Here is a list of enhancement/features that are currently on mytweetmark. 

• User Signup, authentication and security.
• Homepage A/B tested.
• Profile page. 
• Importing address book, viral tracking and dashboard.
• Admin dashboard
• Retention dashboard
• Geo ip lookup
• Referer tracking
• Email integration
• External ip monitoring
• Memcached
• Country setup via ip lookup
• Feature toggle (on/off)
• url mapping for vanity URLs
• Search (lucene)
• Twitter api integration
• Ajax interactions
• mytweetmark brand on twitter
• Blogs
• Posts
• Bookmarks

Internal referer tracking is important to have real-time information where your traffic is coming from.  We have all learned to love Google analytics, but sometimes it's too slow.  If you need to know right away the impact of your campaign, you need to have real-time information, captured via dashboard.  I am a great believer of metrics.  You can find a lot of interesting patterns, and see things way before they will actually happen.  Complicated metrics presented in simple terms is an art.  Because at the end of the day, it's your investors and founders that have to see them and make important business decisions.  The more the metrics, the more visibility into the business.  However, most startups don't qualify to have metrics because you need to have traction before applying any real metrics rules. 

For mytweetmark, every time the mytweetmark link is clicked, the referer tracking will tell right away whether google, bing, facebook, twitter, mashable, techcrunch, any location's viralness is accessed and the impact of the campaign.  Lots of useful analytics can be built around this fashion.  It would also be very interesting from this traffic, mapping CTR to viral factor since every sign up or registration counts into the viral factor.  This is definitely something to be investigated and built into analytics in the future.

Here is a simple snapshot of mytweetmark internal dashboard screen fo referer tracking which is only accessed if you are an admin user.

The following charts will describe viral and retention internals within mytweetmark.  Also, imagej was recently installed for resizing imaging.  It is a much better solution than imagemagick.  imagej runs within your app server space memory via java, and is the fastest imaging technology in Java!  There are a couple of algorithms that need to be mixed together to get the right 'social imaging' best for photos of people.  I feel that's what social networks need where most pictures are of people.



Above is viral factor which can be dissected in any way, shape or form.  Some very useful information can be found and patterns can be discovered so your organization is not wasting empty cycles.


Retention is also key because how the users consume the offering determines who pays the bills.  Please see my other blogs on how to work with viral factor and retention co-efficient.

http://www.mytweetmark.com is up!  It took some grinding development effort, but the site is up and running.  The most fun part was developing the key components like user module/authentication, search, profiles, importing address books, viral, retention, administration and tracking dashboards.  Currently, ad campaigns are running to measure the brand name on google, facebook and bing.  The internal referer tracking by ip and address helps to see where the traffic is coming to the site which is essential for starting any new business.  The click rate from ad campaigns need to be optimized.   

Follow us on twitter, www.twitter.com/mytweetmark

Cheers,
mytweetmark.com

 

Measuring retention is necessary to figure out user traction.  There are a few questions to ask:

1.  How many people registered in 30, 60, 90 days?
2.  How many came back to the site?
3.  Gather retention co-efficient by dividing registered with logged in.
4.  How many logins per user in those days?
5.  How much content?
6.  How much social graph?
7.  How much communication?

The key ingredients for making a successful viral factor:

1.  Email domain.
2.  Country.
3.  Age.
4.  Gender.
5.  Language.
6.  How many address in the address book?
7.  How many invites sent?
8.  How many opened?
9.  How many clicked?
10.  How many converted?
11.  If the child converted sent is bigger then the people that signed up, you got a viral factor > 1 which means you are in growth mode.
12.  If you don't have a viral factor > 1, then you need to analyze your content and see whether that justifies your existence.  You can measure that by how much content per user per time frame is being produced. 

    People ask what is viral?  Viral is the ability to add more value to an existing business in the form of people sharing the service with their connections.  If the viral growth is over 1, that means that the growth is in full affect.  To simplify for example, if a set of # users are inviting more than their # of users, you have a score more than 1.  Calculating viral formulas is a science, and every business needs it.  However, the formula has to be tailored to what business it's applied to.  For that, first data is calculated accurately and presented.  From there, patterns are observed and a strategy needs to be developed.  The strategy is applied and the measured factors are observed.  Sometimes, the duration of the factor is very important.  For example, if you are sending out invites, you need to wait a couple of days for the users to receive those invites before taking their statistics seriously. 

    Retention is also key, because people focus so much on viral and end up spamming the crap out of their users.  This creates a disjointed affect in the space, and people feel that they receive too many emails for no reason.  Affective email strategy is key.  Timely emails can not only increase retention, but also viral growth.  Retention can simply be measured by looking at a cohort, i.e. a group of users and determine how many times they are coming back to the site.  Retention can be measured not just by users, but also content, for example how much content is being published per month.

    Finally, it's key to have a good grasp of your engagement metrics.  This is the ability to drill down in various measurement categories, and develop a combined score.  For example, for a social networking site, the amount of users, photos, friends, content, time spent, etc. are all key metrics that need to be paid special attention to.  If a user wants to upload a photo album and the service fails, chances are that they will upload the photos to a different site and in the future, continue to go to that other site. 

    To run a successful company, all the above areas should be monitored very carefully and historic data should be looked at for trends.  I highly recommend the use of tracking tables and then build summary tables.  Groovy on grails is excellent for Java developers, that want to quickly look at database tables and build dashboards.  Google charts is also affective way to display charts, without spending a tremendous amount of time learning charting tool kits.