at Stripe, I would still join Stripe as an operations manager over a low growth company as a PM.  As a company expands and scales quickly, you will be given opportunities to move around or try new roles.  It may take a year or two, but if you persist you will transfer over to the product team.

Alternatively, if you are really set on a specific role or function you may want to join only a handful of companies who do it well.  In this case you are optimizing learning about a role over growth opportunities.  For example, Google and Facebook are probably the two best places to learn how to be a great product manager.  But it is unlikely you will become VP of Product at Google 4 years out of school.  In contrast, if you join a startup you may end up with a role larger then your experience level if the company does well.

If I wanted to join a rocket ship company, I would underweight exact role and overweight getting in early.  If I wanted top optimize for role, I would join the company that does that role the best.

2. Compensation.
This is the least important factor to consider.  It is probably the one people focus on most.  However, owning 1% of a crappy company instead of 0.1% of a great company is the wrong way to look at things.  The crappy company equity may literally be worth nothing.  Companies wipe their cap table more then you think.  In contrast, the great company may be worth 1000X more.

In general, I would trade equity for cash unless I really needed the cash.  While salaries tend to go up over time if a company does well, it is much harder to get substantial equity later.

If I had to choose to go to a great company with (what seems like) average compensation or a poor company with (what seems like) great compensation, I would join the great company instead.  The equity value should increase over time for the great company, and drop for the poor company.  Even if you make a lot of short term cash, you will miss out on the future network and branding factors mentioned above.

In contrast, at the great company, your cash compensation will also go up over time as the great company succeeds.  Google used to be a cash-poor, equity-rich company.  Now it pays above market cash or cash equivalents (RSUs).  Google effectively shifted from an equity to cash company over time, as break out companies tend to do.

Don't get me wrong - money is important and the ability to pay the rent and not constantly worry about your financial position is crucial.  I would just think long term about it, and also realize that salaries typically go up to market (and eventually exceed market) at the companies that do best.

Like a startup, it is best to think of your own career as a long term thing, rather then a set of short term increments.

If you plan to work in tech, move to Silicon Valley.  The set of opportunities and networks out here are much stronger.  If you do not go to Silicon Valley, go to New York.

International market experience can be beneficial in your career if you plan to be in a business role and want to primarily join late stage companies.  This is a good back door into certain high growth companies where you might not otherwise fit.

[1] This post assumes you have decided to join, rather then start, a company.  I think the decision to start a company or not is outside of the above framework.  The framework is meant to help someone who wants to join an existing company and is biased towards people early in their careers.

[2] Certain things become increasingly important as you have kids, you help family members face illness or other problems, or other life events happen.  For some people the ideal job is largely about flexibility, proximity to family or home, or other factors not mentioned above.  These factors can be crucial to a person's life and happiness, and my intention is not to underweight them.  If you want to e.g. be close to a sick relative, a lot of the other items above may become irrelevant.

Building a Recruiting Org
Recruiting Is A Grind
What To Look For When Hiring Execs
5 Signs A Candidate Just Isn't That Into You
5 People Who Destroy Your Culture
Should You Hire A COO?
Reference Check Candidates
How To Hire Great Business Development People
How To Choose A Co-Founder
How To Choose A Board Member
When And How To Fire An Employee At An Early Stage Startup
How To Fire A Co-Founder

Tuesday, January 13, 2015

Hot Markets For 2015

Being in rapidly growing markets (or ones perceived as hot) increases likelihood of success of a company dramatically[1].  Being in a hot market increases the ability to hire great people, get press and awareness, raise money, and eventually exit via M&A or IPO.

The average startup exit takes 7 years.  Market hotness increases the likelihood of a fast exit dramatically.  In the late 1990's the average time to acquisition or IPO was just 2-3 years due to Internet mania.  The fastest exits usually come via M&A.  Markets with the most natural acquirers will lead to the most exits in a segment.

To successfully IPO you usually need ~$50 million in revenue and a few quarters of profitability behind you.  If you are in a hot market, the profitability constraint may lesson and you can even loose money for a while (see e.g. Hortonworks IPO and big data hotness).

Hot Market Sustainability.

Caveat emptor - about 50% of the markets that are considered hot at any given point turn out to be false alarms.   Examples of past hot markets that turned out to largely be duds include First Wave AI (in the 1980s),  Nanotech (as an industry in the early 2000s), CleanTech (early to mid 2000s) and Geo (smaller scale in late 2000s).

Hot markets that yielded huge companies and large exits include social networking (mid 2000s -Facebook, Twitter, LinkedIn), and mobile social (early 2010s - WhatsApp, Instagram).  Some large market trends are still playing themselves out as per below including big data, sharing economy and other segments.

Hot Markets For 2015

Below is my view of both what markets are hot in 2015, as well as the likelihood of these market segments being medium term duds[2].

1. Gold Rush.
Markets That Will Yield Large Stand Alone Companies and Many Acquisitions
Big Data.
"Big data" as termed in the press has 4 subsegments[3]:
(1) Dealing with large amounts of data (Hadoop, Spark, etc.)
(2) Smart data.  I.e. doing something intelligent with the data you have regardless of the number of petabytes.  This is more analytical tools or tools for data scientists.
(3) Data center infrastructure (sometimes this gets clustered into "big data", sometimes not).  Mesos (and Mesosphere) would be an example of this.
(4) Verticalized data apps (e.g. data store and analytics for medical insurance claims).

In general this market segment has a lot of legs and will continue to create both stand alone public companies, as well as has a large number of natural acquirers.  Potential acquirers include the traditional enterprise companies (HP, IBM, etc.) as well as the earliest companies in the space who support liquid public stock or large market caps (e.g. Cloudera, Hortonworks...).  Additionally, the verticalized data companies for healthcare (and 2-3 other key verticals) will see large of exits to more specialized acquirers (e.g. UnitedHealth for healthcare data companies).

SaaS [Software as a Service - including APIs/developer tools]
As recent players with explosive growth (e.g. Zenefits and Slack) have shown, SaaS still has a lot of legs for everything from enterprise collaboration tools to HR back office management.

I would not be surprised if there will be 1-2 very large companies (or exits) created here per year for the next few years.  The key will be to find differentiated organic distribution (Slack) or business model (Zenefits).

To prevent the listing of too many submarkets, I will toss APIs/developer tools into this bucket as well.  There are lots of services that make sense as an API that traditionally have been performed in a more cumbersome way.  Stripe and Twilio are canonical examples of this trend, a more recent one.

Genomics hasn't hit the mainstream hype cycle yet.  However I think in late 2015 or early 2016 it will emerge as a hot area of investment due to the fundamental underlying shifts in the market.  I think this will yield both a large area of future investment, but also exits ranging in the $100M to multi-billion dollar range.  The genomics wave will include both stand alone genomics software companies (lots of natural buyers including IBM, Oracle, Google, Illumina, and others) as well as more traditional biology centric genomics (with large natural buyers in the pharma and traditional biotech markets).  I think a small number of large, public companies will emerge in genomics.

2. Silver Mines.
Markets That Will Yield Lots of Acquisitions, Less Clear on Independent Stand-Alones
There are two types of AI companies:
(1) Companies trying to develop general purpose AI or are trying to build a "general AI platform".
(2) Companies applying AI to solve a very specific problem or customer need (e.g. machine translation of web pages or screening pathology samples).

The first group of companies will be small to large acquisitions by Google, Facebook and handful of other companies as talent buys.  The second class of companies may yield a small number of large, stand alone, independent businesses.  I am more bullish on the prospects of the second group as truly value creating.  However, if your primary interest is fast time to exit companies in group (1) will likely sell quickly and at a good valuation 1-4 years post founding as Google and others try to stock up on machine learning talent.

IoT [Internet of Things].
IoT is a sexy rebranding of "consumer electronics and appliances".  IoT is modernizing our clunky old school devices in the home and adding software and APIs to allow for seamless interoperability and broaden logging and use of data.

Today's traditional consumer electronics and appliances remind me of the Motorola Razr right before the iPhone - great industrial design but no real use of software.

From an exit perspective large companies like Google, Apple, Samsung, Philips, GE, and others all have an interest in acquiring companies that will accelerate their own efforts in this market. So there are a lot of natural acquirers in the space.  I expect more small to large $500M+ exits in this market, but it is unclear to me which of the new batch of companies will create a long term, sustainable public company of their own.

Now that Nest is gone, I would love to hear what others think are the likely long term stand-alone companies in IoT.

This is a tougher market to crack as a startup hoping to become a massive standalone company, but I expect more startups here in 2015.  On the enterprise side there will be ongoing high-profile hackings and the need to purchase security products.  Barriers to entry in this market are higher due to the need for both a strong sales channel as well as a differentiated product, which will temper overall market momentum.  Basically, a small number of startups will have ongoing small and medium (hundreds of millions of dollars) exits, but the total carrying capacity of this industry is more limited for startups due to sales channel bottlenecks (CIOs will only want to buy security software from a handful of vendors, and too many new startups will focus on a "feature" rather then comprehensive solution).

Recently public Palo Alto Networks and FireEye will likely be industry consolidators as will other traditional enterprise security companies.

3. Roulette.
Binary Markets - Create A Few Huge Stand-Alones, Lots of Failures
Sharing Economy & On Demand Economy.
Distributed labor and work forces, or the sharing of resources will continue be a hot market from a startup founding perspective.  I think the vast majority of the new startups will fail although a handful will still emerge as big hits.  Just as Facebook, Twitter, and LinkedIn where the first wave giants in social, AirBnB, Uber, Lyft, Instacart are the first giants of this wave (by market cap).

Similarly, just as there was a second social wave that yielded break out companies (WhatsApp, Pinterest, Instagram) as well as tons of duds, the shared economy/distributed labor trend will have a few more huge companies emerge.

In general this is an area that will be fraught with lots of failures offset by a handful or truly massive outcomes.  Too many entrepreneurs will do derivative "Uber for X" for tiny markets ("Uber for sports equipment delivery").  The key will be to figure out how to capture an existing large market (e.g. Uber and transportation) or expand an existing market dramatically (Uber again) with a simple use case and product.   The people who win here will win big a they upend entire markets.

4. Tough Short Term Markets For Tech Investors?
While I am bullish on the long term prospects of crypto currencies and blockchain, I wonder if many of the current crop of companies will succeed.  A number of larger structural events need to occur for truly widespread adoption of bitcoin to occur.  Existing bitcoin companies have a ticking clock (aka burning through fundraises) relative to this market timing.  Profitable (or cash rich) bitcoin companies may make it long enough to see this transition just as AOL did with the Interent[4], but any company burning rapidly through its cash will likely fail.  Once a company succeeds sufficiently there will be a large number of potential buyers for BTC companies (including Google, Apple, Microsoft, eBay, and the entire financial system).

I expect there to be an eventual culling of existing bitcoin companies followed in a few years by a massive expansion of cryptocurrency companies when the markets are more mature.  This may be a hard slog for a few years punctuated by one or two large, misleading, exits [5].  Then there will be an explosion in cryptocurrency companies that dwarfs the current trend.   So, I am extremely bullish on this area long term, but worry about the shorter term dynamics.

Biotech Investments By Software Investors.
Outside of genomics, I have increasingly seen technology investors invest in traditional biotech companies.  While genomics has a clear "why now" statement due to its rapidly dropping costs, old school biotech does not share this big shift in market dynamics.  In my opinion this market is going to be a fiasco for tech investors as they misunderstand the industry structure (regulatory issues, IP issues etc.) as well as don't have a good sense for the underlying markets.  While biotech investors may or may not do well in biotech over the next few years (I honestly don't know the market well enough to be certain) I think a subset tech investors may end up loosing big sums of money here (similar to the CleanTech fiasco of the early 2000s).

Other markets I missed?  Comments on existing ones?  Let me know on Twitter.

Thanks to Avichal Garg for comments on this post.

[1] "Success" is defined for the purposes of this blog post as the creation of a large stand alone company or a as a large financial exit.  This is used as a proxy here for impact to the world, as "impact" is very hard to quantify.  How many lives were saved by Google?  Yet Google has transformed the world for the better by providing information access to billions of people.  It is hard to come up with a good metric for doing good for the world.

[2] Like all prognostication, I will undoubtedly get a bunch of this wrong.  This is just my current view of the world, and is obviously subject to change as more data gets generated by that wonderful physics simulation software that we call reality.

[3] From a founder perspective.

[4] AOL is a similar example for the Internet.  AOL was founded in the 1980s and managed to work out an existence until the early 90s, when the bigger Internet wave really hit.  By the late 90s AOL was one of the largest companies in the world by market capitalization.  The next wave of Internet companies were founded a decade later then AOL, when enough infrastructure (markup, browsers, more physical wiring upgrades) allowed for the real Internet boom to occur (Amazon, eBay, Yahoo!, Google, etc.)

[5] Big companies often make large, stupid buys for "strategic" reasons.


M&A Ladder
What Is Your Startup Acquisition Really Worth?
5 Reasons To Sell Your Startup
Acquisition Blues

Saturday, January 10, 2015

The 3 Types of "Platform" Companies

People use the word "platform" to describe products with fundamentally different characteristics.  OSs (e.g. Android), infrastructure products (e.g Twilio), and platforms (Facebook APIs e.g. Connect) may all be called "platform".  However, the distribution approaches and product strategy for each differs. Conflating what makes a platform work versus e.g. an infrastructure product can backfire and cause a team to have the wrong strategy for building a product or getting customers.  These startups tend to fail.

Below I attempt to define and differentiate between these different types of companies and their products.

1. Infrastructure.
Infrastructure products are ones that multiple companies have to build over and over again.  Eventually some smart entrepreneur realizes this and builds the common infrastructure product that other companies will pay to use.  An example of this is the founders of Mailgun, who built versions of the same email server for multiple employers until they realized they could build this as a general service for all developers.

Infrastructure products are often necessary for a product to function (every ecommerce site needs Stripe for payments) but are not often a "strategic" differentiating buy for their customer (although Stripe has managed to differentiate strategically based on its fast iteration on new features and its simplicity as a product).  Early on, many users of Twilio didn't care if they were using Twilio or another telephony provider - they just want it to work quickly, simply, at a good price (which ultimately meant using Twilio due to its ease of use).

The best infrastructure companies have clear economies of scale or network effects.  Twilio is probably able to negotiate better and better deals with carriers on pricing the more volume it aggregates from its customers.  Similarly, large amounts of payment data can provide scale effects for fraud or risk management.

An infrastructure company's success often boils down to a handful of factors:
-Ease of use and integration.
-Up time.
-Differentiated features or historical customer data.  This helps you lock in your customer base.
-Economies of scale.  This can lead to network effects on costs (pricing power of the infrastructure provider relative to its own suppliers) or features (fraud detection).
-Developers or sales channel.  In some cases a developer ecosystem emerges around an infrastructure product (note: this is different from developers using or adopting a product).  This is less common for infrastructure then people think, and is more common for a true "platform" (see below).

In rare cases, an infrastructure company can move up to become a "platform" in its own right.  This only works if the infrastructure company is able to collect and re-position unique end user data, or build direct brand recognition with its customer's customers.  Platforms typically have more lock-in and differentiation then infrastructure, so moving in this direction if possible can be a strong strategic move.

2. Platform.
A platform almost always grows out of an existing vertical product (e.g. Facebook, Twitter, etc.) and is ultimately a generalized extension of some aspects of that product (e.g. Facebook's internal user graph / identity and Facebook Connect.  The resulting platform allows third party developers to take advantage of the unique data, services, or userbase of the original vertical application.

A platform is not easily commoditizable.  Only Facebook has the social graph that underlies it, or the ability to drive distribution of certain types of content to a billion users.  If a developer tried to use another social product's API (e.g. Foursquare) instead of Facebook's, they would not get access to the right type of data or the same distribution.  Alternatively, their own customers would not want to use the "Login with Viddy" button.  In contrast, a company could probably swap one infrastructure product for another without a fundamental change to how its own application functions.

Almost always, a platform is valuable due to some unique characteristic of the original vertical product from which it grew.  For example, Facebook Connect worked in part because Facebook itself was a representation of a person's identity.  It was natural for consumers to feel comfortable logging into other sites with their personal identity.  Contrast this to federated approaches like OpenID, to which the user had no brand association.  These types of "build it and they will come" approaches to platforms tend to fail (in reality, you are building infrastructure in this case, but calling it a platform, but with no user recognition or branding for your product which comes from being an actual platform).

A platform usually has the following characteristics:
-The vertical app the platform is based off of owns the ultimate "end user" directly.
-The core functionality and key features of the platform are derived from the vertical app that spawned it.
-Provides proprietary, non-commodity data and/or distribution to applications using it.  The word proprietary here is key.
-In some cases, provides a monetization mechanism for apps on the platform and almost always takes a revenue share.  (Contrast this to infrastructure, where the infrastructure provider is paid for use of its product).

Many entrepreneurs I know set out to build a "platform" without any real vertical application underlying it.  In reality, they are building infrastructure.  Most companies that confuse these two things tend to fail.

The key way to tell if your "platform" product will fail is if you need to build the first "killer app" for the platform yourself for your platform to succeed.  In other words, you end up trying to build both a vertical application and a platform simultaneously.

3. Operating System (OS).
This is a pretty reasonable definition of an OS.  In general, adoption of an OS is driven by the following:
-The hardware the OS is typically bundled with gets a lot of distribution.
-There exist (or quickly emerge) a small number of killer apps that differentiate the OS causing more distribution and adoption (e.g. spreadsheets and the early PC market).
-An app ecosystem emerges around the OS, which creates a positive feedback loop.  The more users on an OS, the more people develop apps for it, the more valuable the OS becomes to users.

In general, OSs seem to follow two phases of adoption:
Phase 1: A combination of the hardware plus a small number of killer apps drive OS adoption.  For the early PC operating systems this was largely spreadsheet applications like 1-2-3 and Excel.

Phase 2: Once the OS has strong adoption, the longer tail of apps is created by the developer ecosystem who want access to paying users.  This locks in users or spreads OS use to a new set of consumers.  After the spreadsheet word processing / desktop publishing and gaming helped to spread adoption and value of PCs.

One could argue that a platform is its own killer app first and foremost.  For example, the killer app on the Facebook platform is really Facebook.  Once Facebook got adoption for itself, other non-Facebook applications followed.  This is the primary way a platform product has similarity to an OS.

Who Cares?
The reason these distinctions are important is that the strategy for building a successful Platform is different from building a successful Infrastructure company.  Many people confuse the two and pursue the wrong approach as a company.

Signs Your Infrastructure Company Is Off To A Bad Start
Many people call their infrastructure company a "platform" and decide that all they need to do is find a killer app as a customer to drive their own adoption (since the overall market they are gunning for is still hazy and unclear).  This type of company is usually started by a non-market driven technologist, who thinks a new technology is really cool.  Unfortunately, most of the companies that start off this way end up as small acquisitions at best.

The reason is three-fold:
1. The company is not starting off with a market problem.  In the case of a company like Stripe or Twilio, the founders were trying to solve a problem other developers such as themselves faced over and over again.

2. The market may be too small.

3. The "killer app" you are seeking is where all the value in the industry comes from.  If the killer app truly took off, it should be able to launch the true platform in the industry and drive you out of business.  Unless you are a piece of infrastructure.  In which case, where are your customers?

Related Posts
Defensibility and Lock-In: Uber and Lyft
Uber And Disruption
Who Cares If Its Been Tried Before?
The Road To $5 Billion Is A Long One
How To Win As Second Mover
End Of Silicon Valley
Social Products