Category Archives: Uncategorized

Thoughts On TechCrunch Disrupt

I wrote some tweets today at TechCrunch Disrupt Berlin.

At a very high level, here are the problems I see with where Disrupt is going:

1. I get that this is turning into a very big business for Aol. But TechCrunch Disrupt should be a conference with a focus on editorial, not sales.

2. Editorial at TechCrunch seems to be completely submissive to sales. If sales wants something, it happens. Way too many sponsors on stage, for example. And the speakers are nearly constantly herded to “VIP events” that are really just sponsor events where sponsors pay to have a captive audience of well known entrepreneurs and venture capitalists. Speakers are being packaged and sold to sponsors, and they are complaining about it.

3. Picking the winner of the battlefield startup competition should be free of any influence of non-editorial people and judges. The only people who should be in the room to decide who wins the battlefield should be the judges and the editors, and the editors shouldn’t be voting or trying to influence decisions.

4. It should be completely reasonable for judges to discuss the option of not having anyone win the battlefield at all if the quality level isn’t there. TechCrunch management today didn’t allow that conversation to happen, saying it would send the wrong message. Those people shouldn’t even be in the room, let alone directing the conversation.

TechCrunch should be strong enough to withstand criticism, and I hope they understand that I give this advice out of love, not malice. In any event, I’m not going to participate in the battlefield in the future.




Google Should Buy Twitter Before The IPO

A few years ago Google had the opportunity to buy Twitter but passed on the opportunity because they were developing Google+ and didn’t think they needed it.

Today Google+ is supposedly the no. 2 social network after Facebook, but I don’t buy it. No one I know uses Google+ much, if at all. And I certainly don’t see people giving out their Google+ names on the cable news networks and other TV shows. Twitter dominates there.

Google+’s user numbers are juiced simply because Google forces the product on everyone, and if you use Google to authenticate yourself to third parties, you are using Google+.

Frankly I don’t see Google+ as being any more successful than Google Video was in competing with YouTube. I might be wrong, there might be a real base of hard core Google+ users out there who start and end their day on Google+, but I just don’t see it making any kind of mark on our culture at all.

Twitter, on the other hand, is the only really massive social network where the network effect has really kicked in. It is less than Facebook in some ways, and better than Facebook in some ways.

Google needs to buy Twitter just like they bought YouTube.

It won’t be cheap.

Twitter filed a surprisingly low range for its upcoming IPO today – $17 – $20 per share. That works out to a $11 billion valuation on the high end.

That’s lower than most people expected. And it’s just pocket change to Google. The company has added around $45 billion in market cap in just the last couple of months.

At the very least it would be a hedge against Facebook if Google+ fails (which I think it has). And on the upside Google could really let Twitter blossom, much like they have with YouTube.

Also, many people I’ve spoken with think that Twitter’s data alone would be worth tens of billions of dollars to Google’s search team.

Of course even if Google did make an offer to buy Twitter, Twitter would have to accept before it could happen. And Twitter’s current executive team is probably more interested in going it alone. “Giving up” and selling just prior to an IPO would feel like losing to them, I’d imagine.

But Twitter would still have to consider the offer because the board of directors has a fiduciary duty to shareholders. If Google offered substantially more than $11 billion, the board would find it difficult to justify turning them down.

As a user I’d like to see Twitter stay independent. But I’ve thought for years now that Google is crazy for not doing anything it could to buy Twitter. Who knows, maybe there really is something interesting going on at Google+. But I doubt it.

Disclosure: CrunchFund owns shares in Twitter. I personally own Google and Facebook stock.

Unveiling Stealth Y Combinator Startup URX

URX1-1Twice a year Y Combinator shows off its most recent batch of startups at a packed-house demo day in Mountain View. The goal of these startups is to attract investor and press attention. Sometimes, though, a particular startup wants neither, and stays in stealth mode until some later date.

URX, from the last Y Combinator batch, is one of those companies. There has been no press about URX at all, until right now.

Note: CrunchFund has invested in URX, although for now the company isn’t disclosing any other information about money they’ve raised.

URX is a mobile advertising platform. But it’s not about pushing app installs (~$300 million/year just for Facebook, around $1.2 billion/year industry-wide). It’s about getting users who already have apps installed on their devices to actually use and make purchases in them via product advertisements in other apps and websites.


URX helps app developers showcase all of the products in their apps with targeted ads. The desktop equivalent of this is Google’s Product Listing Ads, which have greatly simplified ad management for e-commerce operators on the web. Deeplinks are particularly compelling for mobile commerce companies (Retail, Travel, Local, Food, Deals, etc.) that want to jump in and offer a product to one of their users in real-time.

Facebook JUST announced that they’ll be supporting ads that deeplink to other apps:

However, app discoverability is only one of the challenges of building a mobile app. Developers are also faced with the challenge of getting installed users to return and remain active within their apps. According to a study by Localytics, a Mobile Measurement Partner, 66% of app users only open apps between one and 10 times.

Facebook’s announcement is a extreme validation of this nascent market. We can expect Twitter, Google and other platforms to provide this type of mobile re-engagement ad unit as well. URX is in a good position to include this new inventory on their advertising platform.

Why is this so important? With these new ad formats, there’s no easy way for developers to manage 1,000’s of deeplink ads for their products today. URX indexes the knowledge graph of apps’ deep content, and uses machine learning to recommend relevant products to individual users.

The potential market size is tremendously large. Mobile is gobbling up an increasing percentage of total ecommerce retail sales. By 2017 eMarketer is guessing that 25% of U.S. retail ecommerce sales will occur on mobile devices, up from 15% today.

And yet…the mobile advertising market is still way behind where it should be given the tremendous growth in mobile transactions. Tens of billions of dollars behind. URX believes that product ads will raise the price floor of mobile ad inventory.

Mobile advertising is no longer just about installs and other crude first generation tools. URX is a company that is poised to introduce what’s next in mobile advertising.

URX was founded by John Milinovich, Andrew Look James Turner and Nathanael Smith. Watch the video below for more information.

(Trying To) Face Down The Evil At TechCrunch Disrupt

I’ll be on stage in just a couple of hours at TechCrunch Disrupt in San Francisco. I’m interviewing a lot of people over the next three days, including CEOs, entrepreneurs, VCs and politicians. See here for my interview schedule, here for the full schedule.

This conference is a little different for me than the dozens of ones that have come before. Why?

Last night Paul Graham asked me if I still love startups. I think what he was asking was if the chemical full on in love affair that I had with startups in the early days of TechCrunch is still going strong, or has morphed into something else over time, as all great loves mellow and evolve.

What involuntarily came out of my mouth had little to do with the question. And I’m pretty sure it’s a theme, bubbling under the surface, that is going to take up a big part of my brain space at Disrupt:

“I’m scared of our government and I’m disgusted by what little Silicon Valley has done to fight it.”

Just today I’m reading yet a new NSA document that laughingly refers to us (the people) as “zombies.”

Who knew in 1984…that this would be big brother [picture of Steve Jobs with iPhone}…and the zombies would be paying customers [pictures of people with phones, tablets]?”

We’ve been dehumanized. At best our government considers us meaningless sheep to be herded/slaughtered at will. At worst they consider us all terrorists or potential terrorists, needed to be watched at all times.

I’ve watched the months of debate. Debates where the government sucks up every bit of information on everyone it can, then argues that as long as it isn’t looking at that data the Fourth Amendment is intact.

But then argues that since they got the data legally, they can certainly look at it, too.

It’s a vicious cycle. Collecting mass data isn’t a violation of the Fourth Amendment. Then once they have it, looking at it doesn’t require a warrant.

As I’ve written before, I don’t believe this is a system to fight terrorism. I believe it is terrorism, against the American people and everyone else in the world.

I’ve brought this up in every one of my preparation meetings with the people I’ll be interviewing.

There has been some pushback. Some people don’t want to talk because lawyers. Others say this isn’t their fight and they can’t effectively lead their organizations from behind bars.

Others say they are willing to speak their mind.

In all cases I will ask the questions. We are inundated with denials and narrowly tailored “transparency reports”, but no company has stepped forward to tell us exactly what is really going on, and why the government seems so optimistic about being able to get user data from companies in real time or near real time, without any judicial oversight.

I’ll ask the CEOs if they feel any responsibility to protect their users against clearly unconstitutional (and just plain icky) government data grabs. I’ll ask them what they are willing to do, if they do feel responsible, to protect those users.

I’ll also ask the VCs and others about this, too.

For example, Right off the bat this morning I am going to ask Ron Conway, who has pushed for gun control via his Sandy Hook Promise for nearly a year now, hasn’t said a word about Silicon Valley’s role in the wholesale destruction of our human rights by the United States government.

He could do so much by leading an effort at real transparency, and a real pushback against the government. But he hasn’t lifted a finger. I want to know why.

He knows this question is coming, we discussed it last night.

And we’ll go from there. At the end of the conference I’ll write again and see what kind of answers we got from our leaders. Because I think we can handle the truth. I think we deserve the truth.

One thing I won’t do is ask the easy “gotcha” questions that I know can’t realistically be answered and that will only serve to make people look stupid. I’m not trying to just stir up the crowd at the expense of the speakers. But there are plenty of constructive and forward looking questions, including personal philosophy questions, that I think can be asked constructively. And secretly I think many of these people are just itching for the chance to say what they really think.

Amazon And Free Stuff

I’m reading Amir Efrati and Jessica Lessin’s article about Amazon toying with the idea of giving a free smartphone to people.

Of course there’s a real issue here with the cost of the device: “Offering a phone for free would be a daunting proposition. Amazon would have to find a way to make up for the cost of manufacturing — on average, $200 per smartphone.”

In 2010 I had a good source saying that Amazon was trying to figure out how to give a free Kindle to its Amazon Prime members – the best and most loyal Amazon customers who pay a yearly fee for free shipping and (now) premium digital content.

They had the same problem then – they’d do it, said my source, “Just as soon as they can work out how to do it without losing money.”

I’m not sure they’ll ever figure it out. Hardware costs and software development isn’t cheap. And if the product sucks because the hardware is dated or the software is iffy, then people won’t really want it. The kind of customer who’ll live with a sub par free product probably isn’t who they’re targeting.

Still, it’s a tantalizing idea, and one that clearly keeps coming up at Amazon HQ.

That’s One Fugly Logo, Yahoo


We forget that Google’s logo is really bad because we’ve been looking at it for so long that we don’t notice any more. But I’m pretty sure that even 10 years from now I’ll still look at Yahoo’s new logo think “That’s one godawful fugly logo right there.”

It’s a serious case of “A camel is a horse designed by committee.”

It looks like a logo that somebody would have created with clipart fonts from those CDs back in the early nineties. It lacks any personality, it’s boring, it’s banal. It’s a great big bag of fail. It sucks, badly.

I never thought a logo could be so singularly uninspiring.

On the upside, it’s definitely got people talking.

Update: It actually doesn’t look nearly as bad on the Yahoo website, although at best I give it a “meh.”


My TechCrunch Disrupt Interview Schedule

Next week is TechCrunch Disrupt in San Francisco. The event is essentially sold out, and there’s good reason – this is probably the best lineup of speakers they’ve had to date.

I’m preparing for some fascinating conversations. The schedule is still fluid but as of today here’s who I’ll be talking to on stage. They’re in the order that I’m talking to them:

    San Francisco Mayor Ed Lee
    Ron Conway, David Lee and Brian Pokorny from SV Angel
    Doug Leone from Sequoia Capital and Sanjit Biswas from Meraki
    Reid Hoffman and David Sze from Greylock Partners
    Marc Benioff/Salesforce
    John Doerr/Kleiner Perkins Caufield & Byers
    Vinod Khosla/Khosla Ventures
    Marissa Mayer/Yahoo
    Mark Zuckerberg/Facebook

I’ve interviewed almost all of these people before and I’m sure they’ll have fascinating things to say. I expect Benioff to be the most likely to say something outrageous.

Mayer will get at least a little razzing about the Vogue Magazine thing – Personally I think she should do the interview in the same pose as the picture. Probably won’t happen, though.

Wrapping the conference up with a talk with Zuckerberg will definitely be a win. Hopefully it’ll be as newsworthy and interesting as our talk a year ago.

Compliance vs. Complicity

It is further ordered that no person shall disclose to any other person that the FBI or NSA has sought or obtained tangible things under this Order.

There’s an old saying that it’s far better to ask for forgiveness than for permission. I’ve seen the many, many statements by Google, Microsoft and everyone else about how they really want to tell everyone just how bad these government demands for information are, but they can’t because it would be illegal to do so.

Today we saw it again. Microsoft and Google continue their lawsuits and negotiations with the government, and ask for credit for fighting the good fight – for example, “The purpose of our litigation is to uphold this right so that we can disclose additional data.”

But at the same time they argue that they already have the legal right to disclose – “We believe we have a clear right under the U.S. Constitution to share more information with the public.”

Ok. So do it. You believe you have the right. You know it’s the right thing to do. Forget the lawsuit. Just release the damn information.

Do it.

Bloggers Fail To Molest John Doerr At YC Demo Day


I was at Y Combinator Demo Day today (it was a very good batch of companies). At one point during a break I went over to talk to the gaggle of TechCrunch reporters there to cover the event.

We were talking about the various companies and what not when John Doerr walks by. Apparently to have a phone call in a quiet corner of the room.

“That’s John Doerr right there,” I said.

They looked at me.

“He’s all alone, probably on a secret phone call,” I added.


“Go get ’em! Get your iPhones out and start the video and ask him hard questions!”

“That would be quite rude,” said Colleen Taylor (or something similar), pictured far left.

“Yeah but he’s practically begging to be attacked by you guys. HE’S HAVING A SECRET CONVERSATION RIGHT BEHIND ALL THE BLOGGERS. IT’S LIKE HE’S MOCKING YOU!”


So I watched, yearning for my younger years when I would have made a nuisance of myself while performing an unwanted and impromptu video interview of one of the most important people in Silicon Valley.

I think it was better back then, when sources were at least a little afraid of the bloggers.


Digital Ocean v. AWS: 10x Performance For 1/3 Cost


Cloud hosting startup Digital Ocean announced its first round of funding today – $3.2 million from IA Ventures, CrunchFund and TechStars. See TechCrunch, GigaOM and Pando Daily for coverage of the news (and see the comments on Hacker News).

The company is growing revenue and customers by 30% a month, and had 50x growth from December 2012 to May 2013. Thousands of customers have been grabbed from competitors like Amazon and Rackspace. When Netcraft called Digital Ocean’s growth “meteoric,” they weren’t kidding.

I first heard about Digital Ocean from Nik Cubrilovic. Nik urged me to take a look and try to invest in the company if we could. Digital Ocean has crushed the traditional price/feature tradeoff for hosting. From Nik:

I first heard about Digital Ocean almost a year ago and its name kept coming up every few weeks, which prompted me to finally take a look. It took me 30 seconds to signup, and I immediately realized what they had done – they broke the traditional price/features tradeoff and were somehow offering a great control panel, a great API and a great service with features usually reserved for high-end services at a low VPS price. They were outside of the traditional hosting tradeoff quadrant of price/features. I fired up one server to test them out and the benchmarks came back with numbers that were as good as what other IaaS providers were charging 3-5x the price for.

They have combined the low costs of low end commodity VPS providers with the features of high-end IaaS providers, it seems so obvious now but nobody has been able to pull it off until Digital Ocean.

I have cancelled two low-end VPS services and 30% of my AWS servers so far, with a view of also migrating PaaS hosted apps at Heroku and AppEngine (such as my own personal website) to Digital Ocean as well. The great recent news is that like AWS, Digital Ocean is now available as a provider in Vagrant:

Digital Ocean provider plugin:

If you are a developer who deploys apps to servers and aren’t using Vagrant I strongly recommend you check it out. It sets up virtual servers for development that can then be pushed to the cloud. After initially setting up Vagrant and provisioning with either shell scripts, Ansible, Chef (solo or server) or Puppet (apply or agent) it is a single command to push your development environment up to a new Digital Ocean instance. You will need to scale it automatically and provide your own load balancing using either DDNS or nginx, but you get very close to building your own Heroku or AppEngine type stack with your own customizations, no lockin and at a fraction of the price.

This combination has changed the way I work and has made it easier for me to drop in and out of other projects with developers without spending time on servers configuring. I can’t recommend Digital Ocean enough to other developers, my tip would be to start small with non-critical apps to get a feel for the platform and then work from there – especially by using Vagrant or any other tools (even in-house continuous integration or deployment scripts using their very simple to understand and use API) to automate. Digital Ocean allows you to assemble cloud stacks like lego bricks rather than the completed tool Heroku or AppEngine give you, and you get service as good as more expensive options, for what they do, at low end VPS prices.

Just how much better is Digital Ocean than its competitors? For the lowest end hosting, we’re talking about 10x the performance for 1/3 the cost:

The 512MB server at Digital Ocean, which is the smallest size they have, is $5 per month or $0.007 per hour. Its UnixBench score is 1060.5 on average, IO is 279MB/sec and bandwidth 21MB/sec. Here are the details:–1-cpu-digitalocean

To get the equivelant performance on AWS from a single server, you have to step up all the way to an extra-large instance, which is $374 per month (!), compare the scores and graph here:

The cheapest AWS server is the Micro, which is $15 per month, its performance scores are approximately 10% of the cheapest digital ocean server:

some might complain that AWS isn’t suited to benchmarks, especially the Micro instance which is throttled, but the difference is just too large to ignore.

Digital Ocean was founded by Ben Uretsky, Mitch Wainer, Jeff Carr and Moisey Uretsky. We are extremely happy to be investing in the company – saving developers a ton of money on hosting while also giving them a product they adore is a surefire recipe for success.

%d bloggers like this: