Posts from 2014

Sidechains

Earlier this year some entrepreneurs walked into our office and explained sidechains to us. I was pretty excited about the concept then and I continue to be excited about it. This past week some of the people who explained them to us and some other people I don’t know published a paper about sidechains called Enabling Blockchain Innovations with Pegged Sidechains. I think this is an important paper and everyone involved in bitcoin, blockchains, and cryptocurrencies should give it a read.

Here’s the basic idea in layman’s terms. I am purposely trying to dumb down and simplify the idea here.

1) The Bitcoin Blockchain is the most liquid blockchain, has the most mining on it, and is likely to remain that way given the network effects it has built up over the past five years.

2) The core Bitcoin system and software is not likely to change very much because making changes to it is risky and there is a lot of capital at stake on the Bitcoin blockchain now.

3) There are things you might want to do that are not well supported or not supported at all on the Bitcoin blockchain.

4) This desire for “other things you might want to do” has given rise to a ton of alternate blockchains, none of which have developed a lot of liquidity and mining on them.

5) So what if you created “sidechains” that are “pegged” to the Bitcoin blockchain that allow “other things you might want to do” while still leveraging the liquidity and mining of the Bitcoin blockchain?

That’s the basic idea and the paper I linked to explains how to do the “pegging” part which is critical to this whole idea working.

I believe that the core Bitcoin system and software will have to be modified to allow these “pegs” and I’m pretty sure (but not positive) that these changes have not yet been made.

It will be interesting to watch how all of this develops over the next year or two. If “pegs” are added to the core Bitcoin system and software, and if sidechains become popular and viable, then Bitcoin would essentially become the reserve currency of the entire cryptocurrency sector and there would be a host of sidechains and currencies that are pegged to it. This would allow a ton of innovation to happen around Bitcoin without requiring a lot of change to the underlying Bitcoin system. Which seems sort of ideal given how much money ($5bn at current prices) is at stake now.

The Cost Of Loyalty

In the local transportation market, we now have lots of options in addition to mass transit. Here in NYC, we have taxis, Lyft, and Uber. In SF and LA, we have taxis, Sidecar (our portfolio company), Lyft, and Uber. Around the country and world, there are various options including our portfolio company Hailo.

I’ve always wished there was an aggregation app that pulled all the prices and availability in real-time across all the available services and got you the best fare at the time. Or allowed you to make the choice between price and ETA (the way sidecar’s app does). It turns out there is a lot of price variability in the market and there is not one choice you can make all the time that will work out well for you. Being loyal to one app costs you.

Then this morning, a blog post popped up in my inbox courtesy of my friend Boris. In this post, they calculated the “cost of loyalty” to one just one app.

cost of loyalty

I mostly use taxis in manhattan when Citibike and subway won’t do and that’s because they are the cheapest and most available option. Uber and Lyft are for times you can’t get a cab and you’ll pay through the nose when you take that option as they are almost always surging at those times.

Another interesting thing about these charts is how taxis are the most expensive service to be loyal to in SF and LA. That is crazy. They are going to go out of business in those markets with that pricing.

But mostly I am proud that our portfolio company Sidecar is the least costly service to be loyal to. That is because they don’t use surge pricing and instead allow drivers adjust pricing in their marketplace model as they desire. Sidecar is committed to using a true marketplace and things like shared rides to deliver the lowest cost rides in the market. It is also true that Sidecar ETAs are a bit longer as this chart of SF shows:

ETAs

Going back to the opening thought, which is that someone should build an aggregation app on top of all of these services so we can replace the app on our home screen that we are most loyal to with an app that works across all services. The authors of this blog post did just that and you can use What’s The Fare to tell you who has the best price in the market. It looks like right now its just a web/mobile web app and all it does it give you the fares. If they or someone else went further, made it into a mobile app, and used the services APIs to actually book rides (if the APIs were available to do that), then we’d really have something.

That’s the way this market should work long term. I hope we can get there soon. Google Maps and Apple Maps are the ideal interfaces to make it happen. Let’s go!

Feature Friday: Etsy In Real Life

This week our portfolio company Etsy introduced Etsy Reader, a dongle for your phone or tablet that allows Etsy sellers to sell on Etsy in real life.

etsy_in-person_payments

The natural reaction to this would be “Etsy knocked off Square” and to some degree that would be correct. But Etsy Reader is not just a card reader. There is quite a bit of software behind the scenes that connects the checkout experience to the seller’s shop on Etsy and all of the seller tools that Etsy provides. The better way to think about this is that Etsy Reader extends a seller’s Etsy Store to the real world of craft fairs, flea markets, and other in person experiences.

Etsy Reader is about coming full circle at Etsy. In the early days, back in 2005 and 2006 when we first invested, Etsy was built seller by seller, at craft fairs, with street teams manning Etsy booths and evangelizing a new way to find customers and meet other like minded people. Etsy sellers still sell a lot at craft fairs and other face to face environments. Now with Etsy Reader, the shop can be online or offline with everything tied together.

I am excited to see Etsy Reader come to market. It’s been a dream for a while now and props to Camilla and her team for getting it out the door. Well done.

The Second Coming Of Sign In With Twitter

First a disclosure. The Gotham Gal and I personally own a lot of Twitter stock. I am not objective. With that behind me, I am going to talk a bit about Fabric today.

One of my great disappointments during the years I was closely involved at Twitter was the failure to make Sign In With Twitter a competitive offering with Facebook and then Google. In the early days of OAuth, Twitter was an innovator and leader in this area. But we did not invest enough in the technology and partnership development, we did not supply email addresses via the service, and we did not have as many users. Over time signing into an app with Twitter has declined as an option with developers in favor of Facebook and, increasingly, Google. And it bothered me a lot. Still does.

Yesterday Twitter announced Fabric, which is a set of tools for mobile developers that is an attempt to change that. The timing could not be better. We talk to a lot of mobile developers who come into our office seeking capital and mostly getting advice. I am alway interested in where they get their logins from. Most offer sign in with Facebook but many users are choosing not to use that these days. Many offer sign in with Google and that is growing in popularity but signing in with Google works way better on Android than iOS. Signing in with email remains a popular option, way more popular than you might imagine.

That tells me that there is an opening for Twitter to get back into this game in a big way and Fabric is their attempt to do just that. The enticements to use Fabric for developers are Crashlytics, a very popular crash reporting tool that Twitter purchased, MoPub, a mobile ad exchange that Twitter purchased, and Digits, the new mobile sign on offering. Crashlytics and MoPub are both really good services for analytics and monetization, but it is Digits that got my attention yesterday.

Digits lets you sign with your phone number without providing an email or a password. According to Twitter:

So we’re excited to announce Digits – part of the Twitter Kit – which is sign-in with phone number done right. It’s built on Twitter infrastructure so you don’t have to worry about managing multiple relationships with carriers and SMS interchanges. Digits is fully themeable so that it fits the user experience you’ve designed for your app. Digits won’t post anything on your user’s behalf since it isn’t tied to their social network accounts, including Twitter. And with Digits, your apps are ready for global adoption: it’s available immediately in 216 countries and in 28 languages, on iOS, Android and the web.

Here are the big things for me in this new service:

– Won’t post anything on your user’s behalf because it isn’t tied to their social network accounts, including Twitter. This is what got Facebook’s sign in product in trouble with users. This is a big deal.

– Available in 216 countries and 28 languages, on iOS, Android, and the web. Getting anything that involves the phone system (phone numbers and SMS) working all around the world is not trivial.

– No password required. It’s a pain to remember user ids, email addresses, and passwords. That’s why signing in with Facebook and Google is such an enticing thing to a user. Signing in with your phone number is an even better option in my mind.

So I’m excited to see Twitter take another swing at providing sign on tools and identity services to developers, particularly mobile developers. I’m hoping it’s a home run for them.

Firebase

Yesterday our portfolio company Firebase was acquired by Google. We invested in Firebase in the summer of last year and Albert wrote this post talking about the investment.

Firebase makes a backend that allows developers to store and sync their data in real-time.  In a relatively short time, Firebase has been adopted by over 100,000 developers. It solves some complicated problems simply and elegantly.

We believe Google will be a good home for Firebase. They have the resources and desire to continue to build out Firebase and scale it. I saw the two founders last night at our dinner in SF and they told me that Google is giving them budget to go out and scale up the engineering team. They had big smiles on their faces when they told me that.

Though our time as investors and partners in Firebase was short, we enjoyed working with them very much and are happy that they have found a good home for their company and their technology. It sure looks like the perfect match.

Getting Feedback and Listening To It

When you are VC, you live in this protected environment. You sit in your office in a glass conference room with lovely views and entrepreneurs walk in and pitch you and you get to decide who you are going to back and who you are not. People tell you what they think you want to hear. That you are so smart. That you are so successful. They suck up to you. And it goes to your head. You believe it. I am so smart. I am so successful.

You have to get out of that mindset because it is toxic. My number one secret is the Gotham Gal who brings me down to earth every night, makes me do the dishes, walk the dog, and lose to her in backgammon. Actually I have not lost to her in backgammon in over twenty years because she used to beat me so badly that I couldn’t take it anymore.

But blogging is another helpful tool in reminding yourself that you are not all that. Marc Andreessen said as much in his excellent NY Magazine interview which was published yesterday. I loved the whole interview but I particularly loved this bit:

So how do you, Marc Andreessen, make sure that you are hearing honest feedback?

Every morning, I wake up and several dozen people have explained to me in detail how I’m an idiot on Twitter, which is actually fairly helpful.

Do they ever convince you?

They definitely keep me on my toes, and we’ll see if they’re able to convince me. I mean, part of it is, I love arguing.

No, really?

The big thing about Twitter for me is it’s just more people to argue with.

Keeping someone on his or her toes, making them rethink their beliefs, making them argue them, is as Marc says “fairly helpful.” That’s an understatement. It is very very helpful.

That’s the thing I love about the comments here at AVC. I appreciate the folks who call bullshit on me. There are many but Brandon, Andy, and Larry are common naysayers. They may come across as argumentative, but arguing is, as Marc points out, useful.

The comments are also a place where people play the suck up game. It isn’t necessary to do that and I don’t appreciate it. It makes me uneasy.

So I would like to thank the entire AVC community for being a sounding board for my ideas, for pushing back when I am off base, and for resisting the suck up whenever the urge presents itself. I appreciate it very much.

The Personal Cloud

Benedict Evans coined the term “personal cloud” in his writeup of WWDC in June. He said:

what you might call the personal cloud – the Bluetooth LE/Wifi mesh around you (such as HealthKit or HomeKit)

I like to think about what’s next.

Paul Graham said, “If you think of technology as something that’s spreading like a sort of fractal stain, almost every point on the edge represents an interesting problem.”

And in that context, the personal cloud is a particularly interesting “point on the edge” to me. It includes the following things:

1) NFC and other technologies that will turn the mobile phone into your next credit card

2) Phone to phone mesh networking like we saw with Fire Chat in Hong Kong a few weeks ago

3) Wearables like the watch, necklace, and earbud

4) Personal health data recording (HealthKit) in which your phone has a real time and historical chart of your heartbeat, blood chemistry, blood pressure, pulse, temperature, and much more.

5) Airplay and Chromecast and other technologies that will turn the mobile phone into both the next settop box and remote

I could probably go on and list another five things that fit into the personal cloud, but I will stop there.

If the first wave of the mobile phone’s impact on the tech sector was driven by applications running on the phone, the second wave will be driven by the phone connecting to other devices, including other phones.

I am particularly fascinated about what happens when our phones connect to other phones in dense environments and form meshes that don’t need the traditional Internet connectivity to power them. Mesh networks don’t just solve the problem of lack of traditional connectivity (Hong Kong), they also produce a solution to the last mile connectivity duopoly in wireline and oligopoly in wireless. In the future we may just opt out of those non-competitive markets and opt into a local mesh to get us to the Internet backbone, both in our homes and when we are out and about.

And phone to phone meshes form local “geofenced” networks that are interesting in their own right. A nice example of this is the peek feature in Yik Yak where you can see the timeline at various universities around the US. These Yik Yak peeks are not powered by mesh networking, they are just using the geolocation feature on the phone. But they could be a collection of mesh networks operating in various universities around the country. And so that example is enlightening to me.

I wanted to end this post with an image of a person walking down the street surrounded by their personal cloud and all the devices that are connected to it. But a quick image search did not produce it for me. That in and of itself is telling. That’s our future. But right now we are still in the imagining phase of it.

The Robotic Taxi Driver

Yesterday morning I made the mistake of leaving my apartment without my Citibike key. When I got to the Citibike station, I realized it and hailed a taxi instead. I got in the taxi and told the driver where I was going which was 6th Avenue and 13th Street. He started to enter the destination into the GPS on his phone which was mounted above the dash to the left of the steering wheel. I told him that wasn’t necessary as all he had to do was go a few blocks down Washington to 10th, make a left on 10th, then across 10th to 6th, then a left on 6th. So he took off down Washington and the preceded to blow right past 10th. At which point, I told him that he had missed 10th and he should make the next left onto Christopher, which he then drove right past. After a couple more missed turns, I told him to stop and got out of the taxi and told him that he should learn a bit about getting around the city before getting behind the wheel of a taxi cab. Then I tweeted this out.

If you click on that tweet and look at all of the replies, you will find an interesting discussion of the current state of the taxis and ride sharing services in NYC, Chicago, London, and a bunch more cities. It seems that my experience of getting into a car and the driver having no idea where they are and where they are going is not unique. It’s happening to lots of people in lots of places.

Now you might say, “well you should have let the driver use the GPS” and you would be right about that. But in that tweet reply stream there are plenty of stories about drivers using GPS and still getting terribly lost. When you have no idea where you are and no idea where you are going, the GPS isn’t as useful as it would seem. And then there are the issues of road work, closed streets, traffic, and other sorts of things that requires experience and local knowledge to navigate. There is a huge difference between an experienced driver who knows their way around a city and a driver just off the plane from somewhere else driving around NYC using a GPS in lieu of that local knowledge.

What has happened in NYC and apparently in many other places is the arrival of ride sharing services has increased the demand for drivers and the best drivers are moving from taxis to the higher end services and new drivers are being recruited to drive the cabs and the lower end ride sharing services. These new drivers have no training and have no idea where they are going without the GPS. And they are totally and completely reliant on the GPS. It makes me feel like the autonomous car has arrived in the form of the robotic taxi driver.

I told this story to my friend Jeremy last night and he observed that the right answer is to use the higher end ride sharing services where all the experienced drivers are now working. He said “price and quality are lining up as you would expect in a market economy.” Of course the other option is to not forget my citibike key or walk or take the subway. Which is looking like a better option more and more these days.