Most creators of widgets have focused from the beginning on utility - in order for a lot of people to put (and keep) a widget on their site, it has to contain something useful. The obvious stuff includes weather, news headlines, stock market info.. sure there are plenty of novelty widgets out there, but I suspect a lot of people who do install them end up removing them once the novelty wears off.

Not that short-term novelty content is a bad thing, but it’s hard to build a business on. I recently came across a different approach, on a food-related site called Nutrition Data. Given that they’re owned by Conde Nast you can safely assume they’re interested in having a solid business model. What makes ND’s widget interesting is that it offers functionality and not just content.

The ND widget allows users to look up the nutritional specs on various types of food. It pops up a separate window for the search results, so as not to annoy host sites by redirecting their user traffic. The popped-up window is on their main site, so they get the pageviews and ad revenues from that point on.NutritionData is a pure media property, their only income is from ads, so the widget is definitely a traffic play. But what they’re doing is offering functional utility in order to draw traffic in, rather than pushing their content out through their widget. It’ll be interesting to see how much of this creeps into the rest of the widgetsphere, stay tuned.

In the physical retail business, it’s common for manufacturers to pay retailers to get the best shelf space. The most common way is called “co-op” or MDF (Market Development Funds), the manufacturer pays the retailer cash that the retailer spends on local ads or other promotional programs to help sell the product. A more extreme version is a slotting fee that a manufacturer sometimes has to pay simply to get their product onto the shelf.

The practice has recently spread to the e-commerce world, facilitated by companies like Guidester. In the old days, when you visit an ecommerce site and type something into its search window, the results you get are ranked by relevance, using some type of search algorithm. What Guidester offers e-retailers is the ability to sell placement in their search results, on a pay-per-click basis. Kind of like Google AdWords, except in the main results rather than alongside them.

Search results for “digital camera” at ritzcamera.com

This of course degrades search results - the whole reason Google runs paid ads alongside their regular results, and never mixes them. Google also gives preference to ads based on relevance to the search keywords. If retailers start selling their search results (via brokers like Guidester) to the highest bidder, you could easily get completely random stuff showing up instead of the products you were looking for (see screenshot above).What will keep this from getting out of control is the same thing that motivates Google - when you only get paid for click-through, you care whether see things they’re interested in. The same thing keeps retailers honest with co-op and slotting fees, the product does have to sell for them to make money, so they won’t put just anything on the shelves. And if they play their cards right, e-retailers can get some extra cash for selling stuff they would have sold anyway.

The real losers are smaller manufacturers - for a while it looked like e-commerce would level the playing field by making more obscure products readily accessible to consumers. But if e-retail search results are sold to the highest bidder, consumers will have a hard time finding smaller manufacturer’s products even if they’re looking for them. It’ll be interesting to see whether e-retail goes all the way and mirrors the physical world, or whether some new twist will throw everyone for yet another loop.

The talk of the day seems to be all about widgets. Widgets on your desktop, widgets on your homepage, widgets on your blog - widgets are going to take over the world. The funny thing is that they actually might if there is every any standardization.

I was checking out NetVibes this morning which is a site I’ve looked at previously and while the concept is pretty interesting, I just cannot get my head around having yet another page I need to check out every morning. There is definitely an advantage to having everything in one place but with tabbed browsers, I don’t find going to multiple sites that onerous anymore.

Yet the concept of being able to encapsulate everything I care about on the net in nice, bite sized chunks that I can organize as I please is really compelling. Now if only everyone playing with widgets would focus on getting it right for the user we might have something worth getting excited about.

The trend here that I think is interesting is a movement away from “web-sites” to small, information rich mini-sites that you can plop down anywhere you please. This has huge implications for branding and distribution of content (and for search actually, and advertising…you get the idea).

Consider a world where there are really no websites to speak of, but everyone shoves out their content in small increments that are very customized to the user’s preference. Interested in the Boston Red Sox but don’t care about all those other teams (ok, I live in Boston…I admit it)? Simple, grab the Boston Red Sox widget from MLB and a similar one from ESPN and drop it on your homepage, your desktop, your phone - wherever you like to satisfy your baseball addiction.

Want to show off your stock picking acumen? Snag a portfolio widget from your favorite financial news source, set it up with all your picks and attach it to your FaceBook page, your blog and again drop it onto the your phone (so you can make sure it is something you want to show off). Basically put it wherever you want (your AppleTV?)

Widgets that give you what you want and follow you where you go - no browser needed. It is a compelling concept and definitely a trend worth watching.

I had my most recent interactive eye-opener at an unlikely venue - a New Year’s Eve party at a ritzy house outside of Tel Aviv. In one of the chill-out rooms below the pounding dance floor, they had a PC set up with a chat window, and people were taking turns talking to “Alan“. Behind the PC was a glass door with a view of a room full of servers, presumably running Alan’s algorithms.

While Alan poses no immediate threat to the Turing test, I have to admit I was impressed. Chatbot technology has come a long way since Alice, though a lot of the same tricks are still in use - like changing the subject when you ask something it doesn’t know. Then again, it did remind me of the presidential primaries…

When chatbots finally get to the point where they can simulate human conversation there are many, many places where they’ll be useful. Most call center operators already rely on computer-supplied scripts to answer customer queries, it’s only a small step (conceptually at least) to replace the talking mouth with a chatbot. The bots might even do a better job than half the overseas call center operators I’ve come across.

While there will always be people who would rather pick up a phone and talk to a human, many of us who grew up with the Internet would actually rather not. If I’m surfing around a company’s site trying to figure out a problem, and their support pages (as usual) aren’t helpful, I’d much rather pop a chat window than have to pick up a phone and sit on hold for 10 minutes.