TechCrunch Disrupt SF Battlefield Semifinals: The Judges React

The TechCrunch Disrupt Startup Battlefield semifinals started off in an unorthodox way, with moderator Paul Carr bringing on TechCrunch co-founder Mike Arrington to join tech celebrity judges Ron Conway, Hadi Partovi, Marissa Mayer, Roelof Botha and Matthew Cohler onstage.

While we wait for the final winners to be announced, you can parse through my notes on all of the fun, below.


Unlike Dropbox which syncs your file to the cloud, Bitcasa allows you to write to a cloud, using your computer the same way you would normally, but the utility is writing the file to the cloud. Bitcasa intelligently determines what files you would access, and caches them on your hard drive. It’s significant because it gives you infinite storage in the cloud, subverting the typical model of writing files to your hard drive and then having to move them when you store.

In beta it’s a free, freemium model, $10 a month for infinite storage.

Judges’ feedback:

Mike Arrington tweeted out last night that, “Bitcasa is a game changer unless they screw up.” Referring to the fact that there might be three Dropbox investors onstage, “How will you grow your user base considering Dropbox is out there?” Bitcasa answered that their value proposition is fundamentally different from Dropbox, “We’re about the new hard drive,” not about file sharing, “You still need to have space on your hard drive for Dropbox.

Judge Marissa Mayer also expressed concern that Dropbox was a competitor and wondered about the sharing process, “When you’re sharing with a mobile device, are you storing things on the client, or is it purely streamed?” The answer is that the device will stream it from the cloud in realtime, over the network.”

Botha brought up the cost factor of streaming all that data, which Bitcasa addressed by saying that they use a caching model, which reduces costs. Matt Cohler wanted to know what was the most important problem the startup solved for consumers, to which the startup responded that they never want you have to pull out another USB stick or hard drive again, whereas with Dropbox you still need the space.

Ron Conway brought up the fact that it’s basically going to be a marketing battle between Bitcasa, Dropbox and Boxee and the startup really needs to focus on what differentiates it. “Dual video caught my eye, that seems to be proof you’re different. Beef up the marketing, because this will be mad.”

After a Partovi question about security, Cohler brought up that he saw much of this core functionality with GDrive. “Why are you going to be successful?” To that the startup said, “My mother could use this and she’s not a computer person. It’s the Apple model, we just want it to work.”


Shaker is a social environment that allows you to use your Facebook profile to interact with people around you, “like a Second Life, except you’re yourself.” The platform is for social experiences, which allows you to visit virtual places like bars or offices, incorporating the Facebook API in order to include stuff like a Like wall where you can see what you have in common with the other people in the room.

Shaker has had almost 10K users since its beta launch.

Judges’ feedback:

Ron Conway began the feedback by asking how similar the app was to Second Life, and what happens when the app morphs into a dating site. The Shaker founder talked about how it used your real identity and how it was more meaningful than Second Life, “They’re creating meaningful relationships. Yes, if you build a bar, expect to see a dating or a flirting scene,” he said, but referred to the fact that the app wasn’t just about dating.

Matt Cohler asked about the app’s peak concurrent user number, and the Shaker founder said that he’s only opened it in Israel without any PR and they had to shut down invites to 540 people. At peak hours the app has 5% to 6% of usership, “Shaker is live and vivid 24/7.”

Botha likened the app to Snow Crash, a Neil Stephenson science fiction book that refers to virtual worlds, “It gives you a sense of what is coming. I like it for stuff beyond dating, like listening to music.” “What has surprised you about the usage?,” he then asked.

“When we tested different graphic environments, we got completely different interaction,” the Shaker founder said. “Even with the same communities, no one’s dancing on the bar during morning.”

“It’s cool to try to change interaction,” Marissa Mayer said. “One concern I have is the number of people. My question is around how you decide to move in the space… If it just is dance mode then it’s just glorified chat.” The Shaker founder said that with regards to movement he considers the figure in shakers more like profile carriers and not avatars, “They carry your profile. You really communicate with the profile.”

Hadi Partovi asked whether they’ve ever split up Shaker into two rooms, “How do you decide who goes in which room?” The Shaker founder said that they had actually opened it up for testing in one room, and kept adding rooms, “That’s not the strategy moving forward, we have a system which allows for infinite number of rooms.”

“I’m a big fan, I’m an investor,” said TechCrunch founder Michael Arrington “I’m continually surprised by the breadth and depth of startups out of Israel … If I was Marissa I would say we’re buying this today, immediately make it Google + related.”


CakeHealth wants to be the Mint of heath, allowing you instant insight, bill reconciliation and personalized recommendations for various healthcare programs.

Judges’ Feedback

Marissa Mayer: There were a lot of things I really liked, recommendation engine key is a core piece of tech, can double in your insurance. Google health didn’t work out very well because of data restrictions. Insurance data isn’t accurate. An insurance company would say that you had Cancer when you didn’t have Cancer.

CakeHealth responded that 80% of bills have errors but they had built-in alerts for errors, “There are common things we can look for by tapping into databases; we started out with insurance claims. As we move up the chain we can see even more detailed information.”

(Yes, Mike Arrington did take a pee break)

Roelof Botha: This would be great if it lives up to the promise of solving problems, every time I’ve had an issue with a health care company it’s because they’ve made an error.

CakeHealth: Machine learning is too strong a word, but we’re creating rules for what can and cannot happen.

Hadi Partovi: I can’t imagine taking out a bill, and taking a picture of it with my iPhone. I’d rather just tell my doctor to send it to CakeHealth.

CakeHealth: Taking a pic of the bill is temporary. We’re looking at other models that are not quite in place yet. We are still dealing with a very paper-intensive process by taking a picture and then transitioning into those other models as they become viable. We’re looking at the Netflix model, envelope to house.

Matt Cohler: Distribution, market?

CakeHealth: It’s consumer oriented, employers are great marketers because it will streamline process, reduce amount of inquiries.

Ron Conway: How does this monetize?

CakeHealth: It’s a Referral model: sign people up for plans and savings accounts. bigger opportunity foundation, action oriented negotiating transaction fee.


TalkTo allows you to text any local business for information as easily as you text your friends. It looks just like SMS, nothing new you have to learn. If a business is not on TalkTo, it reroutes the request to a call center which calls the business and texts you back your answer.

Judges’ Feedback

Ron Conway: This reminds me of Yext. How are you communicating with businesses, customers that are texting you to TalkTo?

“We’re going through a call center,” the startup said. “If you sign up for the premium plan we’ll keep making the calls.”

Marissa Mayer: I like this business a lot, we actually bought one of your competitors Talkbin. I’m worried about non-responsivenes. What kinds of guarantees of responsiveness do you have?

Matt Cohler: I would encourage you guys to think about not charging the consumer at all.

Roelof Botha: One of the things that I worry about is the consumer expectation around direct messaging. It will take a very long time. Most of the time you’re going to end up with a “can’t get ahold of them response.”

Hadi Partovi: You should give an email address to every business.

Mike Arrington: I’m worried about the cost getting ahead of this business, but it’s a huge win. I honestly think Marissa should buy this, and Shaker and Bitcasa. It sure would be nice as a consumer.

Marissa Mayer: We do think the space is interesting which is why we acquired Talkbin.

Prism Skylabs

Prism Skylabs uses its video intelligence technology to fuse images together into realtime story boards showcasing a space. It aims to change the way businesses deal with and highlight customer flow.

Judges’ feedback:

Roelof Botha: There’s a lot there. What’s the crisp online hook?

Prism Sky Labs: Video is a huge problem for SMBs.  We could probably build a whole business just doing video for businesses,  but we want to get into the social media aspects. We really want people who are out there to see great views of what is going on.

Marissa Mayer: Obviously, this is related to what we do. Business owners are very particular about the state of their ship. People want to put their best foot forward. People want the privacy silhouettes to be perfect.

Prism Sky Labs: A lot of the things we’re dealing with go away as we accumulate more images. Background, we’re learning what the background of the place looks like over time. We have all that data. We can give businesses merchandising tools.

Hadi Partovi: I want Mike to say Marissa should buy this.

Mike Arrington: Marissa should buy this. Ron has created a lot of value out there. He could start a lemonade stand and I’d invest.

Hadi Partovi: In terms of trying to solve hard problems technically you’re actually doing that. I would try to figure out which of the 10 things you could do and pick the most exciting one.

Ron Conway: The use case for this product is going to morph a lot. This could turn into a huge company. Do you have a unique IP that enhances clarity?

Prism Sky Labs: We’re registered 20 inventions,  provisional patents, since July 1st.


Farmigo is an online local food subscription service that aims to bring people fresher, tastier, healthier, cheaper and more convenient food by connecting consumers directly to farmers.

Judges’ feedback:

Mike Arrington to Hadi Partovi (who is an investor): “Why do you think company should win? ”

Hadi Partovi: The food industry  is 10% of the GDP. It’s a trillion $ industry, not billion, and disrupting it is good for our health.

Marissa Mayer: Groceries are a low margin business, but full disclosure: I don’t cook. There’s a lot of overhead managing behavior as a consumer. What happens when you sign up with too many vendors?

Farmigo: There shouldn’t be a concern about the activation goals, as we’re exceeding them. Once it’s hit exceed goal in a week, it’s not going to dip back down. Food subscription is a change in the way people buy their food.

Roelof Botha:  I’m concerned about adoption from producers. Also it might suffer from a Netflix subscription problem, how people never watch things in their queue. I can’t anticipate if I want peaches next week, what if I want apricots?

Farmigo: We’re focusing on UX, so each grower could use it easily. From the producer’s side it’s usually an admin using the system which means they’ll have comp literacy.

Ron Conway: Awesome example of a concept I was trying to explain to Mike, Collaborative Consumption. I’m selling my Safeway stock.


Trello is simple collaborative software that lives in the cloud. Since it’s launch yesterday, it’s amassed 26K users.

Judges’ feedback:

Ron Conway: Who do you compete with?

Trello: Salesforce, Findbugs, Microsoft Project construction teams. But we’re vastly simplified, our way of doing it is different.

Roelof Botha: This would cut away half the email I receive. But what if you’re in a group of 10 and two don’t adopt? How do you draw people in?

Trello: The way I use this product: make the board, and update it with what I think he’s working on.  He starts to get value out of looking at what I think he’s working on.

Marissa Mayer: I think it’s impressive. I worry that Salesforce is a lot more customizable. What happens when there are 1000 requests? Can it scale?

Trello: I probably wouldn’t use this if I had 1000 requests, there is an inventory of feature ideas that I wouldn’t implement. Nobody’s going to use Excel to do their taxes. The vertical applications will be better in some cases, Salesforce for your sales pipeline, Bug Tracker for bug tracking,

Hadi Partovi: One of the hardest challenges is explaining to people what it will be used for. How do you say that in one sentence?

Trello: Organize everything together. You kind of have to rely on people’s imaginations.

Paul Carr: Mike you have the last word.

Mike Arrington: Now you’re giving me performance anxiety. I think this pretty good.