We also have five new part-time partners: Michael Seibel, Steve Huffman, Dalton Caldwell, and Andrew Mason. As the name suggests, part-time partners advise startups like regular partners, but part-time. Michael was cofounder of Socialcam (YC W12) and now works at Autodesk, which acquired it last year. Steve is cofounder of Hipmunk (YC S10) and before that was cofounder of Reddit (YC S05). Dalton is cofounder of App.net and before that was cofounder of Imeem. Andrew was cofounder of Groupon and till recently its CEO. We've known all these guys for years and we can already tell it will be great to work with them.
I'm proud to say that I have accepted a “part-time partner” role at Y Combinator. I have known PG & Jessica for years, and have tremendous respect for them. The upshot of taking this role is that I'm going to be spending a few hours a week helping out YC companies.
Given that running App.net is a full-time job, I had to carefully weigh the pros and cons of taking on an additional time commitment. After consulting with my co-founder and some of my advisors, I came to the conclusion that spending this time with early-stage startups will make me better at my job at App.net.
Specifically, the #1 thing we can do at App.net right now is ensure that we're building the best social platform for developers creating the next big thing. As PG has put it, understand your users. I like to think that being exposed to this many startups will help keep me honest.
I look forward to getting to know more of the YC community, learning as much as possible, and (hopefully) offering some useful advice and perspective.
An app can afford to be prescriptive about the user experience and means of interaction; in fact, the best apps have a point of view on how the user ought to use their service.
Platforms, on the other hand, are just that: a stage for actors (i.e. apps) of the user’s choosing to create a wholly unique experience that is particular for every individual user.
It follows, of course, that no successful platform can be built on advertising. Advertising demands eyeballs; platform success demands the ability to fade into the background as said unique experiences take center stage.
One thing I have noticed studying web traffic logs is that daily human behavior at scale resembles a sine wave:
Over the years, my mind's eye has constructed an image of the daily ebb and flow of web traffic resembling the ebb and flow of ocean tides. I've learned that code tends to break at the “high tide”, and the best time to do site maintenance is at “low tide”.
So why does human behavior at scale look like a sine wave? It's because of our daily cycles of waking and sleeping. In other words, we have a 24 hour cycle because of circadian rhythms, which occur because of the cycles of darkness and light caused by planetary movement.
In 2002 I went to see a live show at the coffee house on my college campus. The show featured the band The Microphones.
“The Microphones” was a somewhat confusing name for the band because there was only one member: Phil Elvrum.
I went to the show early, and watched Phil personally haul in his own gear. He then proceeded to take out a marker and white butcher paper and make some larger-than-life drawings. Once completed, he hung his drawings behind the stage as scenery. Next, he set up his microphone and amp and played a show. When it was over he gave away the drawings to audience members and stood next to the stage to sell LPs and T-shirts. Each LP had original hand-drawn art on it. I bought one and talked to him for a minute.
Phil Elvrum was 23 when that show happened, and in retrospect, it is difficult for me to imagine what it must have been like to be in his shoes. He seemed very brave to me… as far as I can tell, no one was traveling with him to help out or offer moral support.
I've been talking to a number of independent writers and software developers recently, and it has made me think a lot about what their day-to-day lives must be like. I've never been a freelancer, or worked in a solo capacity myself, but I have to imagine it must sometimes feel like doing a solo music tour. You show up alone, you do your thing, then you haul out your own gear… whether or not the show went wonderfully or terribly.
Going “indie” is clearly not for everyone, but I have the utmost respect for the people that are able and willing to do it.
My gripe is that, as usual, Microsoft has jumped its own shark and is out stomping through the weeds planning and talking about far-flung future strategies in interactive television and original programming partnerships with big dying media companies when their core product, their home town is on fire, their soldiers, their developers, are tired and deserting, and their supply-lines are broken.
xBox’s primary critical problem is the lack of a functional and growing platform ecosystem for small developers to sell digitally-/network-distributed (non-disc) content through to the installed base of xBox customers, period.
Throwing money at media companies & marginalizing independent devs: always a winning strategy.
Microsoft is living in a naive dream-world…any lead they feel they have is simply not defensible by licensing other industries’ generic video or music content because those industries will gladly sell and license the same content to all other players. A single custom studio of 150 employees also can not generate enough content to defensibly satisfy 76M+ customers. Only with quality primary software content from thousands of independent developers can you defend the brand and the product.
Marc Andreessen recently made some bold statements about the future of the retail business:
“Retail chains are a fundamentally implausible economic structure if there’s a viable alternative,” he says. “You combine the fixed cost of real estate with inventory, and it puts every retailer in a highly leveraged position. Few can survive a decline of 20 to 30 percent in revenues. It just doesn’t make any sense for all this stuff to sit on shelves. There is fundamentally a better model.”
I am not sure if I agree with this prediction of his. However, these sentiments have caused me to start thinking about how 3D printing could increase the likelihood of those statements coming true.
The death of music & video retail chains: instructive?
Picture the amount of capital and energy it took to create companies that stocked and sold CDs/rented DVDs. In a music or video rental retail location, the only differences in the stocked product was the visual packaging and slightly different encoded information, i.e. pieces of roughly identical shiny plastic disks.
Having large amounts of physical infrastructure dedicated to running a supply chain of roughly identical pieces of plastic seems ripe for disruption. It's not as if the CDs were lovingly handmade by craftsman with one-of-a-kind album art.
Perhaps the era right after the availability CD burners (but before iPods were introduced) is a decent approximation about what the world will be like with widespread availability of 3D printers. A CD burner is somewhat like a 3D printer: they both take digital information as input and use it to create a useful physical object.
Big Box Stores
I tend to associate “big box” stores with places I used to go to purchase music. Remember how much square footage that used to be dedicated to music and DVD sales at Best Buy and Circuit City? There is still some square footage devoted to CDs at Best Buy, but it seems to occupy about as much space as is dedicated to iPhone accessories.
What percentage of the square footage of Home Depot is devoted to items that could cheaply and easily be 3D printed? Let me give you an example: the lowly plastic drywall anchor. Would anyone care if their drywall anchors came out of a 3D printer?
What percentage of items stocked in a Target could easily be 3D printed? My thought first goes to toys. Also sports equipment. What percentage of Target's square footage and revenue consists of these kinds of items?
The role of debt
In the debates I have participated in regarding 3D printing, one of the points made is usually something like “high-end clothes, food/perishables, heavy-duty tools, and hand-crafted custom items are not going to be replaced by 3D printing.” Let's assume that is true: the vast majority of products have very little risk of being displaced by 3D printing using current technologies. The problem for retail stores is the revenue sensitivity caused by debt.
Taking the example of Borders (or Blockbuster or Circuit City or Tower Records or Virgin Megastores or Linens 'n Things), it's not as if revenues went to zero overnight and so they decided to close their doors. No, the issue is that they were highly leveraged. This is also the reason that Best Buy could declare bankruptcy in the near future: It's not that their revenues will go to zero, but that relatively small revenue declines could cause a ticking time bomb of tight cashflows and corporate debt to explode.
If we accept Andreessen's argument that most retail companies could be put out of business by a 20-30% decline in revenue, 3D printing could be plausibly be the vector by which this scenario is manifested.
Aside: I have been working on a blogpost about the major label marketing machine during the 90s, and thought of this story. Since this anecdote didn't fit anywhere in my other post, here it is.
Radiohead is a band that first gained popularity in the early 1990s. They managed to break out and get heavy radio and MTV support for the single Creep. In a lot of ways they were a conventional early 90s band.
Their third album was called OK Computer. This album has been heralded as one of the best records of all time. Seemingly out of the blue, after a couple of conventional records, they came out with something critically and commercially fantastic.
I have heard that the opposite of “jumping the shark” is “growing the beard”. So yes, they grew the proverbial beard.
This is a story I heard from someone that worked at Radiohead's record label during this time. Given that this story was told to me as a casual anecdote, it should be treated as such.
Finally, after a lot of drama, they sent us over the final mix of “OK Computer”. We all got in a conference room, sat around the table and put it on.
We listened to it all the way through in one sitting. Some people took notes.
After the album finished playing, there was silence, and the people in the room looked despondent. Someone finally said: “what the f-ck did we just listen to… how are we supposed to market that?”
After he told me this, I went back and listened to that record, and tried to pretend that I had no context at all: no album artwork, no reviews, no testimonials to it's greatness. That made it much easier for me to imagine the emotions they felt in that conference room.
Making sense of something on the fly, without any context, is difficult.
Knowledge Representation is a central topic in the field of artificial intelligence. Research in this area tackles the challenge of mapping the external world into simplified, machine-readable models. Creating a sufficiently sophisticated model is as important, if not more, than the algorithmic processing that occurs once the model is built.
The more external states that can be represented in a model, the more complex we could say it is. Simple models can cause people to falsely believe that successful results can easily be “scaled up” to more complex models. Unfortunately, it's rarely that simple.
A well-known example of this in AI was the early promise demonstrated by the SHRDLU project:
The result was a tremendously successful demonstration of AI. This led other AI researchers to excessive optimism which was soon lost when later systems attempted to deal with more realistic situations with real-world ambiguity and complexity.
To summarize: don't assume that a successful AI project which utilizes a simple model will seamlessly scale up to more complex circumstances.
The complexity of large companies
It's the job of a large company to understand what the outside world wants to buy, and then mobilize their corporate infrastructure into delivering the right product, at the right place, at the right time.
For the sake of illustration, let's use Apple as a stand-in for a large company. A single human being could never understand all of the constituent pieces of work that goes into shipping the hardware and software components contained in a new iPhone. But no single Apple employee needs to understand all of the pieces. Apple has a number of well-run subsystems which each tackle a piece of the process that ultimately manifests as a white box with an iPhone 5 in it. Apple is very complex organization… but that complexity is what allows them to ship very complex products.
There is, of course, a downside to organizational complexity. Complexity tends to breed more complexity. When you talk to employees of a large company, it seems as if their entire world is made up of byzantine internal politics with no relation to customers in the external world… a large percentage of every day is spent attempting to navigate ever-shifting politics surrounding the org chart. It would seem that the larger a company gets, the more the company tends to allocate resources towards inwardly focused (as opposed to outwardly focused) issues.
Organizational complexity can show up in odd ways in a shipping product. For instance, why does a single product, the Microsoft Surface, give consumers the option between two very different OSes: one model with “Windows RT” and another with “Windows 8 Pro”? I am sure there is an official reason why this is Totally A Good Idea, but it mostly seems like Microsoft decided to ship their org chart.
Early-stage startups aren't very much like large companies, and that difference certainly applies to organizational complexity. Nonetheless, a startup with one or two employees must create some sort of internal model of the external world regarding what do people want, when is this ready to ship, what is the best technology for us to use to deliver our product, and so on. Even with essentially zero bureaucracy, it's easy to get hung up on inwardly-facing projects that don't actually matter in the long run.
When we talk about things like “focus” and “simplicity” and “doing just one thing right”, we are more or less discussing complexity management. If a company tries to do too many things at once, the company org chart and workflow will naturally begin to model that complexity.
From the perspective of a startup, having tools at our disposal such as AWS, outsourced HR, or streamlined billing providers let us keep our internal model of what it is we are supposed to be doing with our time as simple as possible. Simple models are the easiest to understand and modify on the fly.
When complexity in a startup is good
As a startup grows, it becomes necessary for the scope of the internal model to grow so that it can effectively serve the demands of the outside world.
The simplest example of this is a dramatic increase in traffic. If a startup gets 100x more traffic, then the organization is faced with an immediate, overwhelming amount of work required to keep their service operating. So the startup spends time creating a more sophisticated hosting setup, or implements database sharding, or hires someone with an expertise in scaling. What once was as simple as rolling out some code to Heroku is now something with far more moving parts. In this example, the more quickly and effectively the startup recognizes and adapts to the newly-introduced external complexity, the better.
To take another example, a startup generating increasing revenue/sales must intentionally create additional internal complexity. To deal with increasing revenue, an organization must create/refine the internal process regarding things like customer support, sales support, refunds, taxes, billing and accounting. Even if the decision is to outsource all of those functions and take on zero additional headcount, there is still systemic responsibility for those functions to be adequately performed and managed. A founder probably shouldn't say: “I believe in simplicity and focus, therefore we will have no systems in place to handle accounting or customer support.”
A successful increase in product market fit, combined with a successful increase in organizational complexity to handle increased demand, can feed on itself to create a virtuous cycle of growth. This virtuous cycle is how, once in a blue moon, a startup can metamorphize into a large company.
Whenever I deal with what feels like an overwhelming problem, or get spun out trying to make sense of the myriad potential things I could be doing better, or hear about a company having a tough time because of things like a founder falling out, lack of focus, employee infighting, bad hires, can't hire fast enough, hired too fast, etc, I can't help but think of this quote:
More companies die of indigestion than from starvation
- David Packard
Years ago, when I first heard this quote, I dismissed it as a useless platitude. However, the older I get, the more I have come to appreciate the wisdom contained within that short and simple statement.
If you look at virtually any Web 2.0 application, whether its YouTube, whether it’s Flickr, whether it’s Photobucket or any of the next-generation Web applications, almost all of them are really driven off the back of MySpace… Given that most of their traffic comes from us, if we build adequate if not superior competitors, I think we ought to be able to match them if not exceed them.
@twitter I've been a long-time user of twitter for news and information. Happy to be joining board of this very exciting company.
Interesting things about Peter Chernin's announcement
It is apparently his first tweet ever(!)
To be fair, the account appears to have been created in 2010. Perhaps he just uses it to read news, or to customize his Flipboard account.
His announcement was formatted as a direct reply to the official Twitter account.
This means the announcement would only be seen by his followers that also follow the official Twitter account. I don't get the feeling he did this on purpose. An experienced Twitter user would know to add a “.” at the beginning of his message so that his followers would see it.
He refers to himself as a “long-time user of twitter for news and information”
Think about that for a moment. Not only does he consider himself a “long-time” user, rather than a novice, he is defining exactly what Twitter is to him: a way to consume news and information.
From a product perspective, Twitter needs to either:
Rethink onboarding of new users to do a far better job of educating people about how things like @-replies, RTs, hashtags, etc work.
Admit failure and give up on trying to get normal people to tweet. Pivot the company at this late stage in the game and completely redefine what the core user experience is.
It seems they are opting for the latter option.
What is pre-pivot Twitter?
“…an online social networking service and microblogging service that enables its users to send and read text-based messages of up to 140 characters” - Wikipedia
In other words, the core user experience of Twitter is the sending and receiving of messages with other people. It's a communications tool.
What is post-pivot Twitter supposed to look like?
The best way to consume “news and information”.
Important content is mostly created by media companies, whether they are blogs, television, radio or movies.
The main reason that “normal users” would write messages is as a backchannel to discuss media events such as the Olympics, Election Coverage, or a new television show. “Normal user” tweets are something akin to Facebook comments.
Even though this backchannel exists, it's not expected that brands and celebrities are supposed to pay much attention to everything that is said. Chernin himself hasn't replied to the numerous replies he received.
The Discover tab is the future. Rather than forcing normal users to make sense of a realtime stream, they can see what content is trending.
Systematic destruction of companies built on the back of MySp…Twitter.
In this paradigm, Twitter's business model is to help brands “amplify their reach”. A brand participating in Twitter can certainly distribute their content for free and get free organic traffic, but if they want to increase their reach, they need to pay.
It's no accident that this sounds exactly like the emerging Facebook business model. As discussed in that link, algorithmically filtered primary feeds are vastly easier to advertise against vs unfiltered feeds. The issue for Twitter is that Facebook already has a far larger userbase which is already trained to read an algorithmically filtered feed.
How is Twitter going to pull off their mid-flight pivot, which entails largely redefining what Twitter actually is, not to mention how most people are supposed to use it? Your guess is as good as mine.
Nonetheless, Peter Chernin's announcement shows us the future of Twitter: a media company writing software that is optimized for mostly passive users interested in a media and entertainment filter.
Severalprominentpeople are up in arms about Facebook charging for access to users who have already Liked their page.
I believe this debate is missing the big picture, and what we are in fact witnessing is the unfurling of the full-fledged Facebook business model. Facebook is showing us how they will cross the chasm from low-CTR low-CPM ad-units into what investors have been waiting for since the beginning: a Facebook analogue to Google Adwords.
The newsfeed optimizer
Picture the firehose of potential newsfeed stories that Facebook could show you. There are undoubtedly candidate stories that newsfeed never shows you originating from people that you have forgotten are on your friends list at all. Why is this? In reality, Facebook is already shipping a sophisticated real-time optimizer. Every time you post a story the algorithm attempts to quantify how interesting it is, and is constantly updating the interestingness score based on comments, likes and clicks.
If you only log into Facebook once every few days, the optimizer will go back over that timespan and show you the most interesting content it can during those few days. If you log in every 10 minutes it will show you the most interesting fresh content it can during those 10 minutes. Think about this for a moment: newsfeed is doing realtime engagement optimization that works well enough to be used by a large percentage of the world's population. Newsfeed already has all of the properties of an ad-server… it's just that it has been built to optimize on engagement rather than revenue.
The best ad is indistinguishable from content
We can expect to see Facebook deemphasizing traditional advertising units in favor of promoted news stories in your stream. The reason is that the very best advertising is content. Blurring the lines between advertising and content is one of the most ambitious goals a marketer could have.
Bringing earnings expectations into this, the key to Facebook “fixing” their mobile advertising problem is not to create a new ad-unit that performs better on mobile. Rather, it is for them to sell the placement of stories in the omnipresent single column newsfeed. If they are able to nail end-to-end promoted stories system, then their current monetization issues on mobile disappear.
Open Graph = candidate stories for paid promotion
The idea of Open Graph is that developers and marketers create new verbs and nouns that can be used to model different actions being taken. ie “Bob just watched a video on Socialcam” or “Jane just planned a trip on TripAdvisor”.
Facebook has been systematically testing the amount of virality and engagement of the newsstories created by Open Graph. (You should also note that a large percentage of them are related to ecommerce transactions.) Every single Open Graph action that occurs is a candidate for paid promotion.
Do you see where this is going? Open Graph outsourced the widespread creation of incredibly flexible ad-units that look like newstories… in fact they are newsstories. This provides a huge amount of potential ad inventory for Open Graph partners. For example, if person X bought a deal on Fab.com, it would make sense that Fab would be interested in paying extra for this story to be promoted to person X's friends. Without Fab's initial Open Graph integration, that newstory/ad-unit wouldn't exist to begin with.
The thing to pay attention to here is the fact that Open Graph partners are creating billions of candidate ad-units every day. The vast majority of those stories will never be seen, but that doesn't really matter, does it?
SEO/SEM = Fan Page Management/Paid placement
On Google.com, the ad-units are paid search results which show up at the top and side of search pages. (I recall reading a study that said that most casual Google users aren't even aware that those are the result of paid placement.) From a marketers perspective, the best kind of Google traffic to get is “organic”, (ie free) that is the result of Googles algorithm ranking your page as highly relevant in search results. In addition to organic placement via SEO, a marketer will purchase paid placement to supplement and expand their reach and search-related revenue..
This is analogous to the Facebook playbook: Facebook users and Pages that create high value content which is algorithmically determined to be engaging will get organic visibility in the feed. Additionally, candidate stories that would not normally be placed into the feed by the engagement optimizer can be paid for wider newsfeed inclusion. Good marketers will implement Open Graph everywhere they possibly can: to get both free organic traffic and an ample supply of candidate stories for paid promotion.
From a Facebook user perspective, we don't really know how the algorithm works, and are already trained to understand that some things magically show up our newsfeed and some things don't. If a normal user sees that one of their friends bought a LivingSocial deal at the top of their feed, they may/may not click on it, but they certainly won't question how it got there.
Facebook newsfeed is an embodiment of our war on noise. We depend on the newsfeed optimizer to protect our limited attention span, and as a consequence, Facebook gets to choose what stories we do and don't see, just as Google chooses which search results we do and don't see. Conceptually, this seems very lucrative: Facebook is auctioning off our limited attention span to the highest bidder, as long as the bidder has a candidate newsstory to promote.