Posts from Online Communities

I guess this should be a Feature Friday post but I write my posts based on what I am thinking about first thing in the morning and this is what I am thinking about right now.

I just went to Twitter like I always do first thing in the morning and this is what I saw in my timeline:

Right at the top of my timeline, Twitter is telling me that my friend Mark is active again on his Mediaeater handle. And they follow that with the suggestion that I tweet at him. This is the first time I've seen this kind of messaging from Twitter. And I like it.

If you go back to my 30/10/10 post, you will see that most web services and mobile apps only get 30% of their registered users to use the app at least once a month. The other 70% have largely gone away. But it doesn't mean they are gone for good. Getting them back should be a primary goal of any consumer web or mobile company.

Twitter has been working on this problem for years. I recall getting a presentation from their growth team at least three years ago that detailed how they were working on this problem. And they are still working on this problem. Most likely they will always be working on this problem.

You can email or spam in some other way your inactive users and that might work. But what you do once they come back is way more important. You have to figure out how to make the experience better than it was when they used it previously. Some of that will likely be that the product is much better because your and your team have improved it a lot. But some of that should be an engaging experience that somehow they did not get before.

When we get our various portfolio companies together, I like to ask them if they can identify one or two metrics that separates their successful and engaged users from their unsuccessful and inactive users. Most of the time the metric has something to do with engagement (they left a comment, they got a reply to a comment, they got a like on their photo, they had people follow them, etc, etc). Engaging with real humans, not just the machine, is the key in social systems. And most systems are social in some way.

Your inactive users are important cohort to focus on. There must have been something that got them to sign up for your service in the first place. So focus on getting them back, retaining them, and most importantly, engaging them.

Related articles

Meet Retention Score, the new metric that predicts your customer churn

Think Tracking DAUs Is Enough? Then You're Not Getting Your Whole Mobile Story

What Twitter can teach you about growth hacking

We had a good chat hanging around Zander's desk yesterday about this line from Matthew Ingram's post on his love/hate relationship with Twitter:

Lastly, I hate that Twitter’s metamorphosis seems to reinforce the idea that being an open network — one that allows the easy distribution of content across different platforms, the way that blogging and email networks do –isn’t possible, or at least can’t become a worthwhile business.

I asked Zander what he thought about that line and he told me he hadn't thought long and hard enough about it to have a fully formed opinion but it was certainly important to our investment thesis and we ought to have an opinion on it.

I do have an opinion on it.

I do not think open conflicts with making money and further I think there are ways to make more money by being open rather than closed, but it takes imagination and a well designed relationship between your product/service and the rest of the Internet.

I also think it is better to open up slowly, cautiously, and carefully rather than start out wide open and then close up every time an existential threat appears on the horizon.

I recall when Etsy first put out an API. It was a read only API. Then they made it read/write. Over time they have added a lot of features that have made it possible for third parties to add value to Etsy and Etsy's sellers and buyers. But they have always protected the essential things that make Etsy's business and marketplace work and hang together as a sustainable entity.

Contrast that with Twitter which started out completely open which allowed anyone to build a third party client, grab a huge percentage of Twitter users, and then threaten to take them away from Twitter. That's not a sustainable relationship between your product/service and the rest of the Internet.

So I do believe that there are many ways to be open, to become more open, and to do so in ways that enrich the overall Internet and your company too. I am quite fond of O'Reilly Doctrine:

Create more value than you capture

And I think doing so means being open, becoming more open over time, but always in ways that allow you and your company to remain a sustainable business that can cover its costs and then some and remain viable and value enhancing for the long haul.

I love quotes. Back in the early days of Twitter, we used to tweet a quote of the day. And I tumbl a lot of quotes I find on the web. So when Max Yoder suggested to me a week or so ago that we do a fun friday on quotes, I was sold before even fininshing his short email.

The idea is you comment on this fun friday post with a favorite quote of yours.

I will start with this gem from my all time favorite person to quote, Winston Churchill:

It is a good thing for an educated man to read books of quotations.

Now it's your turn.

This isn't exactly about a feature. Features are the verbs of a web/mobile product. Objects are the nouns. And one thing I always like to think about is what is the most fundamental object of all in your service. I like to call this the "atomic unit."

Here are some examples:

In Twitter, the atomic unit is the tweet

In SoundCloud, the atomic unit is the sound

In Turntable, the atomic unit is the room

In Tumblr, the atomic unit is the post

In Codecademy, the atomic unit is the lesson

In Wattpad, the atomic unit is the story

In Etsy, the atomic unit is the item (although one could argue it is the seller)

In Kickstarter, the atomic unit is the project

In Disqus, the atomic unit is the comment (although one could argue that it is the thread)

In Instagram, the atomic unit is the photo

In Kik, the atomic unit is the private message

In LinkedIn, the atomic unit is the resume

In Foursquare, the atomic unit is the checkin (although they may have just moved it to the venue)

In Gmail the atomic unit is the email, in Gcal the atomic unit is the appointment

In Brewster, the atomic unit is the relationship

In Dropbox, the atomic unit is the file

In Google Docs, the atomic unit was the document, in Google Drive, they would like to it to be both documents and files. And that creates some cognitive dissonance.

I could go on and on, and I apologize to all the USV companies I left off this list. I am not picking favorites. I am just doing stream of consciousness examples.

When you think about an MVP, it's really important to identify the atomic unit and make sure you focus the product crisply and cleanly on that object. If you think you have three or four atomic units, you are going to end up with a clunky and bloated experience and that is what you want to avoid at all costs with your MVP (particularly if you are mobile first).

Can you identify the atomic unit of your product or service? If you can't, then you might want to sit down and think about why you can't and what you might be able to do to address that.

I remember when I first met Joshua Schachter. It was in our first office back in early 2005 and he was still working at Morgan Stanley and running Delicious out of a server in his closet. I was immediately struck by the way Joshua looked for lightweight simple ways to do things. He told me that reduction was the key to getting something right.

Delicious was so simple to use. It didn't do much. But what it did do, it did well. I once asked Joshua why he didn't let users rate links. He said "we do. if they post it to Delicious they are ranking it as interesting. if they don't post it to Delicious, they are ranking it as not interesting". That's classic Joshua.

Skills.to is also classic Joshua. For those who haven't seen it, skills.to is a super simple lightweight way to tag people. It currently leverages the Twitter name space to find people, tag them, and tweet out the tags. You can also tweet out to your followers that you'd like them to come tag you. That's about it. Delicious for people.

There's a chrome extension that allows you to see people's tags in line in Twitter. I just installed that so I haven't had much experience with it yet. But I like the idea of a super lightweight people tagging system that is transportable across the web. I hope this takes off. Joshua is the right person to build it.

Check it out and let me know what you think.

Full disclosure. USV is an investor in Tasty Labs which has built Jig and Skills.to.

Related articles

• Joshua Schachter Launches Newest Tasty Labs Project, Skills.to (techcrunch.com)
• Delicious Founder Creates New People Search Engine, Skills.to (readwriteweb.com)

By now, regular readers of this blog should be pretty familiar with USV’s investment thesis centered around investing in large networks of engaged users that have the potential to disrupt large markets. I’ve mentioned it frequently here and my colleagues have also blogged regularly about it.

Today I got a tweet from Jon de la Mothe about the growth of professional networks on Facebook:

Professional Networks Built On Facebook Are Growing Like Crazy zite.to/H9KjuP via @fredwilson

— Jon de la Mothe (@delamothe) March 27, 2012

And I replied with this tweet:

@delamothe are they networks if they are built on top of another network?

— Fred Wilson (@fredwilson) March 27, 2012

I think it is possible to bootstrap a network on top of another network. Stocktwits did this on Twitter and Zynga did this on Facebook. But both of them eventually built their own networks directly on the web and mobile. Zynga still gets a ton of game play on Facebook and Stocktwits continues to benefit from tweet distribution on Twitter, but they have made the necessary investments to operate their businesses in such a way that they are not entirely dependent on other networks. In the case of Stocktwits, they did this early on. In the case of Zynga, they waited for a while to do this.

But my question to Jon is a bit different. I didn’t ask if you can build a network on top of another network. I asked if it is a network if it is built on top of another network. I think in that case, the answer is no. 

There is a third way, which are networks of networks. My partner Albert has blogged about this and I am 100% in agreement with him that this is the way the market should evolve. I believe that the Internet is an operating system and the networks that operate on the web and mobile via the Internet should interoperate with each other, share traffic and distribution with each other, and act as peers with each other. This is in keeping with the architecture of the Internet and is the most sustainable model and the one I am betting on in the long run.

Code Year, which I blogged about a couple days ago, has now signed up over 100,000 in two days. That's a lot of signups for a brand new service in just two days. How did they do it? Here's some suggestions on the key drivers:

1) an awesome idea. "give me your email address, we'll send you interactive coding lessons weekly" is a damn good idea. tim o'reilly told the codecademy guys "i wish i'd thought of this". that's the definition of a good idea.

2) well timed – launching as a "new year's resolution" is genius. but also launching in a "dead news period" was equally genius. jan 1st and jan 2nd of this year were slow news days. so Code Year got plenty of airtime in the tech blogs and news aggregators over a sustained two day period.

3) the landing page is clean, simple, and well designed. the call to action is simple. here's a blog post from the designer explaining how that page was designed.

4) the use of twitter and facebook to spread the word is simple and powerful. after you give your email address, you are given the option of tweeting out or posting to your wall. TechCrunch says 50% of the site traffic comes from Twitter and Facebook (with Twitter coming in at >33%).

5) a small ask. they didn't ask for money, the service is free. they simply asked for an email address, something everyone has and most are willing to share in return for real value.

So kudos to the Codecademy team and everyone else who was involved for great execution of a service launch. I am looking forward to getting my first coding lesson and getting started.

Last month, I participated in a PII event and sat on a panel (a format I dislike and try to avoid) moderated by Kara Swisher. Kara posted a video of the panel on ATD a few days later.

Sometimes when you are in a public setting you say things that have been coming together in your mind but you have never articulated before. That happened that day. I said the following about 40 minutes into the panel:

Twitter is default public and everyone knows that's what it is. Your Twitter identity is the lightest weight, most public, and therefore the best identity on the web.

Many other online identites we are all developing are heavier weight. They have more private information about us in them. When the companies that operate those identity services share our information with others we get nervous, upset, and anxious.

Twitter and other default public identities (like my daughter Emily's "Things I Like" tumblog) contain only the information we are willing to have the whole world know about us. And therefore they are better identities. They can be portable without our permission. They are crafted by us with the full knowledge that they will be seen by others, potentially many others.

We just completed a long and taxing hiring process. We had over 250 applicants to our analyst position. We asked each and every applicant to share their online profiles with us. Most shared default public profiles like blogs, twitters, tumblrs, etc. We learned so much about each of the applicants through reviewing those public identities. Default public online identities are very powerful and very revealing about people, maybe more so than default private identities. They can be used for almost anything that default private identities can be used for. But they can be used without asking for permission from the owner of the profile because the information is public already.

This is counter intuitive for many. But I'm pretty sure that default public identities are the future for most things (maybe not healthcare and personal finance and the like) and that they are the best form of online identity.

About a month ago, William Mougayar, an AVC regular if there was ever one, emailed me about an idea he had to create a web service to make social conversations easier. We went back and forth on the idea. I pushed him to come up with a simple name and a simple UI. I told him "make it like gmail for social conversations."

Over the past month, he and his crack team of something like two or three developers did just that. And today, I'm pleased to tell all of you that William and his team have built something pretty special.

He calls it engag.io. It's a good name. An io channel for social engagement. Here's what my engagio inbox looks like this morning:

My inbox is dominated by AVC discussions. For most people, their inbox will have a multitude of conversations from many services; Twitter, Foursquare,  Hacker News, Facebook, and hopefully AVC too.

There are a few other nifty features shown on the left sidebar. Engagement shows who you are engaging with. Shared links stores all the links being passed back and forth in your conversations. And Sites shows what sites your friends are having discussions on. So engagio is not just your inbox for social conversations. It is your dashboard for social conversations.

William's got a blog post up on the engagio blog explaining his vision for the product. It's a good read.

Engagio is offering 100 invites this morning to AVC readers. The first 100 users will get in. The invite code is avcengage (lower caps). Give it a try and let us know what you think.

Just In Case You Were Wondering: Neither me nor USV has any financial interest in engag.io or eqentia, it's sister service.