Posts from Facebook

The Gotham Gal and I went out with friends last night. As can happen, we got talking shop towards the end of the night. And specifically we were debating the significance of Snapchat. The debate was about whether the feature that makes Snapchat special (you know your photos won't end up on Facebook) is the basis for a standalone app and business. My view, having lived through this debate with Twitter and Foursquare, is that mobile apps are features in the mobile OS and that Snapchat can and likely will own this feature in the leading mobile operating systems even though institutionalized copycats (ie Facebook who copies everything) can and will copy it. The irony that Facebook has copied a feature that is specifically designed to avoid Facebook is precious in and of itself.

But I digress. The thing I want to talk about here is the emergence of privacy as the defining feature of the next breakaway app on the social internet. What does this mean for where we are and where we are going? Is open social out and closed private in?

At times like this, I like to talk to my kids and their friends. Here is a typical college aged woman I know. She uses Twitter, Instagram, Cinemagram, Foursquare, iMessage, and Snapchat. And Facebook too. She uses each of them for what they are good for. Each of them is on her home screen – one click and she is sharing something with someone. Each app offers a different graph – that she has curated specifically for that app – and each app offers a different type of engagement. If it is something silly that she wants to share with a friend but would be mortified if it ended up on Facebook, its Snapchat. If it is something she wants out there broadly, it is Twitter. If it is something she wants to share with a wide group of curated friends, it's Instagram. She has a private Instagram account so she controls who follows her there. She is a sophisticated user of social media. She was on Facebook in middle school and has grown up with this stuff. She knows how to use social media and she adopts whatever is useful to her. Snapchat is useful to her. Privacy is an important feature at times and she is happy to have an app with that as the central value proposition.

So that is my way of saying that I think privacy is an important feature and kudos to Snapchat for figuring that out. Further they invented a mode of engagement (the photos self destruct) that is new and novel. And the result is they are on the home screeen of millions of mobile phones and that number is growing by the day.

I expect we'll see a rash of copycats and other approaches to leveraging privacy as the central value proposition in the coming months. I am not sure that is the right thing to learn from this though. I think the right thing is to think about what other features are missing in the mobile OS and figure out the right mode of engagement to implement that feature. That is what Twitter did with status, Foursquare did with location, Instagram did with photo sharing, and Snapchat did with privacy. That got each of them on the home screen. Figure out what the next thing is.

I've thought this for a long time but never really articulated it publicly until the Q&A session at ad:tech last week.

There are a ton of social services that sit on top of the world of e-commerce and allow users to curate items they like and may want to buy in the future. These experiences can be highly social. And there are services that allow for transacting and payment inside of large social platforms like Facebook, Twitter, Instagram, and Pinterest. They bring commerce to social platforms. This entire category of services is called Social Commerce.

I have not been particularly bullish on these services because I think social commerce is most naturally e-commerce with a well implemented social layer built natively into the commerce platform. When you look at conversion rates in e-commerce broadly what you see again and again is that the more friction and overhead there is between discovery and transaction, the lower the conversion rate. Something as simple as logging into a commerce platform to complete a transaction can lower converstion rates by an order of magnitude.

Conversion rates are critical. They tell you what systems perform best for the end user. When a system converts north of 5% of users visits to a transaction, it is working extremely well for the end user. When a system converts 0.1% of user visits to a transaction, it doesn't work as well for the end user.

When a retailer or e-commerce service implements a highly social layer into their service with hooks into the major social platforms, the conversion rates can be significant. I have seen this first hand. This is an indication that users enjoy and benefit from social commerce when it is built into a native e-commerce service.

When users start in a social system that is divorced from the e-commerce platform, I believe the conversion rates are significantly lower, often by an order of magnitude or more. This, to me, suggests that the overhead of multiple systems reduces the effectiveness of the experience for users and is suboptimal.

So this is the thinking that led me to say what I said on stage at ad:tech this past week. I've felt this way for a long time and I am glad that I got the question and had an opportunity to address this issue publicly. I am eager to hear the discussion of this issue in the comments.

Google Analytics has a relatively new feature that allows you to look at your "social sources" of traffic. According to Google, about 27% of the vists to AVC in the past month came from social sources. For those who are curious about the rest of the traffic, 30% is direct, 15% is search (much of which is really direct traffic), and of the rest, about half is from social sources.

Here are the top social networks that drive traffic to AVC:

Twitter and Hacker News have been the mainstays of the social traffic to AVC for a long time. Last year, StumbleUpon was driving a ton of traffic to AVC, but that waned early this year and it is much less of a factor today.

Facebook, Techmeme, and Disqus are the other big social drivers of traffic. 

And the traffic that Disqus drives is markedly different than all of the other social sources. These folks hang around longer, read more pages, and engage more.

If you have a blog or some other form of online media and have a Google Analytics tag on your pages, I suggest you take a look at your social sources. I think you'll find it interesting.

Android is fragmented and geting more so. This is a challenge for those that develop on it for sure and has been often cited as a big negative for the Android ecosystem. But it also a big plus.

I have a Kindle Fire on my bedstand. I use it primarily to read on in bed having moved to a Nexus 7 as my primary tablet device. The Kindle Fire uses Android as its OS and then puts a Kindle shell on top which makes it look and feel like something other than an Android. But almost every app that I have on my Nexus 7 is also on my Kindle Fire. The reality is that if you build for Android, you are also building for Kindle Fire.

When Amazon launches a phone, it would be my expectation that the experience will be a lot like Kindle Fire. Meaning it will be running Android with a Amazon designed shell on top.

And then there is Facebook. I have to believe that Facebook will build a phone in the same way. Start with Android and then put its own wrapper and apps on top. If that happens, I would imagine I would be able to run all my favorite Android apps on the Facebook phone.

So imagine a world in which three of the top four consumer tech companies have phones running Android. Does that sound like a fragmented world for Android? Yes. Does that sound like a recipe for having a massive number of Android devices out there to build to? Yes.

In my view, we are in a two OS world for mobile and I think we are going to stay there. I think Apple will own the high end with the best and most integrated experience. And I think Android and its many variants will own the rest of the market. I think everyone else is playing for crumbs in terms of market share and would be better off joining the Android variant parade.

What does this mean for developers? It means build for iOS and Android and ignore everything else. And I think it increasingly means you have to be on both iOS and Android as soon as you can. I have advocated for building for Android first and iOS second. I think that strategy will start making more and more sense for apps that aren't looking to be paid.

Fragmentation cuts both ways. It's bad and it's good. Long term, I think it is a big plus for Android.

We have finally gotten videos up on the web from our Hacking Society event that happened in late April. A number of participants are going to blog this week about some of the specific topics and my contribution will be about our conversation about money and politics.

This 14min video clip starts with Larry Lessig outlining the issues around money and politics and then goes into some potential solutions. I advocated for a hack of the campaign finance system that starts with the internet industry contributing something like 5% of its combined ad inventory to a pool that is available to politicans who agree to certain conditions.

There was a debate about what those conditions should be. One group thought the candidates need to sign onto the Internet Freedom Pledge. Others thought that the candidates should renounce traditional campaign finance. Others thought candidates should do both.

Of course, this idea requires the Internet industry (Google, Facebook, Twitter, etc, etc, etc) to allocate 5% of its inventory to this kind of a pool. I have had a few conversations with the industry leaders about this idea but honestly it hasn't gone anywhere. It would be great if our industry could rally around this idea and make it happen. Hopefully this post and other actions we are taking this week might get this idea rolling.

Here's the video:

By the way, interested folks should check out the Hacking Society web page. In addition to videos, we now have a transcript of the event as well as audio recordings, photos, and lots of quotes. There's a lot there and its interesting and engaging stuff.

I have written so much about free here at AVC that it should be called AVFree. In fact, I've even written a post with this exact same title (almost exactly seven years ago, the summer we invested in twitter, zynga, and tumblr).

I haven't touched this subject much in the past few years but given that the topic has been in the air in the tech blogosphere, I thought I'd address it again.

First, let me say this is specifically not about Twitter. Folks should be encouraged to compete with Twitter. Competition is good for everyone as we talked about here a week or so ago.

This post is in reaction to the idea that services should be paid to ensure that they are appropriately focused on the consumer/user as opposed to the marketer/advertiser/sponsor.

Let's start with advertising. I do not believe it is evil. In fact, I believe it is a fantastic way to support services that want the broadest adoption and want to be free. Think about the Super Bowl, the World Cup, the Olympics, the Oscars, the Presidential Debates, the news coverage of important events. These things are ad supported and free for anyone to watch who has a TV and an antenna. It is good for society for these things to be available to the broadest audience.

There is a paid TV business and it is flourishing. The fact that at its base level TV is free does not mean that all TV has to be free. Free TV does not commoditize paid TV. They co-exist nicely. But we had free TV well before we had paid TV. Free is the foundation that creates a paid tier. That's what Freemium (a term that was coined here at AVC) is all about.

Now let's examine services that have a free and paid tier. Spotify is a good one to look at. Spotify's core business model is a subscription music service. All you can eat music for a fixed price every month. And yet this is what they show you when you land at Spotify.com for the first time.

Why do they show you that? Because most people prefer free to paid. I don't have the exact numbers but I would suspect less than 10% of Spotify's users pay for music. Why is Pandora so successful? Because it is free. Why is over the air radio still the most popular way to listen to music? Because it is free.

Now let's look at servces where the users provide all the value. Wikipedia, Craigslist, YouTube, Flickr, Instagram, Facebook, Twitter, Tumblr, WordPress, etc, etc. There is no value to any of these platforms if the users don't create the content. The users create the service, curate it, and make it what it is. I do not believe it makes sense to charge users to create the value. We've seen folks try this model. Typepad (where this blog is hosted) charges to host a blog. How well did they do? Phanfare charged to host photos. How well did they do? The list could go on and on but I don't want to focus on failed services.

When scale matters, when network effects matter, when your users are creating the content and the value, free is the business model of choice. And I don't think anything has changed to make that less true today. If anything, it is more true.

I understand the frustration of certain folks about the commercialization of Twitter, Facebook, Tumblr, and a host of other services. I understand the frustration over the increasing lock down of the APIs and the control these platforms are exercising on their ecosystems. I would encourage folks to compete with them to keep the web, the mobile web, and the Internet free and open. But I would not encourage those same folks to build paid services. I think their goals will be undermined by that choice.

Twenty years into the personal technology revolution and we are still using address books that work pretty much like physical address books. It makes no sense. The mobile address book should be hyperconnected to our digital life, informed by it, and responsive to it.

I remember back in early 2011, Steve Greenwood walked into our old offices on the 14th floor and told us that he intended to build that hyperconnected mobile address book. He showed us a spreadsheet he had been personally using for the past five or so years to manage all of his relationships. I had never seen anything like it. He was obsessed with relationships and managing them. That's what we are always looking for, someone who just can't sleep because they want to fix something that isn't working in their world and have been trying for a long time to fix it.

Steve had built a prototype that ran on the web and the vision was all there. An address book that knows what you are doing, where you are, who you are most engaged with at any time, and is search and context driven as opposed to a directory of names and addresses.

But Steve knew that he had to launch this as a mobile app that will eventually replace your current address book and he knew he had a lot of hard technical problems to solve in order to do it right and do it at scale. So he asked us for the seed capital to build a team and build that product. We said yes.

Today, Steve and the amazing team he put together are launching Brewster, initially as an iOS app. If you have an iPhone, you can download it from the iOS app store and check it out. The NY Times has a good post on Brewster today.

Here's how Brewster works. You download the app. You connect it to google apps, facebook, twitter, foursquare, linkedin, and soon a bunch more services. Brewster builds you a new address book that runs on your phone and also in the cloud.

This new address book is smart because it knows a lot more things about your relationships than you have ever entered into your address book and it is adapting in real-time to all of this data. It knows who you probably want to talk to right now and it also knows who you are losing touch with and displays all of this data in a feed. Your Brewster address book is also de-duped and hot linked to all the social activities you want to do from calling, texting, facebooking, or whatever.

This is an address book that can handle a search query like "knicks game" or "sushi tonight" or "band of horses concert". We are always querying our brain with questions like that. Now we can ask our address books those kinds of questions.

I have really enjoyed being involved with this project over the past year. It fits neatly into many themes I have been writing about and thinking about for years. But mostly I am excited to finally see this product out in the market where folks can use it and experience Steve's vision of how an address book should work in the mobile social world we live in.

If you look at any of the top web properties on comScore, Quantcast, Alexa or any other third party reporting service you will see that they all have been fairly flat over the first half of the year. You might think that all these big web services are flatlining.

We have seen this in our portfolio too. From board meeting to board meeting, we are seeing a similar pattern. Web is flattish. But mobile is growing like a weed.

I alluded to this in a post last week where I wished for an aggregated audience measurement service across mobile and web.

There is a significant shift going on this year, much more significant than we saw last year, from web to mobile. It is most noticeable in games, social networking, music, and news, but it is happening across the board and it presents both great opportunity and great challenges.

Mobile native services like Foursquare & Instagram have the most to gain from this transition. Big feature rich web apps like Facebook and Google have the most to lose from this transition.

Mobile does not reward feature richness. It rewards small, application specific, feature light services. I have said this before but I will say it again. The phone is the equivalent of the web application and the mobile apps you have on your home screen(s) are the features.

That is why Facebook should (and it looks like will) break its big monolithic web app into a bunch of small mobile apps. Messenger, Instagram (not yet owned by Facebook), and Camera are the model for Facebook on mobile.

User experience is not the only big change/challenge for companies trying to navigate this transition. Monetization is different too.

Approaches like display advertising don't work as well on mobile as they do on the web. And they don't work that well on the web. ARPUs (avg revenue per user) on mobile are lower for display based revenue models on mobile across the board.

On the other hand, commerce works great on mobile if you have a well integrated (one click) purchase experience. The freemium model (whether it is virtual goods in games, in app upgrades, or something else) works very well on mobile.

If you are going to operate a media model on mobile, look at Twitter's model for inspiration. The ads are the default content object (the tweet) and are delivered right in the primary user experience (the feed/timeline). It's not surprising that more than half of Twitter's ad revenue is coming from mobile.

All of this is good news for entrepreneurs since they are in the best position to take advantage of all of these changing dynamics. It is not as good news for those who find themselves operating a big Internet business started more than five years ago. You are going to need to make a hard right turn super fast without flipping over the car.

In the past fifteen years, we have seen Microsoft go from being an unstoppable force to being a non-factor in many important new markets, we have seen Google go from being an unstoppable force to being a non-factor in many important new markets, and I suspect we are going to see Facebook struggle with the same thing. RIM is dying quickly now. Yahoo! is a question mark.

In technology the more things change, the more the stay the same. You cannot ever rest. Because the big change that is going to upset your nice apple cart is right around the corner. Today that is mobile. Tomorrow, who knows? I am trying like hell to figure out what that will be and jump on it. Because that's how you play this game.

I don't disagree with PG when he says that Facebook's IPO performance (or lack thereof) has the potential to impact valuations in startup land. I think it will be particularly impactful on the late stage and secondary markets where most of the IPO valuation speculation is happening.

But let's put Facebook's current valuation in perspective. At the closing price of $26.90, Facebook commands a valuation of $57.5bn (according to Google Finance). Facebook had around $4bn of cash prior to going public and raised about $10bn so let's assume they have $14bn in cash on the books. That means Facebook has an enterprise value of roughly $43bn.

In its last quarter Facebook had $1bn of revenue and 40% pre-tax operating margins. If we annualize those numbers, that would be $4bn of annual revenue and $1.6bn of annual pre-tax operating margins. Let's use pre-tax operating margins as a proxy for EBITDA, because this whole post is back of envelope quality analysis and please take it for what it is.

That means that Facebook's enterprise value is greater than 10x current revenue run rate and greater than 25x EBITDA. These are big multiples folks. I am happy to take those numbers for any company out there.

Clearly Facebook is a premium company and commands a premium valuation and entrepreneurs should not expect to get 10x revenues and 25x EBITDA for their companies in a sale or an IPO. But even at half those numbers there are fantastic returns for investors and entreprenuers to be had.

If speculators are disappointed with the performance of the Facebook IPO it is because they had ridiculous expectations of what rational investors would pay. The market has put a premium valuation on a great company and we should be happy about all of that. I certainly am.