Visual search seems to have come and gone. In 2004, Groxis released the colorful Grokker search client and received $12 mil in funding. Apparently, the consumer market didn’t more than nibble at the innovation; Groxis has pulled the product and refocused on the enterprise knowledge management market.

KartOO is a French company with a skillset somewhat similar to Groxis’. They sell turnkey search engines for the internet and intranet, information visualization, intelligence monitoring, and geomapping software to the enterprise. They don’t seem to have – or at least don’t anymore – consumer ambitions. However, they do have a fully functioning public showpiece of their search visualization technology.

(Here's the first level search results for "Aeon Flux" with the theme "animation" hovered over:)

It could very well be that Grokker and KartOO are too ahead of the mass market and that with time the list of textual results we accept today will seem anachronistic. So even if not popular, the search presentation innovations of today are worth examining as sketches for the future. Google is already changing the results page by partitioning the results into functional categories such as images, news, blogs, etc.

KartOO is a meta-search engine that like Grokker, groups the results into themes or concepts. Grokker represents results and their sets as concentric and zoomable colored circles. KartOO – a rich flash application - draws concept maps: a set of concepts are overlayed onto the same map as the search results (a dozen or so results per map). Hovering over a concept activates colored lines connecting all relevant results – weblinks - to the concept. Hovering over a result activates colored lines connecting the result to all relevant concepts. This idea is basically isomorphic to tags and tagclouds but done on a map. But here, the tags are machine generated.

Clicking on a result obviously opens the webpage; less obviously, clicking on a concept “explodes” that subset of results, ie KartOO seems to recursively apply the same main level algorithm to the sublevel.

KartOO’s results may look cartoony compared to Grokker’s fractal-like spectacle, but KartOO is a lot more concise because each result is depicted only once on a map and more information dense and space efficient whereas Grokker seems to revel in redundancy by finding excuses to make circles out of different concepts which often contain the same weblinks; also zooming in Grokker is purely an optical effect whereas clicking on “themes” in KartOO really does expand the results; and you can keep drilling indefinitely into the results.

User-generated advertising has become a hot space lately. Sony, Toyota, L'Oréal, Converse, MasterCard, Chevrolet have or are flirting with user-generated ads. Comcast Interactive Capital and Turner Broadcasting invested in ViTrue, "a startup providing technology to enable user-generated advertising campaigns."

Yaadz joins the space, announcing on Oct 31, 2006 a service specializing in user ads. Corporations have held their contests either on general video sites like Current TV or on their own sites like Chevrolet Apprentice (contest over so now redirects to Chevrolet sales). Yaadz would be the first site to exist expressly for user ads. Yaadz is a French venture but certainly they can host English language ad campaigns. If they fail to bridge the language or culture divide, at least the model can serve as an example for someone considering the idea in the US.

Here’s my cleaned up translation of Yaadz’s main objectives:

(I cleaned up the English. You'll have to read the French for the original intent)

The Yaadz's Concept in 3 Points

What Is YAADZ?

YAADZ proposes that the advertisers entrust the creation of their videos viral to the Net surfers. YAADZ holds contest all Net surfers in which can take part. YAADZ focuses on those advertisers seeking to reach the new social networks, videos site or blogs. YAADZ proposes completely original and creative content.

How YAADZ contests work?

To launch a YAADZ CONTEST, the advertiser proposes a brief (short scenario of the video) to the community of the creative users registered on YAADZ. The creatives can register with the YAADZ CONTEST and can then submit their videos. The advertiser announces the winner at the end of the contest. The contests can be public, the members of the YAADZ community can then take part in the selection of the winner as voters.

Are contestants paid?

Yes! Prizes are paid on YAADZ. The awards start at $1,000 (presumably EURO not dollars) and can go up to $10,000 per winning video. Other prizes are also awarded according to the success of the video.

This is a very interesting concept. Corporate contest sponsors have had to co-ride programs like the Apprentice or advertise the contests on their or other websites, traditional media or indirectly through PR. YAADZ proposes to be a contest nexus, a YouTube for contests. In contrast to YouTube's increasingly becoming a generic junkyard for all kinds of content, a well-focused contest site can thrive. Corporations, creatives, and interested viewers could have one central place to go for user-generated ads, and the winning videos can be distributed virally to other sites. The million dollar question is how well YAADZ executes. Key issues will be signing up corporations and getting sponsorship for juicy contests. Long term issues will be whether companies will want to commingle their marketing efforts in one location, outside of their comfortable open silos and whether YAADZ can draw a breadth and continuity of sponsorship to make the site worth returning to. Given the enormous marketing budgets of large corporations, even sponsoring one contest a week is a piddling sum. The ROI on a viral winner is huge. The downside is limited by the selection process. Ad agency created advertising sometimes seems too slick and misdirected and it would be nice to see a platform for the creation and viewing of user ads. Let's hope for YAADZ' sake, enough corporations demonstrate boldness.

In my continuing effort to scour the globe for new internet ventures, I present you SubMate, a pal finding service for commuters. Submate is still under wraps and apparently based in NYC though they have ambitions of offering the service in “9 countries and 20 cities.” One of the co-founders is Laurent Kretz. SubMate was supposed to beta test this summer in NYC or Paris but that hasn’t materialized yet. Whether the slippage is due to the unnamed big competitor unveiling a related service in their home market or the typical delays that plague startups is unknown.

In a nutshell, SubMate allows users to create profiles and enter their travel routes and use that information to find others to befriend on (or off) their commutes. The matching can be done algorithmically or by browsing profiles. They founders pride themselves on having a rich feature set and social networking to differentiate SubMate from original dating services, including the emerging related category of “missed connections” sites.

Let’s jump into positioning issues since a beta isn’t available to demo and marketing is most interesting anyway.

What is this animal? Essentially SubMate exploits an arbitrary context for people to connect. Many services do this creatively:

It's Just Lunch - Lunch dating
It's Just Coffee - Coffee dating
Dinner at 8 - Group Dinner dating
Table for Six - Group Dinner dating
HurryDate - Group Speed dating
DogMeet - Dog dating / networking
MeetUp - Group interest networking
Craigslist - Missed connections (regret dating)
terevoire - Missed connections
Dilelui - Missed connections on commute

Read the rest of this entry »

A few weeks ago I covered the new cohort of social activities sites. In the subcategory of basic events listing services, notably comprising Upcoming.org, Zvents, and Eventful, I neglected to include BusyTonight. Ironically, I missed it despite the presence of an active blog on which developers write thoughtfully about the philosophy and execution of the service. BusyTonight is a subsidiary of privately held Team Gigabyte and based in NYC. The CEO is Joshua C Lerner, COO is Richard Mintz and CTO is Matt Kangas.

BusyTonight differentiates itself from the other services by targeting the big mass of less popular events would go otherwise unlisted, the Long Tail. Everyone will list a Madonna concert and likely some or all will miss a lacrosse team fundraiser at the local community college. That’s because they all subscribe to the same duopoly of commercial sources, EventSource and EventCrazy, which naturally capture economically incentivized events, and rely on the imperfect methods of pulling calendars, eg Google Calendar, and RSS feeds and gathering user contributions for the remainder of events. These methods are imperfect – imcomplete and slow - because they require a mass of individual effort. Some of the “loose” events may be captured this way, but the majority will sit unreplicated in their home web pages.

Read the rest of this entry »