Benjy Boxer's Blog

Two weeks ago NewsCred launched the NewsRoom.  We dedicated significant efforts to outreach and PR.  We were able to capture data on where our applicants heard about the product.  This demonstrates that directly emailing people was the most successful technique for reaching our writers.  It was even more successful than all of the traffic that originated from the 3 very positive articles that covered the launch.

All data used in this post can be found here

The success of a media company depends on its ability to deliver its content to a receptive audience.  Distribution is an extremely important part of any content strategy, and it determines which sites grow into successful media brands. 

The Huffington Post was worth $315mm because it became the master of delivering content via search engines.  BuzzFeed and Upworthy will be more valuable than the Huffington Post because they’re mastering a newer and more potent content distribution channel – Facebook.  Building a successful content business depends on adopting a unique distribution strategy and focusing on that one strategy.

Many content sites try to optimize for different methods of distribution, but to be successful, a company must place a bet on a specific strategy.  Older content sites trying to adapt to a new distribution channel face an Innovator’s Dilemma.  The older content companies are optimized for specific channels, but when a new one emerges, they face a choice – abandon the old channel for the new or continue to grow revenue incrementally through the current channel.  As the Innovator’s Dilemma predicts, most choose the latter.  This leaves a gap for new content companies to grow and take advantage of the emerging distribution channel. 

The changing landscape of online media brands is an example of the Innovator’s Dilemma at work on content distribution. The Huffington Post, which launched in 2005, was built to deliver content through Google Search.  The technical engines that power the Huffington Post are extraordinary at optimizing content for search engines.  Whereas, Upworthy and BuzzFeed, which were founded during the Facebook era, continue to optimize their site and content specifically for social sharing.  The Huffington Post is falling victim to the Innovator’s Dilemma as growth slows and its customers find content through the increasingly dominant Facebook channel.  

*Huffington Post data starts after the merger with Aol

Download the data file here

This trend repeats whenever a new distribution channel emerges. Newspapers built on delivering dailies via local newsboys were disrupted by regional media companies that built sophisticated distribution networks based on cars, trucks, and highways.  These media companies were then disrupted by newer publishers that focused on distribution via the internet and later via search engine optimization.  As evidence of this, the Washington Post and Boston Globe (media companies built on sophisticated regional delivery networks) were both sold for much less than the Huffington Post.  This trend is also happening in TV production.  Since the 1960s, TV producers have hitched their product to distribution via TV channels, but as Kevin Spacey recently pointed out, this is changing.

The company focused on search engine optimization, the Huffington Post, is losing ground to those that have been built to distribute their content via Facebook.  You can see the different distribution strategies of these companies via their incoming traffic sources (data from Comscore).

*Adjusted for Aol incoming traffic

The data clearly demonstrates how these popular sites distribute their content.  The faster growing sites (Upworthy may be the fastest growing media company ever) are dependent on Facebook as a distribution channel.  Based on internal data, these numbers may be even more extreme.  When Aol bought the Huffington Post, internal data showed that 40% of the Huffington Post’s traffic was coming from Google.  In a conversation with a BuzzFeed executive, I was told that almost 40% of their traffic comes from Facebook.  I’m willing to bet that Upworthy gets an even higher percentage of its traffic from Facebook than the 40% reported by Comscore.

It’s also interesting that, of the companies analyzed, none have a significant amount of traffic coming from Twitter or LinkedIn.  I’m guessing that the next great content company will succeed based on optimizing for distribution via Twitter, Linkedin, Instagram, Snapchat, or some future distribution channel.

My experience is limited to the time I spent on the corporate development team at Aol and anecdotal conversations, but I want to share what I’ve learned about the process of being acquired.

If you want to be acquired by a large company, your number one goal should be meeting with the senior manager who will sponsor the acquisition and getting her to trust you.  You need to have a strong relationship with this senior manager and convince her that you and your company will help her achieve her goals.

Of all the angel-funded companies (already a very select group of companies) that received funding between 2004-2008, less than 0.7% were acquired between 2008-2012 (these are very rough estimates based on combining two datasets from the NVCA and Center for Venture Research that may not use equivalent criteria).  The odds are against you, so you need to know the best path to acquisition is through trusted relationships at a strategic buyer.

The path to acquisition at Aol was similar for almost every company we acquired between 2010 – 2011 (that’s when I worked there).  During this time, Aol acquired 9 companies.  While I was on the team, we executed and/or integrated 7 of these acquisitions.  Additionally, we came close to acquiring 3 or 4 other small companies, but didn’t close those transactions.  The two acquisitions I worked closely on were Pictela and the Huffington Post.  These two, along with all of the other acquisitions, followed a very similar path.

At Aol, there is a senior management team that is responsible for specific products or brands at the company.  In order to be acquired by Aol, a startup needed one of these senior managers to be its internal champion – the one exception being the Huffington Post, which due to the transformative nature of the acquisition and its impact on the organizational structure of Aol was championed by Tim Armstrong. 

It’s absolutely essential that the CEO of a startup understands how important this internal champion is to the acquisition.

Essentially, at the time of the acquisition, the senior manager must vouch for the acquisition and commit to a budget that your little startup is supposed to affect.  For this reason, the senior manager must believe that you and your company aren’t as risky as they are and that you’re either going to help her achieve her goals or make her job more difficult if you remain independent or are acquired by a competitor.

Of the companies we acquired, I saw four strategies that seemed to convince the internal champion that a company deserved her trust.

Companies made themselves known as thought leaders in the industry.  The founder spoke at conferences and made sure that his product was featured at those conferences.  Sometimes, if you get a senior manager to like you enough, she can help get you featured at the conference.

Founders began a conversation with the senior manager as a business deal and made her realize that she needed the company to drive innovation internally.

Founders networked and established relationships with the senior managers at every potential acquirer.  Senior managers are similar to VCs - they’d rather invest in lines than dots, and they gossip about companies and products.

The CEO convinced the senior manager that she needed him to lead her new initiatives and execute on an innovative plan within the company.  This is a typical path for an "acquihire".

Once you get the senior manager excited about an acquisition, it’s almost impossible for a corporate development team to derail it.  If the senior manager believes you’re the key to her success, she will push the acquisition process and the corporate development team to execute it. 

There are times where diligence stops an acquisition, but this is almost always the fault of the entrepreneur not setting realistic expectations of the state of his business during the initial conversations with the senior manager.  A good corporate development team will quickly find where your business problems exist.  If you set the expectations properly prior to the signing of a Letter of Intent (LOI) to acquire the company, there won’t be a surprise later.  Even if sharing these expectations result in a lower valuation, it's much better to share that information upfront.  I assure you the corporate development team will find the problems and lower your valuation later or terminate the process.  Surprises are very bad during an acquisition; it means you just violated the most important part of getting acquired – the senior manager’s trust.

Simple stock option calculator – I used this for all of the calculations in this blog post

Several of my friends have been considering job opportunities at startups in New York.  Typically, their job offers have included below-market salaries and option grants.  When I chat with them about their offers, they’ve been attributing way too much value to the stock options offered.  I’ve generally explained to each of them why stock options are nice, but are worth much less than may think.  I don’t persuade them against accepting the offer because I think you should be willing to take a pay cut to learn and do more at a startup than you would learn and do at a big company.  I just tell my friends that they should never optimize for options, but optimize for an experience.  Mark Suster wrote something similar to this in a post – Is It Time for You to Earn or to Learn.  Throughout the rest of this post, I’ll explore the math behind the value of a stock option in a seed-funded startup and explain why those options aren’t as valuable as you may think.

Many companies convince engineers and other talent to join those companies at a below-market salary in exchange for stock options.  I believe stock options are valuable, but only in that they make everyone an owner of a company.  Owners work harder and put up with more to make a business successful.  I brought up the fallacy of the value of stock options at a VC-panel, and was quickly shot-down by this panel.  Since they didn’t answer my questions, I thought I’d do some research and present the mathematical case as to why the monetary value of an option is much less than you might expect.

For simplicity, I made some very important assumptions (below) that anyone who is reading this should be aware of.  Generally, I tried to choose the assumption that would make an option more valuable at the time of an exit to over-emphasize the value of stock options.  All of the calculations in this post can be found in this file – you can also download and customize it for your own use.  I invite you to change the assumptions to fit your specific situation.  On a side note, please let me know if you find any errors in the file – thanks.

I’ll quickly explain why your options aren’t worth very much even if you’re one of the top 10 employees at a seed-funded company.  You would be very lucky to receive much more than 50 basis points of a company (0.5%) when you join after a seed round.  That 50 basis points will be diluted by future funding rounds.  At the time of the unlikely large M&A acquisition or IPO, you will have to buy your options at the valuation those options had when they were granted (the strike price).  Additionally, since you’re going to exercise those options and immediately sell them, the gain in value will be taxed as ordinary income.  Finally, this all assumes that you will stay with the company that granted those options for more than 4-years.  You will lose any options that didn’t vest if you left earlier.  How many of you have worked for a company for more than 4 years? 

If you assume the company you work for is the average seed-funded startup that has an average M&A exit (defined by the assumptions listed below), your 50 basis points will be worth approximately $179,000 (after taxes) in 4 years.  Unfortunately, the probability of this happening is low – in fact, it’s about 10% for seed-funded companies.  So the expected value of that option is really only $17,600 in 4 years or presently valued at $10,100 with a 15% cost of capital. 

Now, let’s assume you really hit it big, and your company IPOs in 7 years.  Your 50 basis points at an average seed-funded startup with an average IPO (again, defined below) would be worth approximately $2,354,000 (after taxes).  Again this is a very rare outcome.  It only happens to about 1.4% of seed-funded startups.  So the expected value of that big payday is only $33,700.  Unfortunately, an IPO usually takes longer than an M&A exit, so you’ve now held onto those options for 7 years.  The present value of those options with a 15% cost of capital is approximately $11,800. 

So your company has hit the jackpot; congratulations!  You’ve probably done more and learned 1000x more than you would have if you chose a different job.  On the flip-side, despite that exit, those options you were counting on changing your life aren’t worth as much as you dreamed.  In the IPO case, you’ve just taken home approximately $2mm.  That’s a huge payday, but the expected value is so much less because of how rare this is.  If you combine the present value of an M&A exit and an IPO, the expected value of the options granted on your first day are $21,900.  You would be indifferent if you accepted a pay raise of approximately $7,700 for the next four years. 

Again, I believe options are very important to generating a great culture in a startup where everyone is an owner and feels strongly about accomplishing the mission of the company.  Just don’t accept a big pay-cut for those options.  Accept that pay-cut because you’re about to embark on a rare opportunity to grow your career at a startup.

Assumptions and Important Data

Important Data:

The number of seed funded companies from 2000-2012 and the average investment in those companies was obtained from the most recent PWC MoneyTree Report.

The average IPO market valuation, age of a company at the time of an IPO, and the number of companies that IPO each year from 2008-2012 were collected from the NVCA 2013 Yearbook.

Investor Assumptions:

At each stage, the VC investment has a 1x liquidation preference and has participating preferred equity.

None of the VCs in a company receive anti-dilution protection.

At the time of investment, the seed investor receives 30% ownership; the 2nd round receives 25% ownership; the 3rd round receives 20% ownership; and the 4th round investor receives 15% ownership.

Option Assumptions:

The employee receives 0.5% of the company in the form of 4-year vesting stock options at the valuation following the seed investment.

The employee fully vests prior to an acquisition or IPO.

The employee’s effective income tax rate is 40%.

The employee’s cost of capital is 15%.

The employee doesn’t receive another option grant in the future.

The employee exercises his stock options 6 months after the IPO – Fyi, most companies lock-up their option holders for 6 months following an IPO.

M&A Assumptions:

An average acquisition happens 4 years after the seed funding of a company.  Fyi, I had to guess this number.

Prior to an acquisition a company receives 2 rounds of financing similar to an “Early Stage” and “Expansion Stage” investment as defined by the MoneyTree Report.

The average disclosed M&A deal is worth $134mm.  This is way higher than reality because of right-skewed data as well as the majority of unannounced M&A deals being on the small-side.

IPO Assumptions:

An average IPO happens 7 years after the seed funding of a company – NVCA 2013 Yearbook.

Prior to an acquisition a company receives 3 rounds of financing similar to an “Early Stage,” “Expansion Stage,” and “Late Stage” investment as defined by the MoneyTree Report.

The average market valuation of a VC-backed IPO on the first day of trading from 2008-2012 was $1.44Bn – again, very right-skewed thanks to Facebook and a couple other huge IPOs in the past couple of years.

The Stratasys acquisition of Makerbot demonstrates that 3D printing is quickly gaining momentum.  To help you visualize the importance of this merger, let's use an analogy to the early days of the computer industry.  Makerbot is like Apple (the leading personal computing company) and Stratasys is like IBM (the leader in mainframes at that time).  If that's true, these big players are preparing to dominate the New Industrial Revolution.  Unfortunately, a story that isn't being told is the impending shortage of people who can actually design the products that will be printed.

Allowing individuals to create and design their own products will enable the long-tail of things similar to the long-tail of content on the internet.  There are literally millions of blogs and tens of billions of webpages.  When 3D printing reaches maturity, there will be a similar explosion in the availability of long-tail products as there is long-tail content.

If you pay attention to the current job situation in the US, it's easy to recognize a very strong trend.  If you can 'code' (i.e build a website/web application), you possess one of the most valuable forms of human capital today.  There's an epic shortage of software developers - that's why some are getting paid millions of dollars to stay in their current jobs.  According to the 2010 US Census, there were 1.3 million software developers and programmers and 458 thousand database and system administrators and network architects.  These numbers are obviously imprecise and probably don't capture the total number of 'hackers' and 'coders', but they provide some directional information.

That same census reported that there were 41 thousand industrial and commercial designers.  If that's true, there's is going to be a drastic shortage of people who can actually turn the nascent 3D printing industry into a personal and local manufacturing revolution.  Again, it's important to recognize that there are plenty of people not counted in the census with these skills, but this directional information is alarming.  

I recently read Give and Take by Adam Grant.  I took a lot of notes, and copied and pasted a lot of quotes from the book.  I’m sharing them for anyone interested in the lessons from the book.  I highly recommend that you read the book.  I learned a lot from this fantastic book.

Fyi, these notes poorly organized and were primarily written on my iPhone.  The notes are in order from page 1 to the end.

Several friends have asked me how I would’ve approached the founding of Arkad today to prevent the mistakes that we made in the past.  Hindsight is 20-20, but it’s good to reflect and think about some possibilities.  These two examples are specific to Arkad, but if I were to start another enterprise SaaS business, I would definitely focus on launching with these two ideas in mind.

Start as a service business –Arkad was an enterprise software-as-a-service (SaaS) product.  The minimum viable product for a SaaS business is actually providing the service manually.  Many people shy away from this (as we did), but I think we could’ve learned a lot about our business if we’d started providing investor relations services manually. 

With real, paying customers from the start, we would’ve known what features to focus on.  While I took the lead on providing this service, Micah and Kenny could’ve focused on productizing the most important aspects of the business.  Initially, the products would’ve helped me provide the services better, but eventually, they would’ve replaced me.  We didn’t learn that we were building the wrong features until we released our product. Micah and Kenny did an amazing job getting the MVP of Arkad built in four months, but during those four months I could’ve been learning from customers by serving them rather than just pre-selling our product to them.

Focused on the need that matches the incentives – Through our service business, I think we would’ve learned exactly where incentives met our customer needs.  In my opinion, a deal flow product that would help investors identify investments before other VC firms would’ve matched the needs and incentives of our first customers.  This deal flow product would automatically alert investors that a company they met with was gathering momentum.  If you ask many VCs where their best deals originate, they’ll tell you that those deal are proprietary (i.e. they’re the first to hear about the deal and there is some connection to the firm).  If they could identify the companies that were getting momentum before other VC firms, they may increase their ability to make an investment before other firms were aware of the round.  That initial product would’ve perfectly aligned with our customers’ need and incentive scheme to find and invest in the best deals.

I support the Crowdfunding Exemption in the JOBS Act, but I believe many people have miscalculated the greatest beneficiaries of this exemption.  Since the JOBS Act was passed almost 15 months ago, there have been many claims that crowdfunding will help retail investors participate in huge financial successes like Facebook and Microsoft.  I disagree.  Crowdfunding isn’t suited for investing in technology companies.  In my opinion, the greatest beneficiaries from crowdfunding will be local businesses and the crowdfunders who support them.

It is really difficult to successfully invest in technology businesses.  The majority of professional venture capitalists have failed to return enough capital to justify these risky investments.  In fact, VC fund returns are pretty abysmal when you consider the amount of risk they take.  According to the Kauffman Foundation, “the average VC fund barely returns investor capital after fees.”  Only 22% of funds returned enough capital to investors to sufficiently reward them for the risks associated with venture investments.  As GigaOM recently noted, most of these returns are earned by one or two venture capitalists in each era of technology investing.  During the social media boom, two investors, Union Square Ventures and Spark Capital, have dominated the business.  If the majority of VCs have such a difficult time returning capital to their investors, crowdfunders can't be expected to outperform them.

Furthermore, the best technology companies easily raise the capital they need.  Preqin released a report that, as of February 2013, North American venture firms have approximately $64Bn to spend on investments.  That means that there is plenty of capital on the sidelines to invest in worthy technology companies.  They are looking for places to invest, but haven’t found them.  This money sits on the sidelines for many reasons.  Generally, the cost of starting a technology business has dropped precipitously over the last decade while investors aren’t finding enough businesses whose potential return is outweighed by the investment risk.

On the other hand, local businesses have been struggling to raise bank debt since the financial crisis.  From 2008 - 2011, small business lending (loans of less than $1mm) dropped by almost 22%.  In 2011, small businesses received approximately the same amount of capital from banks as they did in 2006.  Thankfully, more recent surveys from the SBA are pointing to a better environment for small businesses.  In Q4 2012, business lending increased for the first time in 10 quarters.  Unfortunately, you still hear many stories of banks cancelling credit to small businesses without considering the specifics of each business.  Inc. Magazine highlights a few examples of banks cutting loans to successful, credit-worthy businesses.  Right now, there is a lot of mistrust between small businesses and their lenders – this shouldn’t be the case.

Crowdfunding campaigns are perfect for financing the gap between the needs of small businesses and the credit lenders are providing.  These are still very risky investments, with a nearly 30% delinquency rate among small business loans in Q1 2013, but they are much safer than technology investments.  The capital invested through crowdfunding should be structured as a loan (since there will be almost no chance of a liquidity event in these small businesses).  Additionally, the small businesses that successfully repay their loans should benefit through credit rating improvements.

Crowdfunding small businesses will work at the local level because Kickstarter and other crowdfunding platforms have proven that people are willing to make investments, and they do so at an overwhelmingly local level (i.e. the crowdfunder knows the fundraiser).  For making these investments, crowdfunders should be compensated for the risk they assume and should have information rights like bankers do.

The SEC and FINRA are taking their time with creating the regulatory environment for the Crowdfunding Exemption.  We should all be happy that they are.  The rules will define the success of crowdfunding.  The minute someone feels that fraud has occurred, it will perpetuate a culture of mistrust that will be difficult to overcome.

This past week, I spent time building the base functionality of my first app.  Users can now sign up, sign in, and sign out.  I am focusing on the functionality and the structure of the backend for now.  Most of the CSS is basic bootstrap.  Once the functionality is all setup, I will start working on the javascript and css.  

Next, I will allow users to edit their information.  Following that, I will create the race model with a relationship to the users through a join table.

On a side note, one very frustrating problem I kept running into was my SQLite3 database kept throwing a busy signal exception (ActiveRecord::StatementInvalid: SQLite3::BusyException:).  This started when I was using my console in sandbox mode to check why I could not access a column of my user table in rspec.  Rspec kept giving me a NoMethodError.

I tried a couple steps and finally got the SQLite to stop throwing the exception.

I shut down the rails server, the rails console in sandbox mode, and Spork.  I restarted these, but it did not change anything.

I migrated my database, but nothing.

I entered: ActiveRecord::Base.connection.execute("BEGIN TRANSACTION; END;") into the rails console in sandbox mode and regular mode. - this did not work.  I received this error - ActiveRecord::StatementInvalid: SQLite3::SQLException: cannot start a transaction within a transaction: BEGIN TRANSACTION; END;

Then I tried ActiveRecord::Base.connection.execute("END;") - Again...no luck, but it did look like something was going to happen, which got me excited.

I restarted my computer.  This let me keep working, but I quickly ran into the problem again while I was working in the rails console.  This time, I decided I needed to find a solution because I could not keep restarting my computer.

Finally, I entered rake db:reset in the terminal window.  This worked!  Huge relief because I was on the verge.

I recently completed the Sample Application in the Rails Tutorial.  As practice, I've also created a few more rails applications that render mostly static pages.  You definitely don't need rails or ruby to create these applications, but I found it helpful to practice some of the skills I learned from the Rails Tutorial.

I built a marketing page for Arkad, which we used for some inspiration for the actual marketing page.  There are sections of this page that don't work, and there are blank images as placeholders.  The app was just inspiration for the real thing.

I also built a marketing page for my friends' business.

Now I am building my first Rails application from scratch.  The application will help me train for my next marathon.  My plan is to implement these high level functions:

Sign up and state a marathon goal

Based on my goal and my recent runs, I will create a personalized training program for myself

From that marathon goal, I will send a daily email or text message to remind myself what the exercises are for the day

I am not sure how long this will take or what I will get done, but I am excited to learn while creating something.  

I started to learn some coding skills last year when I followed Code Year on Codecademy.  On Codecademy, I learned basic html, css, and javascript.  I used these skills to help very simple parts of the Arkad App (none of the work I did still exists.  The current version of Arkad is all Kenny and Micah).  To keep up with the development at Arkad, I finished Rails for Zombies.  After that, I started the Rails Tutorial.  

I am not trying to be a professional programmer, but I want to have the ability to communicate intelligently and build basic tools that I find useful.  

I've spent a bunch of time trying to learn D3.js for my web app.  First of all, D3 is a beast, but amazing.  There are also some amazing examples and tutorials on the d3.js github.  

I've been focusing on animated donut charts.  The first place to start is the Pie Chart I example on the d3 documentation.  Then follow the link to the animation page.  Once you understand those, if you want to put labels in the chart, I recommend checking out this circle with labels page.