Posts Tagged ‘mashup’

Can a cool search mashup be more valuable than Google?

Wednesday, May 13th, 2009
Google mashup - only with google sources?

Google mashup - only with google sources?

Today I’ve been asking people about my idea for a search mashup that can be more valuable than Google. Can it really be true, why do I even dare to be so bold? Some considerations:

  • The amount of popular Web 2.0 frameworks are growing continuously - people like online tools for certain reasons. Google doesn’t know what your favorite Web 2.0 tools are, or else Google might provide a more personalized ranking?
  • It takes time to keep track of all interesting changes in all those frameworks, personally I like Google, Delicious, Slideshare, Twitter, YouTube, Facebook, Linkedin. And I just know that I should spend more time on LastFM or Flixter, but I just don’t have time to really check all the cool things that are available. So can a mashup save me time and allow me to see more interesting content at the same time?
  • It would be even more interesting if I can mix it with my favorite websites, even if they are behind a login. Can I have a search through our wiki / jira in the same mashup?
  • Can the search be more smart and allow me to organize my interests into ‘parts of life’ like work, research, travel, sport, family, friends, movies …?
  • And can the search be so smart to also move along with me through time - during the week or weekends - morning, afternoon, evening, night?
  • Can the search already get me started with the preferences that I have available in facebook, lastfm, delicious, wordpress or tripadvisory?
Google is so good nowadays …
Google is so good nowadays, we don't need the Top 10 anymore - the first 2-3 are enough

Google is so good nowadays, we don't need the Top 10 anymore - the first 2-3 are enough

The picture above compares how people click on Google - 2005 and 2008. It shows that we now trust the first 2-3 results and we don’t need the Top 10 anymore. With mashups you can have the first 2-5  results of all your favorite Web 2.0 frameworks.

A basic ‘Your search’ mashup

I think we can make a first start from the MarCom mashup to a basic ‘Your search’ mashup with the following features:

  • first ask me to identify with either my Facebook, Gmail or Hyves account
  • it should use the same panels like the MarCom mashup (twitter, delicious, youtube, slideshare, amazon, google search, google search - favorite website)
  • the panels now listen to a ’search keyword input’ or a ’search cloud’ panel instead of a pre-configured tagcloud panel
  • if the user has identified with the Facebook or Hyves account, then use the relevant interests metadata from Facebook or Hyves to generate a starting search cloud
  • the search cloud will remember new search key words that are used
  • the search cloud will only show 30 most often used keywords - if there are more - the ones that are least used will not be shown
    • the keywords and related counts are not thrown away, they’re just not shown in the search cloud
  • the search cloud should offer a possibility to show all tags with a link ’show all tags’
  • a user should be able to save a search cloud with different names (eg work, private, travel, sport, cars, etc.), a max of 25 search clouds should be sufficient I think
  • a user should be able to ‘hide’ tags that are not relevant and save the tag cloud under different name
    • so actually there is actually 1 big list of search keywords for each user
    • but through different ‘names’ the user can chose to organize different views of the search clouds

The search mashup app would be cooler if

  • if the panels could be rearranged in size / length
  • the layout would be a ‘click, drag & drop’ to allow users to define their own mashup combination to show results
  • for each view of the search cloud a different organization of the mashup can be remembered (used panels, position, sizes, lengths)
  • if a user can also import tags from:
    • LastFM (favorite artists)
    • Delicious
    • Google bookmarks
    • Wordpress (used tags)
  • If the searchcloud could use variation in density of colors to show how ‘recently’ the search keyword has been used
    • a dense but small keyword means a new interest, but recently often searched
    • a light but small keyword means that it’s been a big interest in the past
  • If the searchcloud could somehow show directly how many results are found in each seracheable Web 2.0 frameworks before I click on it
    • this can be cached every hour?

If we do it for ourselves … then we can believe it’s also cool for others
Personally I know that I’m a reasonable Web 2.0 poweruser and with me I know that Componence team members are above average Web 2.0 users. But statistics show that there are millions more out there that use different combinations of Web 2.0 tools. But they probably all have the same isse, they want to know / see more - but have less and less time to browse the internet. So if the mashup we make is really usefull for us and we really see advantages of it above Google search, then it will be time for us to launch the app into Facebook to see how many more user internationally will find the app usefull.

Mashup example - LastFM in Facebook with Google Maps & Amazon

Saturday, April 4th, 2009

The first Componence mashup example for the world to see

At last - Componence has achieved a major goal at the end of this first quarter! Our first example mashup is ready! It has been a about a month ago when we began to work on the LastFM - Facebook mashup. And after a month, with approximately 500 - 600 hours of work, we’ve put it to a rest. The result:

  • Different demonstrations to music producers showed interest, the producer of The Dream might be interested.
  • Mubito probably will do a Proof of Concept with us in May - we should add their APIs to it.
  • A demo of our new Chameleon technology that I can easily use & show in our sales. The goal from last summer was to have this at the end of Q1.

The background of this mashup
The idea for this mashup was triggered by this e-mail response of Yannick De Rechapt, CTO of Mubito:

” What type of applications do you want to expose into the communities?
- Basically expose content such as, audio, video, member data, feeds, shop widgets with search based on our catalogues + single cart etc… this is a non exhaustive list.

What kind of communities are you thinking about?
- Again I don’t have an exhaustive list but : flicker, myspace, imeem, lastfm, facebook and so and so

Although they had switched their WLP platform to a PHP platform, Yannick was keen about our Chameleon technology, that would allow him to make widgets to be posted to different social networks. We just took it a little bit further …

Not enough real data
Before I want to tell about what was mixed & matched, here is a list of everything that is hard coded due to lack of time / real data:

News not yet implemented

News not yet implemented - it can be a combination of different news sources and Twitter

Photos & Videos - no reliable real source - can be user generated content or Webservices

Photos & Videos - no reliable real source - can be user generated content or Webservices

The mashup ingredients
The main source of the mashup is LastFM, but the mix might make it more valuable for end users.

  • LastFM
    • User preferences / top artists, artist data, event data
  • Google Maps
    • Allowing us to show the events in a map
  • Amazon
    • Allowing us to connect e-commerce to the app

Why Facebook as first choice?
The combination above has been implemented in a Facebook look & feel for the following reasons:

But if this continues into a real project, of course the app would also be available in other communities.

Value for potential business?
Although we don’t know the music industry is not our own comfort zone, I do think this kind of application can help them to deal better with their pains:

  • Revenues - due to illegal downloads - their revenues are continuously under pressure
    • The theory is that if the mashup is valuable enough, the user can take it along into their comfort zone (community, desktop, mobile). And then probably we might get fans to buy albums / merchandise more easily. From once a year to twice a year can already be a huge increase in revenues.
  • Better cost / benefits
    • An insider told me that the music industry are like dinosaurs, they’re so far behind on technology. Just look at the websites of most artists, most of them just don’t really appeal for engagement / interaction with endusers. A mashup can incorporate many popular, of course Twitter on top, webplatforms that are already favorites for user generated content (YouTube, Flickr, Picasa, blogs). If there is better interaction, there are more chances there are sales.
  • Easier way to engage and keep fans happy
    • By being in their comfort zone, artists can consider giving out special freebies to fans who have this app in their profile. Just think of live webcam feeds of concerts? Or free downloads of the latest single within the first 24 hours? Or a sweepstake contest with the ultimate prize being a private appearance.
  • New business model - different subscriptions
    • Silver, Gold, Platinum? With maybe just $1-$3 dollars a month, you can get access to more cool features. With the ultimate privileges like access to special webcam chats, or another free download, or the details to secret appearances or even invites to hangout with the artist.
  • More prolonged return for investments
    • For any artist - starter or already a star - eventually millions is put into marketing. How much return is their in an ad in a magazine or a billboard? How the return of those investments be prolonged? I believe that such offline promotions can easily be connected to the online Fanbase.
    • I think if a fan spends $50 for a single concert and merchandising, by giving discount coupons at the concert to the online shop and special access codes to behind the scenes reports, we might get the fan to spend another $20 online in the week after the concert. And if we’re really good, the fan will make use of the special discount on the Gold membership, that can last for years.
    • Personally I think a really cool mashup would cost between $50.000 - $100.000. With 10.000 fans - online sales between $5-$10 per fan would already be enough. But what if there are 25.000 or 100.000 fans?

Standard Chameleon panels ready
With this mashup we already have different panels that can be easily reused for our other mashup projects:

  • Check login with LastFM
  • Retrieve & show favorite artists from LastFM
  • Retrieve & show Events list from 1 artist from LastFM
  • Retrieve & show Details of 1 event from 1 artist from LastFM
  • Search & show results of artists from LastFM
  • GoogleMaps to show events from LastFM
  • List of related downloads from Amazon

We might come up with cooler names for the panels, but for now, we’ll use the word ‘Panel’ from the Wicket technology - the base of Chameleon technology.

Why our Chameleon technology is really cool
Although the above sounds not to difficult, the real reasons why our Chameleon technology is so cool:

  • The panels can function as different applications - and listen to events from each other
  • The whole mashup can be shown in different comfort zones
  • Each panel can be configured to become a widget that can function in different environments (Java portal with JSR-286, iGoogle, Facebook, OpenSocial, iPhone web app, gadgets on your MacOSX or Vista

Yesterday Eric van den Wildenberg, Director at Brand republic called our technology trendsetting! And he invited me to give a presentation to the whole group of Central Park.

More mashup following
To show and proof the value of mashups, we’ll implement a few more mashups in the coming months:

  • April: social media marketing mashup - easily invite & share from different social networks
  • May: knowledge management mashup - combine different media sources about a specific topic
  • Summer: some more … to come