Category Archives: technology

In this interview,  Andrew Mason shares an interesting bit of web trivia: Groupon originally launched as a subdomain of The Point: groupon.thepoint.com.

After unsuccessfully attempting to buy groupon.com from an entrepreneur in England, they were finally able to get the owner to sell by enforcing their trademark rights.

It is interesting to think how many major web properties originally launched with a sub-optimal domain name: thefacebook.com, twttr.com, and playfoursquare.com come to mind. And both Google and Yahoo were both originally launched on the stanford network.

Any others to add to the list?

Like many people, I was a bit shocked and saddened to hear the announcement that yahoo was going to “sunset” delicious last week. Amidst all the talk of a new bubble, it was sobering to hear that one of the signature web 2.0 sites  might shut down.

Given the response to the announcement, it seems likely that delicious will be sold, if the site can be decoupled from yahoo’s login system. If it does, that creates a new question for the new owners: how to turn delicious into a business.

At the current level of traffic, delicious probably doesn’t have the scale to make an advertising model work. And one of the initial concepts for a business model, using tags to improve search results, isn’t well-suited to running delicious as a stand-alone properly. So the future of delicious probably depends on how well its future owners apply DHH’s secret to making money online: charging for the product.

This raises what I think is one of the most interesting questions for a web service: how to craft the right package of features and usage limits into a premium version that people will pay for. Embedded in this question is how to define the right “unit” for a web service — that is, the best metric to enable variable pricing based on usage.

Web services that foster communication or collaboration often have natural “units”. For example, the primary unit for 37signals Basecamp is the number of projects. You can create one project for free, but adding additional projects requires a paid plan. The unit for MailChimp is the number of subscribers — the first thousand are free, and the cost of a subscription grows after that.

Interestingly, both of these services have other potential “units” that they chose not to charge for. While other business software, like Salesforce.com, is frequently priced based on the number of “seats,” Basecamp allows you to add an unlimited number of users, which drives growth of the service. Paid plans on MailChimp allow you to send an unlimited number of messages, which encourages active usage.

Services that target towards individual users are more challenging to price. I think flickr’s model is  pretty effective. First, uploads on flickr’s free account is capped to 300MB and 2 videos per month, which provides a reason for active users to shell out $25 for the pro version. The interesting part of flickr’s model is how it encourages retention: if you let your pro account expire, it only displays the last 200 images in your photostream. This provides an effective hook to drive retention. (At least it did for me, anyway.)

So what is the right “unit” for delicious? I’ve used delicious to bookmark almost 1,500 pages over the last five years. I can’t say I visit my archives regularly, but I do find value in knowing that they are there. I haven’t got much out of the social features on the site, since I primarily interact with delicious through bookmarklets. So for a user like myself, I think I would pay for storage / access to my bookmarks and want unlimited tagging for free.

It will be interesting to see how delicious’s fate plays out over the next few months. Let’s hope the new owners can find a viable model to keep the site around for the long-haul…

One of the trends we discussed at Marketspace is that now “Every Company is a Media Company”.

(The underlying premise — best described in Jeffrey Rayport’s book Best Face Forward — is that more and more, companies are competing on how they go to market, rather than what they sell. This raises the importance of forging and sustaining relationships directly with consumers, particularly through new digital channels.  To give a couple examples, this describes why Nike is shifting its marketing budget from TV to online services, and why Dove is blurring the line between content and advertising.)

Upendra Shardanand provides another example of this trend: Purina and Kellogg’s are using Daylife to create their own customized news properties. Purina’s “Pet Charts” is a social news site for pet lovers and the Team Kellogg’s site aggregates news and photos on ten members of the British Swimming team. Neither site has a strong community element: Team Kellogg’s does not appear to take submissions, and on Pet Charts users can vote on, but not submit articles.

It will be interesting to see what reaction these types of sites will receive.

I would expect it to get a great response from marketers, since Daylife’s capabilities will allow them to create full-featured sites on limited budgets.

For consumers, I’m not so sure: they should be able to attract an audience, but I suspect that they will have a hard time driving repeat usage (at least in their current form), unless they add additional features to make the sites more interactive.

Whenever I visit the Globe, The Times, or the WSJ’s site, I usually turn to the the “most e-mailed” feature first. It rarely contains the top stories, but it almost always points me to something something “interesting“. While startups like digg and reddit have attracted most of the hype surrounding social media, I actually think the greatest impact of social news could be redefining how we experience traditional media.

In late June, the NYT pushed this concept forward with the launch of their new TimesPeople feature (PaidContent review), which allows you to recommend articles to a set of your friends. The beta version is currently available as a Firefox plug-in, and the final release is slated to be incorporated into the site later this year.

I have been playing with TimesPeople for the past week or so, and while I’m excited about the concept, I find the current beta version lacking. Two complaints:

First, it actually took a while how to figure out how to recommend articles. The plug-in adds a “Recommend” at the top of the article tools box. This approach is too subtle: since it blends in so well with the existing site, I didn’t recognize that it was connected to TimesPeople. I would add a big “recommend” button to the TimesPeople toolbar to make that association clear — the model would be like the “tag” icon that the delicous plug-in adds.

Second, as you can see from the top right, I don’t have anyone to share articles with. Currently, there is very limited support for building your network: you can search TimesPeople users and scan your Gmail contacts to see if they are TimesPeople users, but you can’t invite anyone else to join. I suspect this is by design (in order to limit usage during the beta) but the effect cripples the service.

Building a social site is like throwing a party: it takes a certain critical mass to get it off the ground. In an old interview, Caterina Fake talks about how the flickr team personally greeted each new user to help get them started with the site. Six months after flip.com shut down, it remains to be seen whether the NYT, and other big media companies, can embrace the type of scrappy tactics that have made the upstart social sites a success.

This morning I came across the new Stock Screener functionality in Google Finance. Like a lot of Google projects, it improves on the current state of the art with a speedy ajax interface and a minimalist layout.

The coolest feature is the use of sparklines to help you create your screen. This displays small graphs showing the distribution of each metric included your screen. I think this provides a really interesting way to analyze the market.

Here are three improvements I’d like to see in this product:

• Ability to filter companies by industry: currently, filters are set up by by sector (e.g., Energy, Healthcare) so you can’t pull up a view of Retailers, or Pharmaceutical companies
• Add more fundamental data, like revenues, or cash flow
• Exporting the results of your screen to Excel or Google Spreadsheets

Fred Wilson asks an interesting question: it is possible to build a business around browser extensions? He thinks the answer is largely no, since the audience is too skewed towards early adopters.

My take is that browser extensions will have their greatest impact as catalysts for related web businesses.

The most obvious application is to use a browser extension to drive traffic to your site. The portals realized the value of this early, particularly for search queries, and have aggressively pursued revenue-sharing partnerships to secure distribution for their toolbars.

But browser extensions can do more than just drive traffic to a site: they can also make it more valuable. Broadly speaking, “Web 2.0″ sites all generate value the same way: by aggregating a particular type of information and allowing it to be searched and shared – think photos on flickr, restaraunt reviews on yelp, playlists on Last.fm. (As Tim O’Reilly says, “Data is the Next Intel Inside.”) These types of sites all face a similar critical challenge: generating enough content so that the site becomes useful. This makes data entry a potential “killer app” for browser extensions. By lowering the interaction costs associated with entering data into a site, browser extensions can jumpstart the value of a service.

Delicious is a classic example. Its Firefox extension seamlessly integrates the browser’s bookmarking functionality with the delicious service; it just takes two clicks to tag a site. Today, the browser extension is driving Delicious’s growth.

Wesabe, a web-based personal finance site is a newer example. Wesabe adds tagging to Quicken-like personal finance functionality. This lets you categorize and analyze your own expenses, and also provides a means to generate tips on saving money (by anonymously aggregating data across users). The major hurdle to using the site is importing your data. Wesabe has created automated connections to most major financial institutions, like ING and CitiBank, but if you bank at a smaller institution, you need to manually download your data and then re-upload it to Wesabe.

To alleviate this problem, Wesabe created a Firefox extension that allows you to automate the upload process. After installing the extension, you log into your bank and download your data, while the Wesabe Extension records your actions. Once created, the upload process becomes a one-click activity. The experience is very similar to recording an Excel macro.

Here’s one to watch: 23AndMe, a California startup co-founded by Anne Wojicki, Sergey Brin’s wife, has launched its initial Personal Genome Service. (NYT Coverage, WIRED Coverage)

To use the service, customers submit a saliva sample, which is analyzed for a set of SNP’s  on an Illumina microarray. Then, the data is analyzed by 23AndMe’s software. The initial offerings from the service include a Gene Journal that provides personalized risk assessments, analyses of one’s ancestry and family tree, and general visualization tools to browse your genome.

It will be interesting to see if this can get any traction in the marketplace, given the $999 price tag, and the positioning toward general interest in genetics, rather than therapeutic applications. I for one, would love to do this. But I can’t quite afford it. Someday.

I saw a mind-boggling article in the times a couple days ago on how McDonald’s is experimenting with using call centers to take drive-through orders. Apparently this helps improve service quality, as well as efficiency, since one operator can serve several restaraunts.

Umair Haque commented that this is an example of the trend towards unbundling the corporation. Since our IT department couldn’t set up a FTP service for me this morning, I am all for it!