Accelerated Decline: Network Effects on FB’s Way Down

To start, this is not another Facebook-is-more-ephemeral-than-Snapchat, look-at-how-good-I-am-at-wordplay post. Also, Facebook is not dead. Just look at their mobile monetization success.

However, when Facebook loses part of a group of people, the rest of the group will disappear simply because the first part did. Sounds similar to the network effects that got those people to join. This is network effects on the way down. Some people did not join because of the single player utility that Fred Wilson talks about here. These people only saw enough value to be worth joining once their friends were on. Put differently, they only saw multi-player utility.

These same people will also not see enough value once their friends leave. Thus viral decline.

My sister is a high school senior. She does not have a smartphone. Nor does she read TechCrunch. She does not know about the social unbundling that mobile has caused and the phone’s contact list and blah blah. She just knows that her friends have left Facebook and she sees very single user utility. I was surprised that she left FB even though she does not have a smartphone (mobile is widely considered to be the Facebook killer).

There was a recent study that modeled Facebook’s rise and fall like a virus. Of course, they used google searches as a proxy for “infection”, which is flawed. Without getting into the statistical issues of the study or the counterclaims, it raised the question, “What does a social network decline look like?”

I think it looks like my sister who stopped using Facebook even though she personally has not replaced it. Her friends were her main source of value and they left Facebook.

Facebook is far from dead, but the network effects that gave them such high enterprise value and customer lock in, could make them decline faster once the ball starts rolling.

Smart Playlist: We Will Program Everything

I just used iTunes smart playlists for the first time and it has vastly improved my iPhone music experience. Not coincidentally, I have also been learning how to program.

My music flow is to download a bunch of new music and go through my recently added playlist a bunch of times. Eventually I find songs that I like. Sometimes I put them into playlist that sync with my iPhone. Sometimes they do not, but I have way too much music to sync my whole library.

My “Recently Added” playlist syncs with my phone. I find that I rely on it quite a bit. One of the most annoying things is when one of my new favorite songs becomes no longer “Recently Added”.

I just started using smart playlists. I setup rules to define my own “Trending” songs. There are a few steps, but if a song has been in my library for a certain amount of time and has accumulated enough plays, it makes the list. The longer something has been in my library the more plays it will need to stay in “Trending”. Now, if I like a song, it will stay on my iPhone even if I do not sort it into a playlist.

For some reason, it took me a while to set this up. The more comfortable I get with programming, the more I see ways for programming and technology to solve my problems. I know many people who do not use smart playlists who would clearly benefit from it. Why?

Fundamentally, technology only works when it understands human psychology. For now, most non-programmers would be prohibitively uncomfortable writing basic algorithms. The confusion costs and steep learning curve make people ambivalent to try programming. That is not going to change with a new app. It will change when the human attitude changes.

In the future, as more jobs depend on technology and as more wealth accumulates in the tech sector, people will gravitate towards technology. It is hard to accurately predict when, but at some point, learning to program will be much more common and programming literacy will be widespread.

I wonder what technology will look like when people feel comfortable programming the last mile of applications and when everything can becomes custom. I might well not be alive. Human psychology is much harder to rewrite than machine’s. However, it will happen.

How I Wrote This Blog Post Efficiently

Do one thing at a time. The idea is so simple, but the execution is so difficult. I always have a lot of things I want to get done. However, I do not always get a lot of things done.

I think the appeal of “multi tasking” is twofold. The first reason is the following broken logic: Very successful people are often doing many things. They must be able to balance them all at once. If I want to be successful, I must posses the same super power of multi tasking. The second appeal is that when I hit a speed bump in one activity, I can switch activities and find more low hanging fruit there. I can work on 5 activities and go at “full speed”. Eventually, I will go back to my first activity and go through the time consuming and slower moving process of sorting out whatever road bump caused me to do something else.

The first point is not valid. Simply, it is not necessary to do many things at the same time. The ability to be involved and successful in many things is about doing each activity as efficiently as possible and not necessarily simultaneously. Multitasking is not efficient, which leads me to my second point. If I have to produce 3 papers and each is 10 pages, I need to write 30 pages. If the background for each paper is the easiest section to write, then I can write the most pages in first two hours by just writing the background for each paper. Then I go back to the paper that is the easiest and write until it gets hard enough that I switch to an easier part of another project. I am optimizing for highest impact first, as measured by pages per hour. However, my end goal is to optimize for tasks completed per day or, getting meta, goals per life. My multi tasking method makes sense if I all I want to do is move as fast as possible for as long as possible.

In high school, I had a friend who took a very long route to school, but it was empty highway the whole way. He got to drive very fast and he never felt like he was wasting time. Of course, if he took a more direct route and just spent some more time in traffic, he would have gotten to school faster and burned less gas. Consistently, I am the most productive when I start a task and do not stop until I reach completion or a preplanned milestone, regardless of how difficult it gets.

And yes, I simply sat down and refused to get sidetracked while I wrote this post even when I hit bumps or saw push notifications for an activity with less friction.

Watch the Throne: Fred Wilson’s A VC Blog Examined with USV’s Thesis

Union Square Ventures is a very successful venture capital firm and they make thesis driven investments. There are many great companies that make bad venture capital investments. USV’s thesis is a great filter to sort out between good companies and good investments. One of the partners, Brad Burnham best summarized the thesis as this:

“Large networks of engaged users, differentiated through user experience, and defensible through network effect.”

I won’t try to explain that because one of the other USV partners, Andy Weissman already did a great job of that here. Instead I will use the thesis to examine Fred Wilson’s (another USV partner) very popular blog: AVC.

AVC is one of my favorite blogs. It is also one of the most popular blogs on tech entrepreneurship and venture capital. Arguably, it provides USV and Fred with a competitive advantage. Many entrepreneurs are familiar with it and value all of the advice they have gotten from Fred’s blog over the years. It might help USV get to invest in some very competitive deals. They can also avoid the need to try and buy their way into a company’s capitalization table with a higher valuation. That is a competitive advantage for VC’s.

Let’s examine AVC through USV’s thesis piece by piece.

There is a large network of engaged users. The blog gets a lot of pageviews per month (around 100,000) and most of Fred’s daily posts get at least 100 comments. I think the disqus comment platform really helps keep the network engaged.

“Differentiated through user experience” is the next element. AVC adds a lot of value to readers. Fred does a lot of cross-linking to his other posts and to other worthy blogs. Many websites try to keep you around to read more articles. They use “most popular” sections and have clickbait titles. I always feel like I wasted time when I spend 30 minutes on those sites. When I spend 30 minutes on AVC, I feel like the content adds value and the linked articles appeal to me based on what I am currently researching or thinking about. Good content and thoughtful linking leads to a good user experience.

The final element is “defensible through network effects”. Here comes the Watch the Throne part: AVC is on top, but does it have something defensible that will keep it there? Enter network effects. All of the other viewers of AVC make my experience better. I love to read the comments. Disqus does a good job of letting a thoughtful conversation evolve. The large number of thoughtful commenters improves all viewers’ experiences. In addition, the likelihood of sparking a stimulating discussion also motivates people to comment on the AVC blog as opposed to another blog with fewer engaged readers / commenters. AVC is a community that has formed a habit of checking and commenting on AVC everyday.

There are also network effects from the people Fred knows at USV and beyond. He links to them, has them guest post and can provoke very interesting discussions across the internet. However, many other successful VC’s also have a network of intelligent people and it does not differentiate Fred as much. Offering a potential bridge between the AVC community with the USV community through the new USV.com may help create further network effects. That is one reason I am optimistic about the USV conversation section.

To summarize, AVC has a large network of engaged users who get a great user experience of high quality content, relevant links and insightful conversation. There are defensible network effects from the conversations on AVC that motivate people to read and participate in thoughtful conversations following Fred’s posts. Finally, the USV conversation has the potential to connect the USV network with the AVC community (there was already a lot of overlap) leading to greater network effects and defensibility. AVC seems like a good “investment” based on USV’s thesis.

Negotiating More Than You Bargained For

There are two all-important elements of a negotiation in general and at startups in particular: what you can get and what you should get during a negotiation.

What can you get? Negotiations have a lot to do with leverage, perception and both parties’ BATNA’s (Best Alternative To a Negotiated Agreement). Roughly, in a negotiation, your strength is equal to the quality of your BATNA minus the strength of the other person’s BATNA. There are other factors and I highly recommend reading Getting to Yes if you want to learn more.

However, once you get some experience with negotiation, you start to get a pretty good feel for it. You can tell when you have leverage or a strong position and you develop a sixth sense for when you can get a good deal. Sometimes you can buy something below its cost. Sometimes it is a revenue split that is unfairly skewed to you.

If you are not familiar with the signs that indicate you are about to get a good deal, I recommend you practice negotiating more. Use Craig’s List or play Monopoly.

The next question is, “What should you get?” What you can get is very different than what you should actually take. Just because you can hire a great employee cheaply, does not mean you should. If they feel they are underpaid, they will be less motivated and may eventually leave. Just because you can get an astronomical valuation because you had an auction, does not mean you should take a valuation you do not feel you can grow into. If you are negotiating with someone who you want to have an ongoing relationship with, then do not take excessive advantage of your negotiating position. Short-term wins can come at the expense of a long-term relationship. It is important to know how much you can get in a negotiation, but it is more important to know how much you should take.

To The Distributors Go The Spoils: Why Netflix & Amazon Will Win

The distributors that control consumers’ preferred distribution channels will win the entertainment media business. Nobody likes to buy bundled Cable content. However, the content we all like can only be gotten through bundled cable, for now.

I love digital distribution channels such as Netflix, Amazon Instant Video and Hulu for entertainment media because I can watch what I want when I want.

Cable seems antiquated. Certain shows are only available at specific times. I have to buy 100+ channels to only watch a few. Watching 30 minutes of TV means ~9 minutes of commercials that are minimally targeted to me, at best.

Clearly, the digital channels have many advantages except one: content. If I can only watch my favorite shows on TV at a designated time with commercials then I will, for now. Previously, cable controlled the means to produce quality content. They also had the cash and audience reach to lure big time stars. This has changed. Digital distributors now have large audiences and  stockpiles of cash.

Yesterday, I was watching Alpha House (great show) on my laptop through Amazon Prime. My roommates were watching House of Cards (another great show) on our Apple TV through Netflix in the other room. It has begun to happen. Our preferred distribution channels have equal or better content to Cable.

Now comes a race. Will today’s entertainment leaders adapt to our distribution preferences or will today’s digital distributors make premium content first? HBO GO is the best attempt I have seen at the former, but it is not digital-only distribution because you still need to pay a cable company for HBO.

Netflix & Amazon are starting to make content that is awesome. They are motivated to de-risk their business by reducing their need to pay for content licenses so they are willing to invest and take risks.

On the other side of the spectrum, Cable is motivated to maintain the status quo. They will be slow to cannibalize their current business model as it is profitable now and until recently, they had no reason to do so. They had the best content and that was unassailable. Now, distributors have the cash, data, confidence and motivation to break the content monopolies of cable companies.

The data piece is crucial. Amazon knows what shows and topics are popular with their audience. They know what cliffhangers compel someone to watch the next episode right away and which scenes in a show precipitate a viewer stopping and writing a bad review for the show. This is the definition of big data. Cable is very behind with Nielson.

Distributors will win in the long term because they have better data on their consumers, own the preferred distribution channels and they have the motivation to take risks while Cable is motivated to maintain the status quo for another quarter.

Digital distributors will win, but what is the best content release strategy? I talk about the different distribution release strategies here.

Binge Releasing TV Show Kills Buzz, but Makes Customers Happy: A Dilemma

Digital distributors of content are moving into original content creation and will likely beat cable companies. (I wrote about the topic here). However, there are two primary strategies for releasing their original content: binge and weekly. Which one is the best? This choice can affect the success and audience size of a show.

Before assessing which strategy is best, it makes sense to first examine the purpose of these original shows. Content creation takes away the dependence of distributors on content creators. A show can be judged on three metrics: (1) how happy are viewers with the show, (2) how fast is the viewer base growing and (3) does this show factor into the customer’s decision to subscribe to a specific service.

I think point 1 and point 2 are related to each other in two ways. The first way is, obviously, good content will make the audience happy and new customers will seek good content at a faster pace than bad content.  Duh.

The second relationship is not obvious. I think there is tradeoff between the happiness of current viewers and the growth of viewers. Current users choose to binge as Netflix alludes to on page 4 of their earnings release when they discuss the way they recognize expenses for original content.  Also, here is a study showing customer prefer to binge. Allowing users to binge makes them happy.

On the other hand, people are more likely to discuss a show that has a spread out release. They will discuss the show offline and through social media. In addition, traditional media and critics have a chance to call attention to the show. In short, there is more peer-to-peer marketing and more time for general buzz with a longer release schedule.

Unfortunately, I can’t know the exact tradeoff between these two forces. Amazon and Netflix will know soon, if they do not already. My guess is that there is an optimal point on the continuum of binge versus weekly releasing to optimize for audience happiness and buzz creation.

I think the optimum will be in releasing batches of shows. I think the viewer is still pretty happy getting 4 episodes at a time. At the same time, a bi-monthly (every other week) release schedule would probably allow the user base to grow faster through social marketing and traditional channels. I wish I had the data to actually conclude and not just hypothesize this.

As for point 3: “does this show factor into the customer’s decision to subscribe to us” – it is not important today, but it will be essential in the future. I know people who subscribe to HBO mainly for Game of Thrones. TV shows produce a few huge winners and if Hulu was the only distributor for Breaking Bad, I bet a lot of people would sign up for their premium service. That is an enormous strategic advantage.

Original shows that will drive subscriber decisions are not important today because these shows do not exist for digital distributors, yet. However, a runaway hit will one day produce new subscribers and lower customer churn in the future as competition becomes more fierce between distributors. One day, content will be optimized for hits that create competitive advantage between distributors. Whatever release schedule that is most likely to produce shows popular enough to affect subscription decisions will become the norm.

Today, distributors must balance viewers’ happiness with their ability to market their shows. I think the balancing point is in releasing batches of shows. Amazon, Netlfix, etc. have enough data that they should be able to figure out the right formula for batch size. However, the future will be choosing the release schedule most likely to produce shows that affect subscription decisions.