Category Archives: Discussion

Google Acquires reCAPTCHA: For the tech or the distribution network?

Recently Google acquired CAPTCHA provider reCAPTCHA one of the leading CAPTCHA providers on the Internet. The key strength behind the reCAPTCHA implementation of the CAPTCHA test is that it pairs a known word (to the server) with an unknown word that an OCR scan has failed to recognise. This allows reCAPTCHA to crowdsource the digitisation of scanned books such as those in the Google Books project as Google outlines in their blog post on the acquisition:

“This technology also powers large scale text scanning projects like Google Books and Google News Archive Search. Having the text version of documents is important because plain text can be searched, easily rendered on mobile devices and displayed to visually impaired users. So we’ll be applying the technology within Google not only to increase fraud and spam protection for Google products but also to improve our books and newspaper scanning process.”

The above publicly stated reasons for the aquisition seem obvious and  perhaps they are a little too obvious, hiding the real reason behind the aquisition. reCAPTCHA hold no specific patents for the technology behind their text CAPTCHA algorithms (At least none they discuss on their website or are able to be found on the US Patents & Trademark Office site) and given that reCAPTCHA operates mostly on open source software the case for buying Google buying the company gets thinner.

Given that Google could easily code their own reCAPTCHA equivalent, this business deal goes beyond the obvious. Google already has their own ‘CAPTCHA Killer’ that operates using images and video, surely they could roll that out if they were really serious about security, so the tech involved with reCAPTCHA is not compelling from this perspective. ReCaptcha’s Prof. Von Ahn has already licenced his ESPgame image labeling program to Google (Now known as Google Image Labeler) so buying reCAPTCHA might have been an attempt to grab this technology as a bundle with the companies other assets.

Certainly using reCAPTCHA to digitise and make searchable Google’s vast collection of books in the Google Books archive project, however given the above they could have designed a similar system themselves. What is really key to this purchase by Google is in fact the existing distribution network of websites that are already using reCAPTCHA and the API they have created to allow new sites/uses of the reCAPTCHA system. Given the 100,000 sites and 30 Million CAPTCHA’s served daily by reCAPTCHA, the decision by Google to buy reCAPTCHA is about the text processing volume it gives them immediately, not the technology behind it.

Twitter is *now* and Facebook is *Last Weekend* (Why Facebook can’t win in real time search)

With the FriendFace deal I’ve written about before the R&D team from FriendFeed was touted as giving Facebook a leg up in the real time search world against Twitter, and while Friendfeed has some great technology to integrate into Facebook, until we see a fundamental paradigm shift in how people interact with Facebook. The service will always lag behind Twitter in the timeliness of posts, and hence, always be behind the 8-ball when it comes to being a true real time search engine.

Why?

In an increasingly connected world, application support on mobile devices is critical to enable people to participate in their social  networks on the go. They need to be lightweight and easy to update your status or participate in your ‘stream’ in the spare moments a person has on the bus, between meetings with clients etc. Facebook has focused a lot recently on developing a new version for the iPhone “Facebook 3.0” which purports to be much easier to use and offers a better optimised user experience than the mobile Facebook site which is “underdeveloped” to put it kindly.

Facebook 3.0 moves to address some of the shortcomings that have limited its use on mobile platforms, but the Facebook ‘environment’ does not translate well from the desktop to the mobile screen. The sheer number of features and plug-ins and other services that Facebook offers can not easily be transferred to a mobile device and this will continue to impede adoption of Facebook by mobile users.

Twitter on the other hand, is inherently optimised for the mobile space. Being limited from early in it’s evolution to messages of 140 Characters (Twitter originally had no message size limit, but this changed soon after the service was born) it has by default a natural place on mobiles, essentially a multi-cast version of SMS. This is both a natural extension of behaviour users are already accustomed to and is an ideal fit for the mobile form factor.

Sure Twitter mobile applications have added many bells and whistles over the basic function of sending 140 character messages, applications such as Tweetie on the iPhone and PockeTwit on Windows Mobile (The screenshots of PockeTwit don’t do it justice, it’s a great App) have added media service integration (uploading of photos, GPS integration with mapping services etc) but the core competancy of these applications is the short succinct ‘tweet’ which drives their function.

Facebook cannot compete with this ‘limitation’ without neutering the experience desktop users have with the service. Which is why when it comes to what is happening now Facebook can’t compete, and is limited to being the repository of the photos from the party you had on the weekend when you finally upload them a week later.

Twitter is the real winner in the FriendFace buyout.

On first inspection of the FriendFace deal, with Facebook paying $150Million (In stock/cash) for Friendfeed it looks like Twitter just lost a potential buyer in Facebook and might be limited to brokering a deal with Google or Microsoft, or actually having to find a way to monetise Twitter on their own.

All is not lost though, Twitter was never a good deal for Facebook, the price was too high and the benefits were limited for Facebook (who was looking for someone they could integrate into Facebook, not an external site separate to the Facebook ecosystem). For Twitter, giving up sovereignty to Facebook was never an option, and there was never a strategic fit between the two companies.

So for Facebook, Friendfeed was by far the better buy. Not only does Friendfeed integrate better into the Facebook environment (Facebook has been copying Friendfeed UI elements for a while now) but because of it’s stagnant growth despite continued innovation (Growing 0.26% in July Vs Twitter’s 16%) Friendfeed was brought to the auction table at a bargain basement price.

What is important to keep in mind here was Facebook bought Friendfeed not because of its user base, but because of it’s technology. Which is where twitter wins out of this deal.

Why?

After being bought out by Facebook, Friendfeed has no future as a stand alone product anymore. Sure there are ruminations about the site continuing, but the exodus has begun and Friendfeed users, unsure about the future of the site, no longer trust Friendfeed to exist in a years time. With this cloud over the future of the site, investing their time through continuing to use Friendfeed has a limited payoff and they are looking for alternatives.

Facebook has made a land grab for disenfranchised Friendfeed users with ‘Facebook Lite’ but given that these users just lost their ‘home’ because of Facebook, this is a psychological impediment to moving to a Facebook property.

Where will they go?

Twitter.

Why Rupert Murdoch’s decision to charge for content online could save the news industry.

After Rupert Murdoch made the decision to charge for online access to content across his suite of newspapers online properties, there was considerable consternation across the Internet. With many armchair pundits in the blogosphere/twitter-verse crying foul at the idea of charging for content online that was previously available for free, citing their own usage expectations and a belief that the content isn’t worth paying for given the perceived drop in journalistic standards and cost cutting amongst newspapers. While I have considerable time for the case against changing to a cash for content mode, I believe in Murdoch’s case it has merit and might actually save the hemorrhaging news industry. Here’s why;

Charging for content online suddenly makes printed newspapers relatively less expensive

So you are running a newspaper company, what is your biggest cost? Sure journalists salaries figure in here somewhere, but it is the actual printing presses and distribution networks that are the significant cost centres in a newspaper operation. With declining readerships of actual newspapers over the last 10 years, advertisers are paying less for space than they used to, and less income is coming in from actual newspaper purchases and the fixed costs of running the printing press aren’t getting any lower and such make up a larger proportion of your total operating costs.

In order to make purchasing an actual news-‘paper’ a more attractive proposition for consumers, one way to drive that change is to make other sources of news relatively more expensive. If you charge for the online version of your newspaper, the actual printed version becomes an attractive option for consumers who had moved to the previously free online alternative. This will drive a percentage of readers back to the paper version and patch up the hole in the traditional newspaper component of Murdoch’s enterprise. For how long, this is debatable, but in localised markets where Murdoch’s papers have near monopolies over content/distribution, it will have a larger effect than in more competitive geographies.

Murdoch doesn’t own the entire news industry, many news sites will remain free and become profitable (or at least, lose less money)

Assuming that all the readers of Murdoch’s online properties do not continue with the site in a cash for content capacity, their eyeballs and the advertising revenue they represent will venture elsewhere. What this means is that free online news sites that were really struggling to meet their costs with a drop in online advertising revenue (exacerbated by the GFC) will be able to stop the revenue bleed with a greater share of the news audience. This ‘tiering’ of the online news audience is good for everyone provided that Murdoch can make a compelling value proposition for readers of his sites to pay for content that they unable to find elsewhere (Ideally, shifting the focus to quality journalism in lieu of the click generating sensationalism seen now).

Will all this save the news(paper) industry?

Maybe. In the long run the competitive forces of citizen journalism and sites such as wikinews will continue to place downward pressure on the operational costs of running a newspaper. With the costs of distribution online almost zero there is nothing stopping an upstart news operation running out of a garage to unseat a large monolithic operation such as News Corp if they can’t continue to offer a decent value proposition while charging for content.

At the very least, this decision will breathe some life back into printed newspapers until some industry shake out has occurred and people adapt to the new online news landscape. This might just buy Murdoch some time to regroup, will it be enough?