Archive for April, 2008

Finding obscure great bloggers and mashup

April 29, 2008 1 comment

As I continue playing with the small application that I’m writing for monitoring positive shifts in bloggers’ Technorati rank I realized that I’m actually finding bloggers writing about almost everything. The only common thing I could find so far is that they are just consistently great.

The tool scans and builds historical data for over 700 blogs so far. I build this growing blogs’ URL list using my favorites (i.e. humanly picked in multiple social ways) and the crawling algorithm I previously explained in this post.

I won’t get into the operation details (and there are plenty of details) but I mange to get a lot done not exceeding the 500 API calls daily Technorati limitation.

I output the result to BlogMon Twitter user for now so please, you are invited to be a follower.

Example of outputs:

Short-term pattern:, rank gain: 18.10 %, since: 4/12/2008, Top Tags: “wordpress”, “themes”

Long-term pattern:, rank gain: 76.10 %, since: 2/1/2008, Top Tags: “Social Media”, “Security”

As you can see I log the URL, the rank gain, since when, and the top two tags to give you an idea what this blog is all about. I found that in most cases this is good enough. Do you?

Why am I doing this?

  • First, it keeps me engaged with the mashup opportunities and there are lots of those available today .
  • Second, I enjoy doing it.
  • Finally, you may find it useful in some way – you can leave a comment on these blogs and maybe get some traffic to your website/blog. I will be happy to hear if you did.

I may be tempted to mashup more web data sources/services in the future or explore discrepancies between Alexa data and Technorati rank .

I’m also using a great early stage service developed by Microsoft called Popfly to build and deploy a small (too small and simple at this time) application to my Facebook profile called BlogTwitt. BlogTwitt will show the recently posted updates to the BlogMon Twitter user I use for outputting the daily findings from the application I’m working on.


At this point I could not share this application – I don’t know why so I left a message on the Popfly Facebook wall. As of this time I got no answer. I do appreciate what that they are trying to do, saving me the time learning/working with the Facebook API.

I think that I will write soon a post about the Popfly and the challenges writing a good mashup. I do encourage people that are just starting their mashup thought process to look at this tool and also at Yahoo pipes (fantastic interface) to play, understand, get ideas and brain-storm with the numerous available web services (API)  out there. This is like working in a software solution architect group for a company that offers multiple products and  findings new way to increase the value of the existing modules by symbiotically integrating them to new offerings.

Finally I don’t think that this is Software plus Service like Microsoft tries to sell it I see it as Service plus Service (the service is build of software, da). Maybe Service X N.

As always, I would love to hear your thoughts so please use the comment section.

Update: I forgot to mention that what that I like about using Twitter vs. my blog to post results is that it does not  add to the blog reactions count. So, it goes under Technorati radar and does not impact the Rank (avoiding the Observer Effect). That may change one day when they will realize that Twitter’s twitt with blog’s URL is actually a blog reaction.

Alpha, Beta, Launch – whom, where, when?

April 14, 2008 3 comments

It seems like the world of web site in alpha and beta phase is booming. Invites, private or public, spreads all over the web-sphere, help getting new online service a try prior to going live, .

I’m looking for a web site that shows what products are now in alpha or beta phase. A place where you can see what service opens its product for the small or large crowd. If you know about this kind of service please let me know in the comment below. In the meantime let me fantasize how it should work.

Every start-up that is working towards releasing a limited or general availability of his product and like to get people involved in testing, providing feedback and suggestions should fill up a small registration form on this web site.

The information should include:

  • The name of the service
  • The company information
  • The URL
  • A short description of the service (try to fit it to a listed category or create a new one)
  • The development cycle phase: alpha, beta, limited availability (e.g. for only English language support)
  • The method: private (#of invites) or public
  • Phase starting date
  • Estimated phase end date
  • Any other relevant information (open text section)

Some features:

The service should provide the company with login to be able to update the development status.

This site will list all the companies and their product information

This site should allow search

This site should allow ordering the products under testing by date

The site could aggregate blog posts and news relevant to the company and product (if you choose to drill down).

Now in the spirit of everything social (aka Web2.0):

  • People can rank, and comment on a specific product
  • Subscribers can offer their service as alpha and or beta testers to available products in a chosen category (they can get alerts from the web site that something in their domain of interest/expertise just listed) .
  • The company can invite its devoted testers to celebrate the launce
  • The company can rank different testers based on their feedback, bugs, creative suggestions
  • And like any good web2.0 service why don’t invite/follow your testers friends or the actual entrepreneurs to share something

The site can follow up with some success story and news (RSS feed, newsletter).

What is the value that this kind of service provide? To whom?

  • VCs –  some visibility into their current or future investments
  • The start-up – visibility and access to experienced resources that can help improve a premature service
  • The tester – finally someone can see who we are:)
  • Business development – visibility for opportunities to join forces and integration
  • Recruiters
    • positive: convincing someone to join a start-up that is making good progress
    • negative: picking fallout
  • Job seekers – knowing what company is looking for employees and some information about the place you’re may end up at.
  • TechCrunch – it may save Michael Arrington some cycles and allow him to rest a little bit:)

I did a very quick search and all I could find is this site. Again, if you know about a place tracking this kind of information in an organized way please let me know.

Even if you don’t I would like to hear your thoughts? Do you agree with me here? Do you see a need?

The three pillars of Social Search Engines

April 12, 2008 4 comments

Inspired by TechCrunch article listing multiple social search engines solutions I decided to check few of the services claimed to be (or categorized as) social search engines. I tried: Rollyo, Eureskster, RecommendBox , iPercisnsyght, and andunite. There were few other names (from this source) that were mistakenly categories as social search engines.

In my previous post I wrote about blog searching and Twingly, a company that leverage the community for building spam free search engine and improving search results. By the way this company is very high on The Search Race list. 

To my surprise I found that most of these solutions lack in the one or two key features of social search engines: building great content and expending the community. After examining each of these services and looking at their process, mainly in the later two pillars of social search engine: content, and community, I yet to see a social search engine that made it simple enough to attract large audience and build an eclectic content base. If I was missing a great existing solution, in this short quest, I’ll be happy to hear about. I’m looking for a Google quality like implementation in each of these pillars.

It is not enough to provide a simple way to upload one’s eclectic web links to the system, it is also important to provide a way to expend the content in the same social fashion. It is not enough to provide the option to  add friends or search in one’s contact list(not that I’m too excited about this method), it is also important to help a person to find new friends; expending his network and at the same time the search engine input base. 

Pillar I – The search

I’m not an expert in search algorithms, and it is not the goal of this post to comment about this part of SSE implementations yet it is assumed. I will say that today’s concept of search expended beyond finding a word or a phrase in text to words, and idioms spell check  (did you mean?), direction (map) and business basic info.

Pillar II – The content

There are multiple ways that SSE collect and refine its content

The interactive way:

  • Allow you to import your favorites
  • Allow you to submit new links and tags from the search web site
  • Allow you to submit new links and tags through a gadget on a web site or blog
  • Allow you to submit new links and tags using a button on your web browser
  • Allow users to add tags, comments and vote for search results

The offline (automated) way:

  • Crawling the web looking for fresh links in addition to the interactive way – there is an argument whether SSE should do that and I can see both point making sense- pro: allow for fresh content to support complete search experience, con: more is not always better and mainly when it looks and feel like spam.
  • Crawling the communality submitted content to find matches

Why stopping at the submitted content only?

If I may suggest use the community submitted content as a source for crawling kind of how Twingly does it. I would also ask the community permission to access any of their other Web2.0 services to find more content.

Services that I found doing it fine:

nsyght – no option for favorite import but allow you to get your links form delicious and few other sources, iPercis – I successfully imported my favorites, yet requires some usability improvements (select all in some palaces)

Services that I found doing it poorly:

Rollyo – failed to import my favorites (single folder)

RecommendBox –  no assistance or automated way to build recommendation

Pillar III – The Community

This was my biggest surprise I assumed that a social search engine will put his emphasize on the social aspect of the application. I found some cases where after registering to the site “I was left alone in the cold”. I had no way to expend my network.

There are multiple ways helping people to do grow their network.

From within:

  • Show a page with the current people logged-in to the web site and a view of everyone else.
  • Allow adding people as a friend or preferably as I wrote in here – use the Twitter “follow” model

New users:

  • Allow the user  to import her contact list
  • Allow the user to grab his friends from her other social services
  • Social crawling – look on other social services for leaders with many contact and lure them to sign in

Nobody has to use any of these automated options if they don’t want to to but at least they could add friends easily if they do.

Services that I found doing it fine:

The best one so far in this area is nsyght – they implemented most of the options listed above.

Services that I found doing it poorly:

RecommendBox  – there is no way to add friends from within. There is no automated way to add friends that are not currently using it.

Update 4/14/08 – my bad, sorry, there is a way in RecommandBox to load conatcts from online address books in the welcome page (after login in). Thank you Scott Rutherford for correcting me.

In Rollyo, and Eureskster there is only implicit notion of community but you can’t “see” it.

From my short inquiry on social search engines I  learned how wide is the range of different solutions that this term cover and I think that I found two promising services nsyght and iPercis.  iPercis is still in alpha and nsyght based on the TechCrunch post is on public beta so expect some hiccups along the way, yet both companies invested, in my opinion, in all the right element of Web 2.0 social search engine solution.

Probably  I will stay with Google and that I really  hope to see coming up with better search capabilities. What about you?

"The Blog Search" – single exit strategy? Thinking outside the search (edit) box.

I see new start-up companies still working on new search engines and search technologies. I can’t seem to see the reason why and the need for it when it comes to searching web sites. I do see a business need for new ways searching blogs, under some conditions.

Google is more than good enough for me. Most time I manage to find what that I need on Google search.

There are many basic things that Google does great beside text search like finding address, directions, maps and telephone numbers. They have built a highly scalable spiders, data center, ad sense technology and many other essential capabilities.

Do you really think that everyone can afford to build it?

The only exit strategy that make sense for these companies is one contingent upon buyout by one of the search giants like Yahoo, Microsoft, and Google.

If I’m right in this case I would suggest to these start-ups not to focus on the amount of data indexed by their new search engine because no one will ever use them under their origin brand name. They will not be able to wave with great statistics to convince someone to buy them because the chances for them to reach wide audience are slim.

On the other hand they should focus on building high technology that could be patent protected, alternatively they should find a niche or geography that Google is very weak in.

I can only offer them to walk directly to one of the large search engine companies, knock on the door and offer the new technology for sale and then to spend their effort in another technology front.


One place still worth making an effort in search technology is the blogsphere. Google is not so great over there as I wrote in here . Technorati is a fantastic blog search engine but it is mainly focused on English speaking blog readers and bloggers.

One blog search engine with promising future is Twingly. This is an example for a blog search engine that is strongly invested in other languages and have deep roots in European countries. See posts on TechCrunch here and here.

I recently signed up to the private beta and I like what that I see so far. I can’t speak much about the search quality but I can tell about few things that I liked. First is the spam free search engine approach. This is a working progress to find information based on few initial reliable sources and then spreading out using links from those known bloggers building a “white” links list to index. The second is the powerful support for multiple languages other than English. Yes, not all the bloggers are speaking English, there are  more than 140,000 registered Swedish blogs, tons of French bloggers to name a few. So there is a need. Even in the US, the big melting pot, there is a great market for multi-lingual blog search engines(like in Spanish). What that I really liked in this beta is the TechPlan section where people can offer suggestions for product improvements features and then allow others to vote on them. This is a great way to collect feedback. I can’t see why not leaving it out there, even after the beta is done. There are more capabilities like voting on blog posts, prefix a search phrase with tags like link:, site:, blog:, lang:, tag:, and tspan: for qualifying target searches.

So, where should Twingly invest their effort?

I have few ideas:

Be Cultural

Don’t try to be American. Europe has a classic culture that could be embedded in the search results. Alternatively Twingly can create countries’ specific web pages the same way that Technorati created the “what’s percolating in blogs now”. E.g. what’s percolating in [country name here e.g. Norway] blogs now?

Be Social

Find a way to bring bloggers from around the world together. Match bloggers from different places by area of interest and create a place for them to interact. Create events centered in Europe for bloggers to meet. Use the Twitter like follow model for bloggers to find one another.

Be digital – build the best Blogsphere system-of-record

As I wrote in here strive to monitor and collect any available data about blog posts, blogs and bloggers, measures, profile information, area of interest, methods, preferred media, activities, patterns. Make sure to organize the data in a way that it is useful for both the general public and businesses.

Well if Twingly grows on their own or not they still have high chances for getting on Google radar:)


Get every new post delivered to your Inbox.