My Perfect Startup

Nicholas Carr had a short riff yesterday on social networking:

…[that] social networks have the effect of turning up rather than turning down the volume of our self-consciousness seems to me precisely right. The Net turns the social instinct inward, which ends up fencing in rather than freeing the self.

There was a related Guardian story recently, “Social networking under fresh attack as tide of cyber-scepticism sweeps US“.

The Guardian article is the more comical and confused. Have you been swept by the wave of skepticism lately? Me neither. The whole story screams of… an attempt to create a story. It’s a controversy! People are very upset! Academics are writing articles!

But the real story here is how astonishingly widespread social networking has become – 600 million people on FaceBook, 200 million on Twitter, 500 million on Skype – with very little social resistance. There was way more controversy about the spread of video cassettes in the 80s than there is now about the spread of on-line social – yet at the same time, FaceBook and Twitter really are changing how, and how much, people interact – a profound change.

How could a change so profound happen with so little resistance?

It can only be because it fits so naturally with how people want to interact.

It allows communication with the people users want to reach, and in the ways users want to reach them.

It allows users to curate their personas – something that everyone does in the “real world” also, but it is easier to do, and requires less use of metaphorical sharp-elbows, online than in person. Perhaps this is the reason why seriously shy folk sometimes turn more chatty on line?

It allows users to communicate without talking over anyone and without having to buttonhole someone directly (the later a big advantage of FaceBook over email).

I’m doubtful of Nicholas Carr’s point that online social is a solitary act of self-presentation, whereas real-world interactions are a chance to liberate the individual consciousness into a group setting. People are presenting themselves in the real-world at least as much; and people can brainstorm and be a group member online just as they do face-to-face.

As a startup, I would be highly skeptical of attempts to address the anti-social or privacy-invading “downsides” of social networking; much of these downsides are thing that most users don’t care about, whatever the media might like to think. The great bulk of opportunities for startups in social will come from making the experiences more fluid, more relevant, and in tying commercial opportunities to social activity in various ways that users find useful or at least acceptable.

Best article I saw on the $315m acquisition of Huffington Post by AOL was at Gigaom.

Most commentary suggests that AOL overpaid out of desperation, driven by the need to do something before their legacy business finally stops generating cash.

At first blush, the metrics for Huffington Post suggest it is a significant asset. Quantcast has huffingtonpost.com at 40m visitors per month, compared with about 14m for the New York Times (nytimes.com).

On the other hand, nytimes.com has 115m visits-per-month (8 visits per user), while Huffington Post has 93m visits-per-month (2.3 visits per user). Huffington Post is flat year-on-year for page-views, and down 40% in the last six months. Low rates-of-return-to-the-site, and falling page-views-per-user, are consistent with the feeling that Huff Post has been drifting more towards being a “content farm” that pulls in users to look at a single article that is designed to score highly on Google search: this one about Superbowl start times is a fine example of search-engine candy. Certainly, the site’s original “draw” as a left-of-center celeb-fest must have had the air let out of it by the decline in political enthusiasm on the American left in the past year.

Given this trajectory – that, and the fact that they were offered 10-times trailing annual revenues – it is easy enough to see why Huffington Post may have been a willing a seller.

Can the deal still make sense for AOL?

The risk of loss of focus by employees, or decline in editorial quality or reputation under AOL, is more than outweighed by the likelihood that superior advertising inventory and targeting can drive revenues, and that the AOL advertising machine, shared across multiple sites, will be more cost effective than one driven by Huffington Post alone. So far, so good.

Nonetheless, in a recent post on “The Daily”, I wrote about a possible outcome for news:

Media conglomerates form that capture enough content that they can be their own aggregator. Though it would not be surprising to see this attempted, I suspect that sources of news are too diverse to make it practical in the long term.

I still think that attempts by conglomerates to be entry-points or in-house aggregators of content will not work. Huffington Post’s relatively modest monthly-visits-per-user (2.3) confirms the thought – for most of its visitors, it is a place to visit occasionally, not a hub or a daily habit. More and more news and comment sites will be of that form, so no shame in it for Huff Post; but when news sites are visited occasionally, not continually, it does suggest that buying up a number of sites and then trying to be the entry-point for a significant number of users will fail.

A more flexible and expansive form of aggregation is still needed.

For startups, beyond the possibility of inventing and executing on the perfect form of next-generation news aggregation, the lesson of this deal seems simple – if your business trajectory is uncertain, and a strategically motivated buyer offers you a huge multiple to acquire your company, you are probably wise to accept.

Fred Wilson (blogging venture capitalist) wrote:

… just spent 20 minutes reading my timeline from this morning back to yesterday morning. I have built an amazing set of people I follow, 564 of them, all curated one by one over the past four years. The timeline is so rich, so full of different things from different people. Tech, sports, politics, music, family stuff, humor, and way more.

This is indeed a fine testament to Twitter, but it once again highlights two major issues.

1. His feed is curated by him, which takes a lot of time and effort – relatively little is done to help automate that for the average user
2. It is – to a surprising extent for a “social” medium – a passive viewing channel only, not an information-flow-with-communication. Passive viewing has its place, but Twitter could be much more – and for many users, it’s not fulfilling that additional potential.

For a typical user, automation will beat curation 9 times out of 10 – as we’ve discussed before.

And so… how’s the “Personalized Trends” code going? Didn’t do much today, but I did move my little test program from being a MAC app to an iOS (iPhone / iPad) app. Really, it was easy.

Comprehensive review of the new iPad newspaper, “The Daily”, at Gigaom.

The Daily seems nicely done, albeit with some complaints about clunky top-level navigation (JPEG artefacts / blocky thumbnails, I mean come on…).

Yet it seems  a relatively conservative effort to rethink the newspaper for the iPad.

What are some of the areas where a newspaper could rethink more radically?

1. Games – this is one area The Daily does seem to have latched on to, at least a little. Newspaper’s always used to be the source of crosswords, sodoku – and a newspaper with a large readership has opportunities in multi-player games, too. The “daily” nature of the newspaper can be exploited in games.
2. Real-time – perhaps the most obvious, but plainly any digital newspaper can provide a through-the-day stream of updates. Even so, there is user-experience problem to be solved here – how to make the stream relevant; and how to make it as obtrusive as the user wants (not more nor less obtrusive). And this includes video, not just text/photos.
3. Self-customization – learn what you’re most interested in, highlight that
4. Group customization and interaction (“social”) – do your friends read the same content you do? Do you want to know what they’re reading? Do you want to know what they’re thinking abut what you’re both reading? Since one of the reasons people read the news is to discuss with friends and colleagues, what friends and colleagues are reading – or could be encouraged to read – is almost certainly relevant. FaceBook, LinkedIn, Tumblr, Instagram, and Twitter all provide potential mechanisms for interconnecting people and their news.
5. Advertising platform – can you target based on interests, based on on location? Does the newspaper provide advertising inventory management?
6. “Finishability” – how do you provide people with the boundaries they want around how much time they prefer to spend reading the news? This is of course a big advantage of a conventional printed newspaper – once you’ve read it, you’ve read it – but how could that benefit be replicated in the digital world?

FlipBoard, a news aggregating startup, is one company that has pushed this as far as anyone. A key thing to note with them, though, is – they are a news aggregator, not (primarily) a newspaper in the sense of journalists producing original content.

It seems like there a few ways this could go:

• 99% of all news content becomes free / advertising supported, and news is consumed largely through aggregators who provide the services 1-6 described above. 1% of news is subscription for highly specialized topics.
• 25% of news content is still subscription based, but the subscription model is transformed to allow it to be accessed via aggregators. This might also drive another change, where star journalists create their own subscription business without needing access to a newspaper brand
• 20% of news content is subscription based, but it is “freemium”, with users being lead into paying by free content available on aggregators (Gigaom itself, headed up by Om Malik, is an example of that style)
• Media conglomerates form that capture enough content that they can be their own aggregator. Though it would not be surprising to see this attempted, I suspect that sources of news are too diverse to make it practical in the long term.

Choices, choices…

In “Devices, Media and The Consumer“, talking about the need to back-up cloud, I wrote:

While I think backup could be part of the story, I’m skeptical that it is a strong-enough motivation – people find it to easy to ignore until the worst happens. I think that i) universal access (any device); and ii) Organizing media and communication in one place are the real drivers, with backup coming in third. Perhaps it’s notable that Dropbox have taken this kind of approach with documents, and Picassa with photos – they emphasize access, and sharing, over backup, even though backup is more-or-less explicit part of the offer too.

So, today comes a counter-example (via TechCrunch). Mirco Wilhelm had 4,000 photos on Flickr (photo-site owned by Yahoo), they were deleted by an administrator at the company. Even though Mirco is a paying customer, there is no backup and no redress.

As a general matter, we can assume that no consumer Internet service, up to and including email, has any backup of user data; some business services, such as a Microsoft BPOS, have at least some measure of backup and/or data redundancy.

Data backup / redundancy is expensive. Obviously, at best it increases storage costs substantially. It also increases networking costs. If the redundant/backed-up data is at a different site – as it really should be, to guard against risks of fire / earthquake / storm – it becomes even more expensive. Then there are the software challenges of what to do when one data copy is destroyed – it is almost impossible to create a complete new copy at a new location given the bandwidth limitations of storage systems and networks.

So, there is an argument as to why consumers should be interested in backing up cloud and non-cloud data alike.

But I still think most people would ignore it, and would need stronger more day-to-day motivations to adopt a service.

Maybe it would be different if a major cloud service failed and wiped out users’ data, not just one at a time, but for thousands or millions of people. Let’s hope that doesn’t happen…

Update: There is a startup targeting back-up of online data – Backupify. There already claiming to cover quite an array of services; Flickr, Google Apps, Twitter, FaceBook, LinkedIn. Pretty nice. Still not sure I can be bothered, though… even though I probably should.

Most Startups Will “Build” Their Solutions By Sticking Together Pre-Existing Components

Yesterday, I started noodling some code for my little “personalized trends on Twitter” idea [See here, and here for previous posts on this topic]. I’m not in love with this idea yet, but I like that i) It can be simple at the start; and ii) It has the potential to grow, for instance by embracing and simplifying other forms of communication, including email.

I thought I’d start by doing it as a prototype iPhone or iPad app.

Not having started a clean-slate new project for a while, there’s a lot I don’t know here. I have just a week or two of iPhone programming knowledge. And I have never used the Twitter API.

When I dig into Twitter, I quickly found that there’s more “new stuff”  – new to me, anyway. Twitter uses “oauth” – open authentication – to enable apps to have access; and the Twitter oauth process is quite complex (I always love when technical documentation includes a “Phew!” at the end). Twitter’s main API is JSON (JavaScript Object Notation, text-based structured-data, like a simpler XML) based. Both oauth and JSON are new to me.

Plus of course the details of the Twitter API is completely unfamiliar.

How to proceed? Turns out, it’s about sticking stuff together.

There is a free objectiveC library, MGTwitter, for iPad/iOS/MAC, that hides most of the complexity of the Twitter API.

OK, but MGTwitter example code doesn’t really use oauth, it uses xauth, which is less satisfactory for various reasons.

Ah – but another free objectiveC library, oauthconsumer, will handle most of the hard stuff – like assembling SHA1 signatures – for oauth.

Another library handles the JSON processing.

Some careful Googling shows up the Apple documentation for popping open a web-page, and the Twitter documentation explains how to open the correct page for app authorization as part of the oauth flow.

Amazingly quickly, considering the potential complexity and opportunities for getting stuck, the app can authenticate and start pulling interesting data from Twitter.

What’s the lesson? The set of available third-party technologies that are freely available has become remarkably rich and broad. There is a whole set of software startup ideas which can be implemented by sticking existing technologies and APIs together in creative ways. The “skill” may not be the ability to go deep on a narrowly defined problem, so much as to elegantly conjoin existing components to create something worthwhile.

“Narrow and deep” startups still exist – for example On2, acquired by Google for their video encoding technology (WebM) – but they seem likely to be the exception.

Back to the code now…

Ars Technica (and others) are reporting that Android-based mobile devices have passed Apple and even Nokia in volume in the final quarter of 2010.

[Credit: Canalys via Ars Technica]

What’s of note about Android? Let’s face it, it has become an imitation of Apple’s iOS. Note: I use an Android phone, the Galaxy S, as my main phone, though I also use a variety of Apple devices. For the user, Android is “like iOS, only slightly messier, and slightly worse.” And Apple’s industrial design is still superior, too. Why then has Android been able to make progress in this way?

It really is because it is open, and open-source – and so free to anyone who wants to use it – despite Steve Jobs’s comments. Any manufacturer, any carrier, can pick it up, do whatever they want with it, make it what they want it to be. Consequently, every route to market, every positioning, every geography, every price point is covered by the creativity of the general market. Just as Linux did in server systems, the open-source approach creates a flood of low-cost experimentation that submerges more established, more expensive, and less flexible platforms, despite the apparent superiority of the existing non-open solutions.

Nokia’s CEO, Stephen Elop, stated that, due to commercial pressures, Nokia is considering “multiple ecosystem partners”, which most took to be a hint of an Android future for Nokia. Further reviewing his remarks, I might guess that he would look to use Android for mass-market devices, reserving Nokia’s high-end for the new(ish) MeeGo operating system.

Yes, Apple’s iPhone remains the most highly valued by users, the BMW(*) of this market while Android perhaps amounts to Toyota. Yet if Nokia does switch partially to Android, Android phones will amount to over half the total cell-phone market, up from 10% just a year ago; and Android will have 3 or 4 times Apple’s volume.

As a startup, the lesson that gets learned again and again is that it is hard to compete with free, and that exploiting “free and open source” gives you a chance to grow, iterate and disrupt in a way that conventional go-to-market approaches rarely do.

To go down that path, you do need to find a way to be operate in a lean fashion and/or live off advertising dollars (FaceBook being the example of the moment), or else find a rich parent as the Android team with Google. And you have to really achieve the potentially massive uplift that free-and-open-source can give you.

(*) Granted, the BMW position can be very profitable, too.

I recently moved my iTunes library – which includes all my music as well as iPhone and iPad apps – to Dropbox, an online file synchronization service, for two simple reasons: 1. To ensure it was backed up to an offsite location; and 2. To automatically synchronize it between multiple laptops (Apple’s “home sharing” system being pretty feeble at actually keping multiple machines in sync).

Media-device and web-service proliferation is already becoming a reality. Here are some of the devices and services many people are using:

• Full-size laptop / desktop computer
• Netbook, tablet and/or mini-laptop
• eReader
• Full size iPod / MP3 player for house or car
• Micro/wearable iPod / MP3 player for walking / running / biking
• Smartphone(s)
• Feature phone(s)
• Portable video players
• Home TV-based video storage and player / set-top-box / DVD-player (Tivo, Apple-TV, Roku etc.)
• Facebook (key location for photos and videos)
• Twitter, Tumblr or other micro/mini blogs
• GMail or other email service

The range of digital media potentially needing storage, and/or needing to be accessible across multiple devices, also proliferates:

• Music files
• Apps (for many platforms…)
• Photos
• Videos (self-shot)
• Professional video including TV and Movies, DVDs and Blu-Ray
• eBooks
• Documents
• Communications – Email, IM/chat, SMS, tweets, blogs
• Other personal information – calendar, contacts, to-do-list
• Web-service subscriptions

I find myself using DropBox for documents and videos – it does a good job being accessible from most devices.

I use Picassa for photo and video storage.

Nothing for professional video – TV/movies – I just have a Sony BluRay player.

My communications – email, IM, blogs – plus personal info – calendar, contacts – are done on the web (GMail, Skype, WordPress, Twitter, Facebook, Tumblr). Is it noteworthy that they are not backed up – if wordpress.com looses this blog, or GMail looses my email, I have no copy of it?

Many people are expecting streaming content to overtake stored content for professional music and video. Apple has always resisted Blu-ray disk support for this reason, believing that streaming should overwhelm Blu-ray. And streaming music is making some inroads in Europe (Spotify). Yet the digital-rights-management required by content providers is a huge barrier; and where traditional media delivery companies (Comcast) are integrated with content providers (E!, Style Network, G4, The Golf Channel, Versus, NBC), it becomes doubly challenging, since the content providers have a vested interest in maintaining the profitability of the traditional distribution method and holding back the rise of Internet streaming.

Even so – surely there must be some way to control this proliferation of content, devices, and delivery mechanisms? Clearly, the average user doesn’t have time for it. Streaming surely must play a part. It could be personal streaming, where the user keeps their content library in the cloud and can stream it anywhere; or it could be professional streaming that connects to hosted content (as Netflix is doing now); or a combination.

Could back-up be a way-in for a startup? For instance, backing up GMail so that I’m not at the mercy of Google’s backup strategies (if they have any). While I think backup could be part of the story, I’m skeptical that it is a strong-enough motivation – people find it to easy to ignore until the worst happens. I think that i) universal access (any device); and ii) Organizing media and communication in one place are the real drivers, with backup coming in third. Perhaps it’s notable that Dropbox have taken this kind of approach with documents, and Picassa with photos – they emphasize access, and sharing, over backup, even though backup is more-or-less explicit part of the offer too.

Perhaps there’s something in the “personal streaming” – perhaps intersecting with the rise of HTML-5 technologies, so you play on any web-enabled device?

Needs more thought to become an actual start-up idea.

Alex Williams, writing at ReadWriteWeb, held up Google Apps as an example for Larry Page, Google’s new old CEO to follow. He offered up a few reasons. Per Alex, Google Apps:

• Contains a credible set of core components
• Fits with a number of associated services
• Fits with other Google products, Android and Chrome OS
• Is a “framework” (can build on it)
• Attracts “young developers” who use the framework
• Does not involve space travel

The last point is undeniable (I think?), but I nonetheless feel that Alex’s conclusion is wrong.

Let’s look at Google Apps differently. It is:

• Derivative (of Microsoft Office)
• 80% baked: GMail scalability and simplicity is great, but the UI is as ugly as any major website; their conversation view is a good idea, but inconsistent in user experience; shared calendars, great but buggy; spreadsheet UI is buggy – and so on.
• Undifferentiated – too much of the “differentiation” – it’s implemented using web technologies! – is geeky stuff many ordinary users don’t care about
• Not quite a corporate-grade product (too flakey), and not quite consumer-suitable (too geeky, GMail aside)
• Money-loosing, and probably always will be (on current trajectory)
• Fighting the last war, against Microsoft’s Office product, not the current one (social) nor the next (uber mobility)
• Hard to transitioning from existing (Microsoft) setups due to weak migration tools
• In fairness, the extensibility is worthwhile

What lessons does this hold for startups?

Most importantly, aiming for the new, the “whitespace on the map” rather than well-explored territory, makes it much easier to build a real business. Established business areas can be lucrative for startups (including by creating strategic value for acquisition), but relevant-to-users differentiation and relevant-to-users problems-to-be-solved are needed.

Secondly, do what you can do excellently. And no more. Focus on that excellence. The 80%-baked problem is a killer for creating actual value for customers – and it’s something Google seems to suffer from, with their bottoms-up development process.

I can confirm from (bitter?) experience that these two points are actually close to being a single point. If you target something new, you won’t be dragged so easily into feature-comparisons with existing products in the space, and won’t be tempted, or forced, to try and do too much.

So no – while Google Apps has its merits, I would not hold up as a model – either for a startup, or for Google.

Spent most of today watching my wife’s U.S. naturalization, as well as driving up and down to the ceremony. It might seem cheesy or sentimental, but surely it means something that people from 104 different countries came to Oakland’s Paramount Theater today to become citizens of the United States.

Meanwhile, there’s been some coverage of Collecta’s decision to terminate support for their real-time news/activity API, for instance ProgrammableWeb’s Is It Finally the End for Real-time Search Engines?

Collecta provided a real-time news feed that was adopted by plenty of web sites, enabling those sites to provide real-time aspect to their pages. Collecta provided web-APIs and sample code to allow web sites to straightforwardly add the Collecta feed. They were getting decent adoption, and could show nice usage graphs for their service.

What’s the problem?

Firstly, it is hard to monetize an intermediate widget when someone else owns the customer. The massive usage volumes that Collecta would have required to monetize successfully were not coming their way, not as a startup, anyway.

Secondly, a lot of the value of information comes from analytics, especially personal, social, or group analytics, not simply from the mere aggregation of content or simple content search-ability.

Of course, aiming at “enterprise” analytics / news, rather than consumer as Collecta did, would be an alternate strategy; but enterprise carries its own set of challenges.

Yet another approach would be to curate the feeds much more actively – have a celebrity feed, an Apple feed, a Justin Bieber feed…

Collecta’s technology looks decent, and they have a nice team. For me, the lesson would be that something that is highly personalized for each viewer, or to each area of interest, is going to be more compelling than simply providing a “stream of stuff”. Real-time and personal and part of an integrated user experience that is more than a widget – that might work.

Which leads me right back to this.